The Story of Hemoglobin

How a special protein in red blood cells transports oxygen around your body

Jun 23, 2026

It might not seem or feel like it, but by cellular count, we are mostly blood. Over three-quarters of all the cells in our body are the 20-30 trillion red blood cells floating in arteries and veins and squeezing between capillaries to deliver oxygen wherever it’s needed. We make about 2.4 million of them every second, and they circulate tirelessly for three to four months (or 150,000 to 200,000 laps) before the spleen and liver break them down and recycle the parts.

Providing oxygen efficiently is vital to our existence, so maybe it should not come as a surprise that a significant part of us is devoted to transporting it around. But despite how crucial it is, our body doesn’t stockpile any; it has to be delivered continuously through blood. It does so with a special protein housed within red blood cells called hemoglobin, whose sole job is to pick up oxygen where it is plentiful and release it where it’s needed. Without this protein, if we relied on oxygen dissolving in blood alone, we’d need sixty to seventy times more blood circulating to stay alive.

This delivery task is harder than it sounds. When a red blood cell, chock full of hemoglobin, fills up on oxygen in the lungs and is pumped by the heart across the body, nothing tells it where to go. There is no GPS, no central coordinator steering it towards the working muscles that desperately need oxygen and away from the ones that need less. And hemoglobin itself has to keep a delicate balance: if it holds onto oxygen too tightly, or if it lets it go too easily, many tissues won’t get the oxygen they need. Getting every tissue the right amount at every moment — knowing not only how to grab and hold onto oxygen but also when to let go — requires a marvel of precision engineering. What kind of structure could possibly do this?

A brief history

For centuries, scientists have tried to figure out what blood was made of. In the 1740s, Vincenzo Menghini, an Italian physician, decided to study why blood was red. The going theory at the time was that it came from iron, the same material that forms rust, after iron’s presence was found in the ashes of incinerated vegetables several decades earlier. Exactly where this iron came from was still unknown, so Menghini decided to do an experiment: he took five ounces of blood from a dog, roasted it into a dry powder, and then hovered a naturally magnetized knife blade over it. Some bits stuck onto the blade; blood did, in fact, have iron.

He continued experimenting on other animals — an ox, a horse, a boar, birds, frogs, fishes — and found that the iron only came from the red portion of the blood after it was separated; very little came from the serum. And so, for about the next century, iron being the cause of blood’s red color remained undisputed.

But Menghini’s work raised a deeper question: why does blood carry a metal at all? Answering it meant looking inside the blood. Some seventy years before Menghini, in 1675, Antonie van Leeuwenhoek, the Dutch draper and naturalist who made the best microscopes of his time, peered at his own blood and saw tiny “sanguineous globules” floating in it. It took another century to discover the true shape of these globules. William Hewson, a British anatomist, found red blood cells not to be round but disk shaped, with a depression in the middle, like “a circular pillow that has just been punched,” as Siddhartha Mukherjee describes.

In the century that followed, attention moved to what was inside that cell. In 1840, Friedrich Hünefeld, a German biochemist, noticed that dried blood placed between two glass slides grew tiny crystals, rectangular and bright red, with sharp edges. They were the first crystals of a protein ever recorded, but he did not know what it did, and so left it unnamed. Twenty-four years later, Felix Hoppe-Seyler gave it one: hemoglobin, from hemo (blood) and globin (for its round, globe-like shape). He also showed that this protein was where the iron lived, what oxygen attached to, and what gave blood its color.

But knowing what hemoglobin was made of was not the same as knowing how it worked. How did this single molecule pick up oxygen and let it go? To understand that, you needed to see the shape. Beginning in 1937, the Austrian-born chemist Max Perutz spent more than twenty years firing X-rays at hemoglobin crystals, painstakingly reconstructing the molecule from the pattern of spots they scattered. By the end of the 1950s he had the first three-dimensional model of hemoglobin. When he compared its oxygen-loaded and oxygen-empty forms, he saw it physically change shape, clamping shut and springing open as it took on its cargo and let it go.

Thanks to Perutz and the work of hundreds of scientists before us, we can now take apart the molecule and inspect it atom by atom. Its structure reveals its function: how it catches oxygen in the lungs, ferries it across the body, and releases it precisely where it needs to be. The closer we look, the more marvelous and more deliberate this nanomachine turns out to be.

The anatomy of hemoglobin

To understand the function of a molecule, we must first understand its form. The mechanics of how hemoglobin catches oxygen and how it releases it are a product of how it’s assembled. So first, let’s assemble it part by part, and along the way we’ll see how the parts work together to enable its function. Here are the components we have to build with:

Let’s start by assembling the heme, the part that catches and releases oxygen. Each hemoglobin molecule has four of them, and each heme can carry one oxygen molecule; so in all, a single hemoglobin carries four oxygen molecules.

We’ll begin with just the ferrous iron ion, which sits right at the center.

To this, let’s bond a single pyrrole, a five-sided ring with a nitrogen pointing inwards, towards the iron.

Then, surround the iron with four of these pyrroles, each joined to one another with a methine bridge (a single carbon). This creates what’s known as the porphyrin ring, with iron clasped in the center. (This ring — not just the iron — is where blood gets its color; it soaks up the green and blue of visible light and reflects back red).

To the back of this ring, attach the proximal histidine, an amino acid that bonds to the iron and anchors the heme to the rest of the structure. In all, iron can connect on six sides. Four are taken by the porphyrin nitrogens lying in the plane of the ring, and the proximal histidine, reaching up from below, is the fifth. That leaves a single open spot on the opposite side, pointing up out of the ring, where oxygen will later bind.

Hovering just above the ring is the distal histidine, so named because it sits a little further away. It’s also attached to the same chain as its sibling, the proximal histidine; as we’ll later see, both play crucial roles in catching and releasing oxygen.

This whole assembly together — the iron, the porphyrin ring, and the proximal and distal histidines — makes one complete heme pocket (not to be confused with heme, which is just the iron and porphyrin ring). A single hemoglobin molecule cradles four of these heme pockets.

Each heme pocket is wrapped in a protein chain, the structure the histidines are connected to. It can be either an alpha chain or a beta chain. One alpha and one beta chain clasp tightly together to make an alpha-beta dimer, one half of the whole molecule.

A second, identical alpha-beta dimer completes the structure; we now have a complete hemoglobin molecule. The two halves fit snugly together, but with just enough room to shift and rotate against one another. As we’ll later see, this ability to rotate also plays a crucial role in its function.

How hemoglobin catches oxygen

Now that you understand the structure, let’s look at exactly how it catches oxygen. Before oxygen arrives, the ferrous iron atom sits a little behind the plane of the heme, about half an ångström off (a fraction of an atom’s width), doming out of the ring. In this state, the iron atom is too large to fit through the ring hole. As oxygen binds, the electrons in iron rearrange, and the atom grows slightly smaller, just small enough to rise into the plane of the ring. (That same rearrangement also changes how the heme absorbs light, which is why arterial blood, freshly loaded in the lungs, runs a brighter scarlet, while the spent blood returning through your veins is a darker, almost maroon red.)

Because the iron is tethered to the proximal histidine, as iron moves up, it pulls the histidine along with it. And since the histidine is connected to the rest of the chain, this pull tugs the rest of the chain as well.

To understand the significance of this tug, we need to revisit the rotation ability mentioned earlier. Without oxygen, hemoglobin lies in a tense (T) state, its structure held rigid by a network of salt bridges — small electrostatic clasps around the tail ends and joints of the protein chain. In this state, the proximal histidine is more restrained, holding back iron with more rigidity, and so this hemoglobin has a low appetite for oxygen. The other state is the relaxed (R) state, reached by a 15-degree rotation of one half against the other. In this state, the proximal histidine can move more freely, and so this hemoglobin has a high appetite for oxygen. There is no stable position in between; like a light switch or a clicker pen, the molecule rests fully in one setting or the other.

When a pull from the bound oxygen travels out through the chain, it strains the clasps — the salt bridges — that hold the tense state. A single oxygen is usually not enough to break them, but once one or two have bound, the salt bridges give way, the two halves rotate their 15 degrees, and the molecule switches into the relaxed state.

This flip means that the remaining oxygen now binds more readily. Thus a positive feedback emerges: the first one or two to bind have the hardest task, binding to a reluctant tense state and paying the higher energy cost to flip the molecule into the relaxed state. But once it has flipped, the rest bind to the eager relaxed state almost for free. One bound oxygen makes the next three easier to bind. This working together is called cooperativity, and as we’ll see later, it enables some very advantageous behavior.

In the binding animation above, we left out the distal histidine, but it too has a crucial role. Left to themselves, oxygen and ferrous iron would combine into iron oxide, also known as rust. Were that to happen, hemoglobin would be ruined and unable to deliver oxygen. But it doesn’t happen, because oxygen doesn’t fully react with iron. Rather than fully surrendering its spare electron and making a complete bond, iron only partially shares it, and oxygen, for its part, only leans in partway, like two frenemies reluctant to properly shake hands. That partway lean, with oxygen settling at roughly a 120-degree tilt where it meets iron, is steadied from above by the distal histidine. It hovers above and forms a weak bond — an electrostatic tether — holding the oxygen at that angle and keeping the binding gentle and easily reversible. Essentially, rusting, the exact reaction that eats ships and bridges, is almost running in your blood, deliberately frozen a hair short of completion so it can be undone and redone trillions of times to catch and release oxygen.

How hemoglobin releases oxygen

Mechanically, releasing oxygen works mostly like catching in reverse: hemoglobin switches from the relaxed state back to the tense state, which grips oxygen more loosely and lets it slip away to a tissue that needs it.

The flood of oxygen in the lungs switches hemoglobin from the tense to the relaxed state, but what switches it back, from relaxed to tense? There are several factors in play, each characteristically clever. Let’s say you’re doing a hard set of squats. Your quadriceps are burning lots of oxygen and releasing lots of carbon dioxide, which dissolves in the blood and makes it more acidic. The extra protons of that acid bind to the ends of the hemoglobin chain and give them a positive charge, forming new salt bridges. These added clasps tip the protein back towards the tense state, loosening its grip on oxygen and letting it go. This response, where the waste product of burning oxygen is the signal for exactly where the demand for more oxygen is greatest, is called the Bohr effect.

Not all the carbon dioxide turns to acid though. Some of it binds directly to the ends of the hemoglobin chain, like commuters grabbing a departing train. Those ends turn into carbamate groups. They form preferentially in the tense, deoxygenated state, hanging onto hemoglobin that has just unloaded its cargo, and like the acid and salt bridges, stabilizing this T state and reinforcing the Bohr effect. Through this, the carbon dioxide hitches a ride back to the lungs, where the high oxygen pressure pushes hemoglobin back to the R state, breaking the salt bridges and carbamate and letting carbon dioxide fall off, eventually into air that is exhaled. Thus, oxygen released in tissues lets carbon dioxide attach; oxygen attaching in the lungs pushes out carbon dioxide. This process is known as the Haldane effect, the structural mirror of the Bohr effect.

There is yet another signal that tips hemoglobin from relaxed to tense state. In the tense state, a small cavity opens between the two chain halves; in the relaxed state, the rotation closes the gap. Red blood cells produce a tiny molecule called 2,3-bisphosphoglycerate, or 2,3-BPG, that slots into that cavity and props it in the tense state, like a doorstop wedged under a door. All three signals — carbon dioxide dissolving to acid, carbamate groups, and 2,3-BPG — work together to push hemoglobin to release oxygen.

Cooperativity

So far, we have seen the mechanics of a single hemoglobin molecule. Binding and releasing of oxygen are two sides of the same loop: one heme binding oxygen makes the others bind more easily; one heme letting go of oxygen makes the others let go more easily. This is cooperativity, and it allows oxygen to be gathered quickly where it is plentiful, as in the lungs, and surrendered quickly where it is scarce, as in a working muscle.

This relationship between the availability of oxygen (measured by its partial pressure) and the amount of oxygen hemoglobin holds (measured as percent saturated) can be graphed across all the hemoglobin in a body.

Cooperativity is what gives the curve its sigmoidal, S-shaped form. At the low end, on the far left, increasing oxygen pressure leads to a very slow increase in saturation. Here, most hemoglobin molecules are in the tense state. But, after a certain threshold, there is a steep increase; small increases in pressure lead to larger increases in saturation as the molecules switch to the R state and the positive loop kicks in. At high pressure, saturation levels off, given that there are fewer and fewer open spots left for oxygen.

Nearly each point on this curve is a potential situation somewhere in the body. At sea level, the pressure in your lungs is ~95mmHg, which loads your hemoglobin to almost full, about 98%. If you climb a small mountain, say 1800m above sea level, this pressure drops to ~75mmHg, but saturation remains steady at ~95%. Both of these are loading points, high on the flat shoulder of the curve. Unloading happens further down. Resting tissue has a pressure of 40mmHg and a saturation of ~75%; since the blood arrived nearly full, a muscle at rest has consumed only a quarter of all available oxygen. That point sits right at the top of the steep part of the curve, where a small dip in pressure releases a relatively large amount of oxygen. A working muscle, at a pressure of 20mmHg, has a saturation of ~32%, meaning it now consumes two-thirds of the available oxygen.

In both cases the pressure changes by the same amount — 20mmHg — but the saturation does not. Up on the flat shoulder, that 20mmHg barely moves saturation — just a 3% drop — which is why you can climb to elevation and still breathe relatively well. Down on the steep part, the same 20mmHg that separates a resting tissue from a working one does exactly what you’d want: it unloads far more oxygen — ~43% here, nearly 15 times the change up high — to feed the muscle.

The switch between states enables this. Were hemoglobin stuck in the relaxed (R) state, it would grab oxygen greedily in the lungs but refuse to release it in the tissues; and were it stuck in the tense (T) state, it would release oxygen easily but never load up properly in the first place. The switch allows it to change its appetite depending on where it is.

The effectors that push hemoglobin to release oxygen — the Bohr effect (carbon dioxide dissolving to acid), carbamate groups (carbon dioxide attaching directly), and 2,3-BPG — all act on the curve too. Each of these shifts the curve to the right. To compare curves, we can use the P₅₀, the pressure at which hemoglobin is 50% saturated. A right-shifted curve has a higher P₅₀, which means that it takes less of a pressure drop to get rid of oxygen. In other words, hemoglobin lets go of its oxygen more easily than normal — it has a lower affinity for oxygen — which is exactly what those signals indicate and cause: a higher demand for oxygen.

Among these, 2,3-BPG, the small molecule manufactured inside red blood cells, acts as a regulator, keeping hemoglobin’s affinity at the right level: a P₅₀ of ~26mmHg. That level is a balance struck from both sides. Without 2,3-BPG, hemoglobin’s affinity climbs too high: the curve shifts left, loading fully in the lungs but holding on too tightly to release much in the tissues. Pushed the other way, it would let go easily but never fill up in the lungs to begin with. 2,3-BPG holds it at the point that does both well, and the body can re-tune that point: at high altitudes, for example, it makes more, nudging the curve right to improve unloading.

This same tiny molecule plays a second role in fetal hemoglobin, which has a slightly different structure, swapping its beta chains for gamma chains. 2,3-BPG doesn’t bind as well with fetal hemoglobin, and so its affinity curve stays left-shifted relative to the mother’s curve. The higher oxygen affinity of this left-shifted curve allows it to literally pull the oxygen from the mother’s blood across the placenta, which is how a growing fetus breathes.

In its last days maturing in the bone marrow, a red blood cell does something drastic: it ejects its nucleus, its mitochondria, and every organelle it owns. It does this to make extra room for hemoglobin, packing it in until it fills a third of the cell's volume, just short of the concentration at which it would crystallize and ruin the cell. About 270 million copies ride in each one. Multiplied across every red cell, that's roughly 6.5 sextillion (that’s 6,500,000,000,000,000,000,000) molecules in circulation at once, some 700 grams, or close to one percent of you.

We might feel unique, having this special protein transport our oxygen, but this design — a heme cradled inside a protein chain — is one of the oldest structures in all of biology. It's older than blood, older than animals, older than bacteria; in fact, an ancient precursor was already in use when life was nothing but single cells in an ocean, where it most likely did chemical-defense or sensing work rather than ferrying oxygen.

Now, the modern heme-and-protein-chain design is used across life. Nearly every animal carries its version of hemoglobin. A trout's, a sparrow's, and ours might differ in DNA sequence, but the shape — two pairs of folded chains, clasping its iron — is the same. Outside blood, other similarly designed protein chains perform different jobs. There's myoglobin, in the muscles of diving whales, hoarding oxygen for their lengthy descent; there's leghemoglobin, in the root nodules of soybeans, doing the opposite job, clearing oxygen away to protect bacteria that feed the plant; and there's a similar protein, in bacteria themselves, sensing oxygen rather than carrying it.

We began with the question of what kind of structure could possibly get the right amount of oxygen to every tissue, knowing both how to hold on and when to let go. It is an ancient protein, a marvel of precision engineering, holding its iron firmly enough to gather oxygen and weakly enough to surrender it, keeping that bond a breath short of rust as it flips between two states to change its appetite. Across our body, every second, this catch-and-release plays out more than a hundred quintillion times, all snapping without instruction, without rest, and without you noticing. And it's just one of the many nanomachines keeping us alive.

Further Reading

• Life’s Blood: The Story of Hemoglobin — Michael H. Rosove (2024)

The 3D models and animations were created with Mol*, an open-source molecular visualization toolkit.

If you liked this and would like to hear when new content is published, please subscribe below.

If you have any feedback, found bugs, or just want to reach out, feel free to DM me on Twitter or send me an email.