Myoglobin Structure and Oxygen Binding

A sperm whale can hold its breath for over ninety minutes. It dives to depths where no fresh air is available. The reason lies in a single small protein packed into its muscle. This protein occurs at concentrations forty times higher than in a cow. Myoglobin is a 153-residue globular heme protein weighing only 17.8 kDa. Yet it does something no naked heme group can do reliably in water. It binds molecular oxygen reversibly. It does this without letting the iron atom rust into a useless ferric dead end. Understanding how it manages this is a masterclass. It shows how a protein fold turns an ordinary inorganic cofactor into a molecular machine with tunable chemistry. This matters because myoglobin is the simplest possible testbed for questions that recur across all of physiology. How does a metal center bind a small gaseous ligand selectively? How does a surrounding protein matrix modulate binding affinity and kinetics? How do we extract quantitative binding parameters from an experimental curve? The chemistry here — iron coordination, ligand-field splitting, steric gating by a distal residue — is the same chemistry that underlies Hemoglobin Structure and Quaternary Change, gas sensing, cytochrome function, and countless heme-based catalysts. Myoglobin is where you first meet a genuinely hyperbolic binding isotherm. You derive it from first principles. This happens before hemoglobin's cooperativity breaks that simplicity. The big picture is a triangle of structure, thermodynamics, and function. Structurally, myoglobin is a compact all-α globular protein. Eight helices (A–H) form a hydrophobic pocket that cradles a heme b (iron-protoporphyrin IX) cofactor. Thermodynamically, the iron sits five-coordinate in the deoxy state, ready to add a sixth ligand. The equilibrium of that sixth coordination site follows the simplest possible binding law. It is a rectangular hyperbola in partial pressure of O₂. Functionally, two histidines — proximal (F8) and distal (E7) — do essentially all the chemical work. One anchors the iron and tunes its electronics. The other polices the distal pocket, allowing O₂ but discriminating against carbon monoxide. This article builds the full mechanistic story. It covers how heme is held by the tertiary fold. It explains why iron must stay ferrous. It explains how the proximal and distal histidines exert their opposing effects. It shows how the single-site binding equilibrium is derived. It also shows how this equilibrium is fit to real P₅₀ data. It closes by previewing why hemoglobin needed to break this simple hyperbola. Hemoglobin is built from four myoglobin-like subunits. This break was necessary to achieve cooperative, allosterically regulated oxygen transport.

Key takeaways

Myoglobin is a single-chain, all-α globular protein (153 residues, 8 helices A–H) built around one non-covalently bound heme b cofactor.
The heme iron is five-coordinate in deoxymyoglobin: four pyrrole nitrogens plus the proximal histidine (His93, F8); the sixth site binds O₂ reversibly.
The distal histidine (His64, E7) hydrogen-bonds to bound O₂, stabilizing it and sterically discriminating against the toxic, linear-binding ligand CO.
Fe must stay Fe(II) (ferrous, high-spin ⇌ low-spin) for reversible O₂ binding; oxidation to Fe(III) gives metmyoglobin, which cannot bind O₂.
Oxygen binding to myoglobin is a single-site equilibrium, producing a rectangular hyperbola in fractional saturation Y vs. pO₂, characterized by one number, P₅₀.
The Hill coefficient for myoglobin is n ≈ 1 (no cooperativity), in sharp structural contrast to hemoglobin's tetramer.

Key idea: A single heme cofactor plus two strategically placed histidines convert a simple Fe–O₂ coordination equilibrium into a controlled, reversible, hyperbolic oxygen-storage function.

Where this lives today

Myoglobin remains one of the most important small proteins in modern biochemistry, medicine, and biotechnology, well beyond its textbook role as the simplest oxygen-binding model. Clinically, serum myoglobin is used as an early biomarker of myocardial infarction: cardiac muscle damage releases myoglobin into the bloodstream within 1–3 hours, faster than troponin, making it useful (in combination with troponin) for rapid emergency-department triage of suspected heart attack, even though its lack of cardiac-tissue specificity limits it as a standalone diagnostic. In diving physiology and comparative biology, myoglobin concentration is a direct biomarker of diving capacity; genomic and biochemical studies of cetaceans, pinnipeds, and diving birds show convergent increases in myoglobin surface charge (reducing self-aggregation at the extremely high intracellular concentrations needed for long dives) as one of the clearest documented examples of molecular-level convergent evolution. In biotechnology, engineered myoglobin and myoglobin-inspired heme proteins are used as scaffolds for designing artificial metalloenzymes: mutating the distal pocket residues repurposes the heme active site for peroxidase activity, carbene- and nitrene-transfer catalysis, and even abiological carbon–silicon or carbon–boron bond formation, exploiting exactly the steric-gating logic described here but redirected toward new reactivity. In food science, cured-meat color chemistry is myoglobin chemistry: the characteristic pink of cured ham comes from nitrosylmyoglobin, formed when nitrite converts Fe(II)-myoglobin's ligand to NO, while the brown color of aged or poorly stored meat reflects metmyoglobin accumulation — the same Fe(II)/Fe(III) distinction discussed above, now visible on a supermarket shelf. And pedagogically, myoglobin's clean hyperbolic binding curve remains the essential reference point against which the sigmoidal, allosterically regulated behavior of Hemoglobin Structure and Quaternary Change is defined and measured, making it the indispensable baseline for the entire field of Cooperative Binding and Allostery.

Ask the tutor: How does hemoglobin's quaternary structure convert four individually hyperbolic subunits into one sigmoidal binding curve?

The physical picture

Myoglobin's fold is the textbook example of a globin fold: eight α-helices, labeled A through H, connect via short loops into a compact, roughly spherical globular domain with almost no β-structure. The helices pack around a hydrophobic crevice between the E and F helices, and it is in this crevice that the heme group sits, held in place almost entirely by van der Waals contacts with nonpolar side chains, plus one crucial covalent-like coordination bond from the protein to the metal. The heme cofactor itself is iron-protoporphyrin IX: an iron atom chelated by the four pyrrole nitrogens of a porphyrin macrocycle, which itself carries two propionate side chains (solvent-exposed, often forming salt bridges with surface lysines/arginines), four methyl groups, and two vinyl groups packed into the hydrophobic pocket. In the isolated cofactor, iron in the porphyrin plane has four equatorial N ligands; the two remaining coordination positions (axial, above and below the ring) are open for additional ligands. This is exactly what the protein exploits. Two histidines dominate the chemistry of the pocket. His93 (F8), the proximal histidine, donates its imidazole Nε2 lone pair directly to the iron as the fifth (axial) ligand, tethering heme to the globin scaffold and, crucially, tuning the iron's electron density through its imidazolate character. His64 (E7), the distal histidine, sits on the opposite face of the heme, not bonded to iron but positioned to hydrogen-bond to a ligand bound at the sixth coordination site. In deoxymyoglobin, this sixth site is vacant, water occupying it only transiently; iron is five-coordinate and slightly domed out of the porphyrin plane toward the proximal histidine, a geometric signature of high-spin Fe(II). When O₂ binds, it coordinates end-on and bent (Fe–O–O angle ≈ 120°) at the sixth site; the distal His64 donates a hydrogen bond to the terminal oxygen, both stabilizing the Fe–O₂ adduct and — through simple sterics — discriminating against the linear, high-affinity binding geometry preferred by carbon monoxide. This is the celebrated "steric gate" that keeps myoglobin from being poisoned by endogenously produced CO, and it reduces mismatch of CO affinity from a free-heme value near 20,000-fold over O₂ down to roughly 20- to 200-fold in the protein — still favoring CO, but survivable.

Figure

Myoglobin fold with heme pocket between helices E and F

Figure

Coordination sphere of heme iron: deoxy vs oxy states

The core ideas

Oxygen binding to myoglobin is described by the simplest possible ligand-binding equilibrium: one binding site, one ligand, no interactions between sites (there being only one site per chain). The equilibrium is

\text{Mb} + \text{O}_2 \; \underset{k_{\text{off}}}{\overset{k_{\text{on}}}{\rightleftharpoons}} \; \text{MbO}_2

with an association constant

K_a = \frac{[\text{MbO}_2]}{[\text{Mb}][\text{O}_2]} = \frac{k_{\text{on}}}{k_{\text{off}}}

where:

$[\text{Mb}]$ , $[\text{MbO}_2]$ — concentrations of deoxy- and oxy-myoglobin
$[\text{O}_2]$ — free dissolved oxygen concentration, proportional to partial pressure $p\text{O}_2$ via Henry's law
$k_{\text{on}}, k_{\text{off}}$ — bimolecular association and unimolecular dissociation rate constants

Because dissolved O₂ concentration is experimentally tracked through gas-phase partial pressure, it is conventional to write the equilibrium constant in pressure units, defining the dissociation constant $P_{50}$ directly as a pressure:

Y = \frac{p\text{O}_2}{P_{50} + p\text{O}_2}

where:

$Y$ — fractional saturation, the fraction of myoglobin molecules with O₂ bound (0 ≤ Y ≤ 1)
$p\text{O}_2$ — partial pressure of oxygen in the surrounding gas or solution (torr or kPa)
$P_{50}$ — the partial pressure at which exactly half the myoglobin is oxygenated ( $Y = 1/2$ ); for sperm-whale myoglobin, $P_{50} \approx 2.8$ torr at 20 °C

Reading: this is a rectangular hyperbola — Y rises steeply from zero, curving smoothly toward saturation, with no inflection point. It reflects the fact that O₂ binding to myoglobin is not cooperative: there is only one binding site, so there is nothing for a bound ligand to "communicate" with. This is the essential structural contrast with hemoglobin, whose four interacting subunits produce a sigmoidal curve instead. A second core relationship, the Hill equation, is used to characterize (and, in myoglobin's case, to falsify the presence of) cooperativity:

Y = \frac{(p\text{O}_2)^n}{P_{50}^{\,n} + (p\text{O}_2)^n}

where:

$n$ — the Hill coefficient, an empirical measure of apparent cooperativity
$n = 1$ recovers the single-site hyperbola exactly; $n > 1$ signals positive cooperativity (as in hemoglobin, $n \approx 2.8$ –3)

Reading: for myoglobin, fitting real oxygen-binding data to this equation returns $n \approx 1.0$ , confirming — independently of structural knowledge — that the sites behave independently, exactly as expected for a monomeric single-heme protein.

Derivation AFrom elementary equilibrium to the hyperbolic binding curve

Goal: derive $Y(p\text{O}_2)$ from the elementary mass-action equilibrium for a single binding site, and show it is a rectangular hyperbola with one adjustable parameter, $P_{50}$ .

Step 1 — Write the equilibrium. Myoglobin (Mb) and oxygen combine reversibly at a single site:

\text{Mb} + \text{O}_2 \rightleftharpoons \text{MbO}_2, \qquad K_a = \frac{[\text{MbO}_2]}{[\text{Mb}][\text{O}_2]}

Because this is a genuine equilibrium (no covalent bond broken, only a dative Fe–O₂ bond formed and broken), $K_a$ is a true thermodynamic constant at fixed temperature.

Step 2 — Define fractional saturation. Let $Y$ be the fraction of total myoglobin that carries bound O₂:

Y = \frac{[\text{MbO}_2]}{[\text{Mb}] + [\text{MbO}_2]}

This is the experimentally accessible quantity (measured, e.g., by absorbance at 580 nm, where oxy- and deoxy-Mb have distinct spectra).

Step 3 — Substitute the equilibrium expression. From Step 1, $[\text{MbO}_2] = K_a[\text{Mb}][\text{O}_2]$ . Insert into Step 2:

Y = \frac{K_a[\text{Mb}][\text{O}_2]}{[\text{Mb}] + K_a[\text{Mb}][\text{O}_2]}

The factor $[\text{Mb}]$ cancels top and bottom (it appears in every term):

Y = \frac{K_a[\text{O}_2]}{1 + K_a[\text{O}_2]}

Step 4 — Convert to partial pressure and to $P_{50}$ . Dissolved $[\text{O}_2]$ is proportional to $p\text{O}_2$ (Henry's law, constant solubility coefficient absorbed into $K_a$ ), so we may simply relabel the equilibrium constant in pressure units and call its reciprocal $P_{50}$ . Divide numerator and denominator by $K_a$ :

Y = \frac{p\text{O}_2}{\frac{1}{K_a} + p\text{O}_2} = \frac{p\text{O}_2}{P_{50} + p\text{O}_2}, \qquad P_{50} \equiv \frac{1}{K_a}

Step 5 — Verify $P_{50}$ is literally the half-saturation pressure. Set $p\text{O}_2 = P_{50}$ :

Y = \frac{P_{50}}{P_{50}+P_{50}} = \frac{1}{2}

confirming the definition self-consistently. As $p\text{O}_2 \to \infty$ , $Y \to 1$ (full saturation); as $p\text{O}_2 \to 0$ , $Y \to p\text{O}_2/P_{50} \to 0$ linearly — the low-pressure initial slope of any hyperbolic binding curve is $1/P_{50}$ , with no sigmoidal lag, which is the operational signature of non-cooperative, single-site binding.

Go deeper · Derivation B · 2 minWhy the distal histidine's steric gate lowers relative CO affinity

Derivation BWhy the distal histidine's steric gate lowers relative CO affinity

Goal: derive, in energetic terms, how a fixed steric/hydrogen-bonding perturbation at the sixth coordination site can shift the relative affinity of CO versus O₂ by orders of magnitude, without needing to change the intrinsic chemistry of the iron itself.

Step 1 — Define the discrimination ratio. For any heme site, define the selectivity for CO over O₂ as the ratio of association constants:

M \equiv \frac{K_a(\text{CO})}{K_a(\text{O}_2)}

For free heme in solution (no protein), $M_{\text{free}} \approx 2 \times 10^4$ – $2.5\times10^4$ : naked iron porphyrins bind CO about 20,000-fold more tightly than O₂, because Fe(II) is a good π-back-donor and CO is a superb π-acceptor forming a strong, linear Fe–C≡O unit.

Step 2 — Relate $M$ to a binding free-energy difference. Since $K_a = e^{-\Delta G^\circ/RT}$ for each ligand,

\ln M = \ln K_a(\text{CO}) - \ln K_a(\text{O}_2) = -\frac{1}{RT}\Big(\Delta G^\circ_{\text{CO}} - \Delta G^\circ_{\text{O}_2}\Big)

so $M$ is set entirely by the difference in binding free energies of the two ligands, not by their absolute values.

Step 3 — Identify the perturbation the distal histidine adds. Free CO binds heme end-on and linear (Fe–C–O ≈ 180°), the ideal geometry for maximal π-back-bonding overlap with the filled Fe $d_{xz,yz}$ orbitals. O₂ binds bent (Fe–O–O ≈ 120°) regardless of environment, because its π* system dictates a bent geometry. The distal His64 sits directly above the sixth site and can hydrogen-bond only to the bent, electron-rich terminal oxygen of a bound O₂ (or bound O of Fe–O₂⁻-like character); it sterically clashes with the linear Fe–C≡O geometry, forcing CO to bend or tilt away from its optimal linear approach.

Step 4 — Express this as unequal free-energy perturbations. Let the protein pocket add a stabilization $\delta_{\text{O}_2} > 0$ (favorable, from the H-bond) to O₂ binding and a destabilization $\delta_{\text{CO}} > 0$ (unfavorable, from steric clash forcing a bent, weaker geometry) to CO binding, relative to free heme:

\Delta G^\circ_{\text{O}_2,\text{protein}} = \Delta G^\circ_{\text{O}_2,\text{free}} - \delta_{\text{O}_2}, \qquad \Delta G^\circ_{\text{CO,protein}} = \Delta G^\circ_{\text{CO,free}} + \delta_{\text{CO}}

(more negative $\Delta G^\circ$ = tighter binding, so subtracting $\delta_{\text{O}_2}$ makes O₂ binding more favorable, adding $\delta_{\text{CO}}$ makes CO binding less favorable).

Step 5 — Recompute the discrimination ratio in the protein. Substitute into Step 2's log ratio:

\ln M_{\text{protein}} = \ln M_{\text{free}} - \frac{1}{RT}\big(\delta_{\text{CO}} + \delta_{\text{O}_2}\big)

Both perturbations act in the same direction on the ratio: any distal-pocket effect that helps O₂ or hurts CO shrinks $M$ . Using representative values $\delta_{\text{CO}} + \delta_{\text{O}_2} \approx 6$ – $7\ \text{kcal/mol}$ (from comparing native myoglobin, $M \approx 20$ – $200$ , against site-directed mutants lacking His64, where $M$ reverts toward the free-heme value near $10^4$ ) shows that a modest few-kilocalorie steric/H-bonding perturbation, applied as an exponential factor in $K_a$ , is sufficient to change $M$ by two to three orders of magnitude. This is the general lesson: protein pockets need not change the intrinsic chemical reactivity of a cofactor at all — they only need to differentially bias a small number of geometrically sensitive interactions to produce large changes in ligand selectivity.

Worked examples

Example 1 — Foundational: computing Y at a physiological muscle pO₂. Resting skeletal muscle tissue has a local partial pressure of oxygen around $p\text{O}_2 = 20$ torr (venous-side capillary values are often quoted lower, 20–40 torr, depending on tissue and metabolic state). Using sperm-whale myoglobin's $P_{50} = 2.8$ torr at 20 °C:

Y = \frac{p\text{O}_2}{P_{50}+p\text{O}_2} = \frac{20}{2.8 + 20} = \frac{20}{22.8} = 0.877

So myoglobin sitting in resting muscle is already about 88% saturated. Now drop to a value representative of exercising muscle where local pO₂ falls to 5 torr:

Y = \frac{5}{2.8+5} = \frac{5}{7.8} = 0.641

The saturation fell from 0.877 to 0.641 — a drop of about 24 percentage points — illustrating that myoglobin releases a meaningful fraction of its bound oxygen precisely in the pO₂ range where working muscle needs it, thanks to its steep hyperbola sitting at low P₅₀. Check the limiting case: as $p\text{O}_2 \to 0$ (anoxia), $Y \to 0$ linearly with slope $1/P_{50} = 1/2.8 = 0.357\ \text{torr}^{-1}$ , meaning myoglobin dumps essentially all its oxygen store under severe hypoxia, functioning as intended as a short-term O₂ reservoir.

Example 2 — Applied: extracting $K_a$ , $k_{\text{off}}$ , and residence time from kinetic data. Stopped-flow experiments on sperm-whale myoglobin give (representative, literature-consistent values) an association rate constant $k_{\text{on}} = 1.4\times 10^7\ \text{M}^{-1}\text{s}^{-1}$ and a dissociation rate constant $k_{\text{off}} = 12\ \text{s}^{-1}$ at room temperature. First compute the equilibrium association constant:

K_a = \frac{k_{\text{on}}}{k_{\text{off}}} = \frac{1.4\times10^7\ \text{M}^{-1}\text{s}^{-1}}{12\ \text{s}^{-1}} = 1.17\times10^6\ \text{M}^{-1}

Convert to an apparent dissociation constant in concentration units:

K_d = \frac{1}{K_a} = 8.55\times10^{-7}\ \text{M} = 0.855\ \mu\text{M}

To translate this into the pressure-based $P_{50}$ , use the aqueous solubility of O₂ at 20 °C, roughly $\alpha_{\text{O}_2} = 1.9\times10^{-6}\ \text{M/torr}$ (Henry's-law coefficient for O₂ in water/plasma-like medium at that temperature):

P_{50} = \frac{K_d}{\alpha_{\text{O}_2}} = \frac{8.55\times10^{-7}\ \text{M}}{1.9\times10^{-6}\ \text{M/torr}} = 0.45\ \text{torr}

This is somewhat lower than the commonly cited textbook value (2.8 torr) because real $k_{\text{on}}/k_{\text{off}}$ pairs and solubility coefficients vary with exact temperature, pH, and ionic strength across reported studies — the arithmetic itself, however, is the general recipe: kinetic ratio → thermodynamic $K_a$ → pressure-based $P_{50}$ via Henry's law, and it is this chain of unit conversions, not any single "true" number, that a student must be able to reproduce. Finally, compute the mean residence time of bound O₂ before it dissociates:

\tau = \frac{1}{k_{\text{off}}} = \frac{1}{12\ \text{s}^{-1}} = 83\ \text{ms}

Each myoglobin-bound O₂ molecule, on average, stays put for tens of milliseconds — long enough to be relevant on the timescale of a single muscle twitch, but short enough that the O₂ is always kinetically available for release to mitochondrial cytochrome oxidase.

Example 3 — Exam/competition level: two-point Hill-plot analysis to test for cooperativity. A student measures fractional saturation of a purified myoglobin sample at two partial pressures and wants to determine both $P_{50}$ and the Hill coefficient $n$ without assuming $n=1$ in advance. Data: at $p\text{O}_2 = 1.5$ torr, $Y_1 = 0.300$ ; at $p\text{O}_2 = 6.0$ torr, $Y_2 = 0.706$ . The Hill equation rearranges to a linear form:

\log\!\left(\frac{Y}{1-Y}\right) = n\log(p\text{O}_2) - n\log P_{50}

Compute the left side at each point. At point 1:

\log\!\left(\frac{0.300}{0.700}\right) = \log(0.4286) = -0.368

At point 2:

\log\!\left(\frac{0.706}{0.294}\right) = \log(2.401) = 0.3805

The slope between the two points gives $n$ :

n = \frac{0.3805 - (-0.368)}{\log(6.0) - \log(1.5)} = \frac{0.7485}{0.7782 - 0.1761} = \frac{0.7485}{0.6021} = 1.243

This is close enough to $n \approx 1$ (within the scatter expected from only a two-point estimate and rounding in the input data) to confirm non-cooperative, single-site behavior — exactly what the monomeric globin fold predicts, since there is structurally only one heme per chain and no subunit interface through which a conformational signal could propagate. Now solve for $P_{50}$ using point 1 and the fitted $n$ :

\log P_{50} = \log(p\text{O}_2) - \frac{1}{n}\log\!\left(\frac{Y}{1-Y}\right) = 0.1761 - \frac{1}{1.243}(-0.368) = 0.1761 + 0.2961 = 0.4722

P_{50} = 10^{0.4722} = 2.97\ \text{torr}

This recovers a $P_{50}$ close to the textbook sperm-whale value (2.8 torr), and — crucially for the exam-level lesson — demonstrates the general diagnostic procedure: a Hill plot's slope diagnoses cooperativity ( $n=1$ : independent sites; $n>1$ : positive cooperativity, the hallmark to look for later in hemoglobin), while its intercept recovers the half-saturation pressure regardless of whether $n$ turns out to be exactly 1.

Limiting cases, scope & common traps

Regime / question	What happens	Why
$p\text{O}_2 \to 0$	$Y \to p\text{O}_2/P_{50}$ (linear)	Hyperbola's low-concentration limit is always linear in ligand
$p\text{O}_2 \to \infty$	$Y \to 1$	Full saturation; no upper cooperative sigmoid possible with 1 site
$p\text{O}_2 = P_{50}$	$Y = 0.5$ exactly	Definition of $P_{50}$ , independent of units used
Hill coefficient $n$	$n \approx 1.0$ for Mb	Single site, no allosteric coupling possible
Fe oxidation state	Must stay Fe(II)	Fe(III) (metmyoglobin) cannot bind O₂ reversibly; only weak-field ligands like H₂O/OH⁻ or exogenous CN⁻ bind
Distal His64 removed (mutant)	CO/O₂ selectivity $M$ reverts toward free-heme value (~10⁴)	Loses steric gate and H-bond stabilization of bent O₂
Proximal His93 removed (mutant)	O₂ affinity and rate constants change drastically; heme may be lost	Loses the covalent Fe–N(His) anchor and its electronic tuning
Temperature increase	$P_{50}$ increases (O₂ affinity drops)	Binding is exothermic; higher T favors dissociation (Le Chatelier / van't Hoff)

Common trapMyoglobin is not "a little hemoglobin subunit" functionally

Structurally, one hemoglobin subunit is nearly superimposable on myoglobin — same globin fold, same heme pocket geometry, same proximal/distal histidines. But isolated α or β subunits, when studied alone, actually bind O₂ hyperbolically, much like myoglobin. The sigmoidal curve of whole hemoglobin is a quaternary property, arising only when four subunits are assembled and can shift between T and R states. It is a common exam trap to attribute cooperativity to some intrinsic chemical difference between hemoglobin's heme and myoglobin's heme — there is essentially none; the difference is entirely an emergent property of the tertiary-to-quaternary transition, previewed further under Cooperative Binding and Allostery.

The chemist behind it

Max Perutz and John Kendrew shared the 1962 Nobel Prize in Chemistry for solving, respectively, the first three-dimensional structures of hemoglobin and myoglobin by X-ray crystallography — myoglobin, from sperm whale muscle, was in fact the very first protein structure ever solved at atomic resolution, published by Kendrew's group in 1958 (a low-resolution 6 Å model) and refined to 2 Å by 1960. Kendrew chose sperm whale myoglobin deliberately: diving mammals evolved unusually high concentrations of myoglobin in their muscle to store oxygen for prolonged dives, and the resulting protein crystallizes exceptionally well, making it an ideal test case for the then-nascent method of protein crystallography built on Perutz's earlier discovery of the heavy-atom isomorphous replacement technique. Kendrew's structure was, at the time, almost disappointing to contemporaries expecting elegant symmetry: myoglobin's fold appeared irregular, with helices packed at seemingly arbitrary angles and no obvious repeating motif — a stark contrast to the clean helical regularity of DNA solved five years earlier. In retrospect this "irregularity" was the single most important lesson protein crystallography ever delivered: a protein's biological function does not require geometric elegance, only a folding pattern that is precisely reproducible and chemically functional, each side chain positioned by evolution to do a specific job — exactly as later borne out by the pinpoint placement of His93 and His64 around the heme iron.

Origins · 1 minJohn Cowdery Kendrew and how this was discovered

John Cowdery Kendrew

1917–1997 · Cambridge, MRC Laboratory of Molecular Biology

Kendrew trained originally as a chemist at Cambridge, then worked on operational research during the Second World War before returning to structural biology under J. D. Bernal and, later, alongside Max Perutz at the Cavendish Laboratory. He selected sperm whale myoglobin specifically for its high natural abundance and excellent crystal quality, and used mercury-heavy-atom derivatives to solve the crystallographic phase problem — the technical bottleneck that had stalled protein crystallography for two decades. His 1958 low-resolution and 1960 high-resolution myoglobin structures proved, for the first time, that X-ray methods could resolve the atomic architecture of a genuinely globular protein, opening the field that would later solve hemoglobin, lysozyme, and ultimately the ribosome.

History

- 1958 — Kendrew publishes the first 6 Å resolution structure of myoglobin, the first three-dimensional protein structure ever determined.

- 1960 — Kendrew's group refines myoglobin to 2 Å resolution, revealing the heme pocket, the proximal and distal histidines, and the all-α helical fold in atomic detail.

- 1962 — Kendrew and Perutz jointly awarded the Nobel Prize in Chemistry for structures of myoglobin and hemoglobin.

- 1965–1970s — kinetic and spectroscopic studies (Antonini, Gibson, and others) establish the O₂/CO discrimination ratio and its dependence on the distal histidine.

- 1980s–1990s — site-directed mutagenesis of His64 and His93 in recombinant myoglobin directly confirms their individual roles in ligand affinity, selectivity, and autoxidation rate.

Why is myoglobin's heme not covalently attached to the protein backbone?

Heme b (iron-protoporphyrin IX) is held in myoglobin's pocket almost entirely by non-covalent interactions: van der Waals contacts between the porphyrin ring's methyl and vinyl substituents and hydrophobic side chains lining the E/F helix crevice, plus electrostatic/salt-bridge interactions between the two solvent-exposed propionate groups and nearby basic residues. The single bond that resembles a "true" chemical bond is the dative coordinate bond from the proximal histidine's imidazole nitrogen to the iron itself — but even this is reversible under denaturing or acidic conditions, which is why heme can be extracted from myoglobin (to make apomyoglobin) and reconstituted, a standard laboratory manipulation. Some other heme proteins, notably cytochrome c, do form covalent thioether linkages between heme vinyl groups and cysteine residues, but this is not the case for myoglobin or hemoglobin.

Does myoglobin ever show any cooperativity at all, even weakly?

Under normal physiological conditions and in solution studies, myoglobin's Hill coefficient is experimentally indistinguishable from 1.0, consistent with there being only one oxygen-binding site per polypeptide chain and therefore no structural mechanism for inter-site communication. Some historical reports of slightly elevated apparent Hill coefficients in certain myoglobin preparations were later traced to sample heterogeneity (mixtures of met- and deoxy-forms, or partial denaturation) rather than any genuine cooperative mechanism. The clean n = 1 behavior of myoglobin is precisely why it serves as the standard non-cooperative reference against which hemoglobin's sigmoidal, n ≈ 2.8–3 behavior is defined and quantitatively compared.

Why does myoglobin have a lower P50 (higher O2 affinity) than hemoglobin?

Myoglobin's functional role is oxygen storage in tissue, not oxygen transport between lungs and tissue, so it is evolutionarily tuned to bind O₂ very tightly (low P₅₀, around 2.8 torr) so that it stays loaded under the modest pO₂ found in resting muscle and only releases oxygen when local pO₂ drops sharply, as during intense exercise or ischemia. Hemoglobin, by contrast, must both load efficiently in the lungs (high pO₂, ~100 torr) and unload efficiently in peripheral tissue (lower pO₂, ~20–40 torr), which requires a right-shifted, sigmoidal curve with a higher effective P₅₀ (~26 torr) and allosteric machinery to modulate affinity — a functional distinction that maps directly onto the storage-versus-transport division of labor between the two proteins.

What would happen to myoglobin's oxygen affinity if the distal histidine were mutated to a non-hydrogen-bonding residue like valine?

Removing the distal histidine's hydrogen bond to bound O₂ removes a stabilizing interaction specific to the O₂ adduct, so O₂ affinity typically decreases (higher P₅₀) in such mutants, because the H-bond was making a differential contribution to O₂ binding that a nonpolar residue cannot replicate. At the same time, because the steric/H-bonding penalty against linear CO binding is also relieved, CO affinity often increases relative to O₂, shifting the discrimination ratio M back toward the much larger free-heme value — this exact experiment (His64→Val and similar substitutions) was carried out via site-directed mutagenesis in the 1980s–1990s and is one of the clearest structure–function proofs in all of protein biochemistry.

How is fractional saturation Y actually measured experimentally?

Oxy- and deoxymyoglobin have distinct visible-region absorption spectra because O₂ binding changes the iron's spin state and porphyrin electronic structure: oxymyoglobin shows characteristic α/β absorption bands near 542 and 581 nm, while deoxymyoglobin shows a broader single band near 555 nm. By exposing a myoglobin sample to a series of controlled, precisely known partial pressures of O₂ (using a tonometer or gas-mixing apparatus) and measuring absorbance at a diagnostic wavelength at each pressure, one converts absorbance directly into fractional saturation via a simple two-state spectral decomposition, then plots Y against pO₂ to extract P₅₀ and the Hill coefficient by nonlinear or log-log (Hill plot) regression.

Is the P50 of myoglobin a universal constant, or does it vary?

P₅₀ is not universal; it depends on temperature, pH, ionic strength, and species. Sperm whale myoglobin's frequently cited value of about 2.8 torr is measured near 20 °C at near-neutral pH; raising the temperature increases P₅₀ (lower affinity) because O₂ binding is exothermic, consistent with the van't Hoff relationship between equilibrium constants and temperature. Different species' myoglobins — particularly those from diving mammals versus terrestrial mammals — show systematically different P₅₀ values and different surface charge distributions, reflecting evolutionary tuning of both affinity and solubility to each organism's physiological oxygen-storage demands.

Why doesn't myoglobin's iron just get oxidized to Fe(III) every time O2 binds, given that O2 is an oxidant?

This is exactly the risk the protein is built to manage. When O₂ binds Fe(II), the Fe–O₂ adduct is best described as intermediate between Fe(II)–O₂ and Fe(III)–superoxide (Fe³⁺–O₂•⁻) character, meaning some electron density genuinely does shift toward oxygen. Occasionally this partial electron transfer goes to completion and the bound ligand leaves as superoxide, leaving true Fe(III) metmyoglobin behind — this autoxidation happens at a slow but non-negligible basal rate even in healthy tissue. Cells counter this continuously using enzymatic reduction systems (e.g., NADH-cytochrome b5 reductase) that reduce accumulated metmyoglobin back to the ferrous, O₂-competent state, maintaining a functional steady-state pool rather than preventing the side reaction outright.

Myoglobin Structure and Oxygen Binding

Where this lives today

The physical picture

The core ideas

Worked examples

Limiting cases, scope & common traps

The chemist behind it

Before this

Where this leads