H&E Staining: Hematoxylin and Eosin Explained

Haematoxylin and eosin is the default stain of histopathology: the section a pathologist looks at first, and usually the only one they need. Its reputation for being simple is misleading. H&E is a two-dye method in which one of the dyes is not actually a dye until you oxidise it, only works when chelated to a metal ion, and is routinely over-applied on purpose and then stripped back. Most bad H&E comes from misunderstanding those three facts rather than from poor technique at the bench.

What the two dyes actually stain

The pairing works because the two dyes carry opposite charges and therefore target opposite classes of structure.

Haematoxylin, once complexed with a metal mordant, behaves as a positively charged (cationic) dye. It binds acidic, negatively charged material — chiefly nuclear chromatin, and also ribosomes and rough endoplasmic reticulum. Structures that take it up are described as basophilic, and appear blue-purple.

Eosin is a negatively charged (anionic) dye that binds positively charged proteins in cytoplasm and connective tissue, producing the familiar pink to orange counterstain. Structures that take it up are eosinophilic.

That is the whole logic of the image: nuclei blue, most everything else pink, with the intensity of each reflecting how much charged material is present. A densely basophilic cytoplasm, for instance, usually means abundant ribosomes — a plasma cell or an actively synthesising tumour cell — not a staining fault.

Haematoxylin is not the dye — haematein is

This is the single most useful thing to understand about the method. Haematoxylin is a natural product extracted from logwood, and in its native form it stains poorly. It must first be oxidised to haematein, the actual colouring compound. Oxidation happens either slowly on exposure to air and light (natural ripening) or immediately using a chemical oxidant such as sodium iodate.

Oxidation is not a single reaction with a single product. Mono-, di-, tri-, tetra- and penta-haematein all form, and the aluminium complexes of each are a different colour: di- and tri-haematein are the purple-to-blue species you want, tetra-haematein is brownish, and penta-haematein is effectively colourless. An over-oxidised, over-age haematoxylin therefore does not simply get weaker — it drifts brown and muddy, because the population of oxidation products has shifted past the useful ones. That is why haematoxylin solutions have a working life rather than merely a concentration.

Why it needs a mordant

Haematein alone still binds tissue poorly. It becomes a practical nuclear stain only when chelated to a metal ion — a mordant — which supplies the positive charge that lets the complex bind anionic chromatin. Aluminium and iron salts are the successful mordants; aluminium is standard for routine work, giving the classic blue-purple nucleus. The resulting dye-metal complex is called a lake, and an alum haematoxylin is often called haemalum.

The practical consequence: anything that strips the metal ion strips the stain. Acid does exactly that, which is the basis of the differentiation step below, and also why acidic fixatives, acidic tap water or residual decalcifying agent will quietly weaken nuclear staining no matter how fresh the haematoxylin is.

Progressive versus regressive haematoxylin

Formulations differ in dye concentration, and that determines how you use them.

• Progressive (for example Mayer’s) is more dilute and relatively selective for nuclear chromatin. You leave the section in until the nuclei reach the intensity you want, and stop. No differentiation step is needed.
• Regressive (for example Harris) is more concentrated and stains nuclei and cytoplasm intensely and indiscriminately. You deliberately overstain, then remove the excess in acid alcohol until only the chromatin retains dye.

Neither is better. Progressive is more forgiving and suits automated runs; regressive gives crisper chromatin detail in skilled hands. What causes trouble is running a regressive protocol without a proper differentiation step, or a progressive one with an aggressive acid rinse — the two formulations are not interchangeable between protocols.

The sequence, and what each step is for

A routine paraffin-section protocol runs in this order. Times below are typical starting points from published routine protocols; every laboratory tunes them to its own fixation, section thickness and reagent age.

1. Deparaffinise — three changes of xylene, about three minutes each. Paraffin is hydrophobic; leave any behind and the aqueous haematoxylin simply cannot reach the tissue, producing pale or patchy areas.
2. Rehydrate through descending alcohols — 100% then 95%, several changes, roughly ten dips each — then rinse in water. You are walking the section from an organic solvent to an aqueous one without shocking it.
3. Haematoxylin — the nuclear stain.
4. Rinse in water to stop the reaction.
5. Differentiate in acid alcohol — literally one to two dips — then wash in tap water for about a minute. Regressive protocols only.
6. Blue in an alkaline solution (ammonia water, roughly 30-60 seconds, or Scott’s tap water substitute).
7. Rinse, then move into alcohol before eosin.
8. Eosin — alcoholic eosin, around two minutes.
9. Dehydrate through ascending alcohols, clear in xylene, and mount in a resinous medium.

Differentiation and bluing decide the quality

These two steps cause most of the variation between a crisp slide and an unreadable one, and they are the two most often rushed.

Differentiation is a weak acid that removes dye from everything that holds it loosely, leaving it only where binding is strongest — the chromatin. It is measured in dips, not minutes. A few seconds too long and nuclei go pale and grey; skip it after a regressive stain and the whole section stays a flat blue that buries cytoplasmic detail and makes the eosin look weak by comparison.

Bluing is not cosmetic. Fresh from acid, the haematein-aluminium complex is a dull reddish-brown. Raising the pH converts it to the stable blue-purple form. Under-blued sections look brown-nucleated and are frequently misread as over-differentiated and re-stained, when the actual fault is an exhausted or too-brief alkaline step. Running tap water blues sections adequately in many laboratories, but only if the supply is genuinely alkaline — soft or acidic water will not do it, which is why Scott’s tap water substitute exists.

Reading a section, and what the colours do not tell you

H&E reports charge and density, not identity. It will show you nuclear size, shape, chromatin texture, nucleoli, mitotic figures, the architecture of an epithelium and the character of the stroma. It will not tell you which protein a cell expresses, and no amount of staining quality changes that. When the question is identity or lineage, the answer is immunohistochemistry or immunofluorescence, with H&E as the orienting section alongside them.

Common failures and their actual causes

• Pale or absent nuclear staining — incomplete deparaffinisation, exhausted or over-oxidised haematoxylin, or acid carried over from fixation or decalcification stripping the aluminium complex.
• Brown rather than blue nuclei — under-blued, or haematoxylin that has oxidised past the useful di-/tri-haematein species.
• Flat, uniformly blue section — a regressive haematoxylin run without adequate differentiation.
• Weak-looking eosin — often not an eosin problem at all, but excess residual haematoxylin flattening the contrast.
• Patchy staining across one slide — sections not fully covered in a bath, floating debris, or wax remnants; check the deparaffinisation series before touching dye times.
• Section detachment — inadequate adhesion or over-vigorous washing, made worse by long alkaline bluing on unsubbed slides.

Sectioning artefacts — chatter, folds, knife lines — are upstream of staining and are not fixed by adjusting dye times; see microtome technique. Resolution limits on what a stained section can show are covered in optical microscopy resolution.

Frequently asked questions

Why is it called H&E rather than haematein and eosin?

Convention, and the fact that haematoxylin is what you buy and put in the bath. The molecule doing the staining is haematein, formed by oxidising haematoxylin, but the reagent and the method kept the older name.

What makes something basophilic or eosinophilic?

Charge. Basophilic structures are acidic and anionic — chromatin above all — and attract the cationic haematein-metal complex. Eosinophilic structures are basic and cationic, mostly cytoplasmic and extracellular proteins, and attract anionic eosin.

Do I need to differentiate?

Only with a regressive haematoxylin such as Harris, where overstaining is deliberate. A progressive formulation such as Mayer’s is stopped when the nuclei look right, and an acid rinse will simply strip it.

Why did my nuclei come out brown?

Most often inadequate bluing — the acid form of the complex is reddish-brown and needs an alkaline step to convert. If bluing is adequate and the problem persists, suspect an over-oxidised haematoxylin producing tetra-haematein.

How long does haematoxylin last?

It has a working life, not just a concentration. Oxidation continues in the bottle, shifting the mixture toward brownish and colourless products, so an old solution stains differently rather than merely faintly. Track it by performance on a control section, not by date alone.

Can H&E be done without xylene?

Substitutes exist and have been evaluated in the literature, including comparisons of non-xylene deparaffinisation against the conventional series. They are adopted mainly to reduce solvent exposure; validate any substitute against your own controls before switching, since deparaffinisation quality drives everything downstream.

References

• Nuclear staining with alum haematoxylin — PubMed 19579146
• Does progressive nuclear staining with hemalum involve DNA, and what is the nature of the dye-chromatin complex? — PubMed 29320873
• H&E staining overview and best practice — Leica Biosystems Knowledge Pathway
• Liquid dish-washing soap as an alternative to xylene and alcohol in deparaffinisation and H&E staining — PMC4196369