Why Is Folding Built to Come Undone? Proteins, Nucleosomes, and a Few Hydrogen Bonds
A folded structure did not survive because it was strong. It survived because it could come undone. Measure the strength of folding and that shows.
Last time we watched a single string fold and acquire a property.
This time, the other side of it. Why is it built to come undone?
About thirty years ago, in a first biochemistry class: enzyme, nucleosome, histone, peptide, nucleotide, octamer, chromatin, chromosome. I was lost before the subject began.
One exam question in the third year has stayed with me. “Describe the difference between transcription and translation.”
Many people stop here. Not because the concept is hard, but because the vocabulary is the wall.
That exam question first
The answer is one line. Does the language change?
| What into what | Writing system | So | |
|---|---|---|---|
| Transcription | DNA → RNA | four letters → four letters | copying within one language |
| Translation | mRNA → protein | 64 codons → 20 amino acids | the writing system changes |
Copying and rendering are different acts. Hence the different names.
Try to memorise the terms and the wall gets higher. Hold on to what turns into what and the wall is low.
There are two kinds of folding
One thing has to be sorted out first. Protein folding and DNA folding are usually discussed together, and they are not the same phenomenon.
| Protein | DNA | |
|---|---|---|
| What decides | the amino acid sequence itself | histones, a separate apparatus |
| External instruction | none | a winding protein is required |
| Accurate word | folding | wrapping · compaction |
| In common | both can be undone |
Mixing them blurs the point. What the two share is not structure but reversibility, and that is what this article is about.
Ten thousand kilometres into a thirty-metre room
Keep the conversion: one cell as a dome 50 metres across. Five million times.
One nucleus holds about 2 metres of DNA. Each base pair is 0.34 nanometres and a human cell has 6.4 billion of them.
Two metres times five million:
about 10,000 kilometres. Seoul to New York.
And it goes into a room 30 metres across. The original draft said “hundreds of kilometres”; the real figure is an order of magnitude higher.
| What | Real | × 5,000,000 |
|---|---|---|
| Cell | 10 μm across | a 50 m dome |
| Nucleus | about 6 μm across | a 30 m room |
| DNA double helix | 2 nm thick | 1 cm |
| Nucleosome | 11 nm across | 5.5 cm · a fist |
| DNA wound on one nucleosome | 147 bp ≈ 50 nm | 25 cm |
| Metaphase chromosome | about 700 nm thick | 3.5 m |
| Total DNA in one nucleus | about 2 m | about 10,000 km |
Ten thousand kilometres of centimetre-thick cord into a room thirty metres wide. Stuff it in and you can never get it out again.
Twenty-five centimetres around each fist
So it is wound.
Two copies each of the histone proteins H2A, H2B, H3 and H4 — eight in all — form an octamer. At this scale, a fist.
DNA wraps that fist for 147 base pairs, about 1.65 turns. Twenty-five centimetres of cord.
That unit is the nucleosome. Fists threaded on a string.
The atomic structure was solved in 1997. The numbers 147 and 1.65 come from there.
Wind it all and nothing can be read
If compaction were the goal, everything would be wound. Wind everything and nothing can be read.
So nuclear DNA divides into two states.
- Euchromatin — loose. The readable zone
- Heterochromatin — dense. The closed zone
If folding were a compression technique this distinction would be unnecessary. That it exists tells you what folding is actually for.
Folding is not concealment. It is selection. A device for choosing what stays open.
On top of that, cohesin and condensin pull DNA out into loops — loop extrusion, observed directly in 2018. Drag two distant sites together and, without altering a single base, you change which gene meets which switch.
How it is folded is how it is read.
The real strength of a fold — a few hydrogen bonds
How much more stable is a folded protein than an unfolded one?
For small globular proteins the difference is typically 5–15 kcal/mol.
A hydrogen bond between water molecules is about 5 kcal/mol. Which means a protein
holds its shape by the value of a few hydrogen bonds.
A structure of thousands of atoms stands with about as much margin as two water molecules holding on to each other. The term for it is marginal stability.
That is why enzymes fail at a fever of 40 °C. It is not that sturdier was impossible. Sturdy cannot be regulated.
Life did not choose this
One phrase needs fixing here.
“Life chose a state it could revise over one that was safe” assumes a purpose. It requires someone doing the choosing.
The truth runs the other way.
Lineages with structures that could not be undone ended together the first time conditions turned over. What is left is what could be undone.
Life did not choose. The ones that could not choose disappeared.
The conclusion lands in the same place, but written this way it is a claim that can be checked.
Does it always come undone? The exception in the eye
Crystallin, the protein filling the lens of the eye, is effectively never replaced. Radiocarbon dating in 2008 found that the carbon in these proteins dates from around the person’s birth.
An eighty-year-old carries eighty-year-old protein, still folded, inside the eye.
And when it does unfold and clump, that is cataract.
Amyloid fibrils are the same. Once a chain enters that structure it does not come back. The prion of the previous instalment belongs here.
Reversibility is not a law but a usual case. When it fails, the exception generally acquires the name of a disease.
The textbook’s 30-nanometre fibre may not exist
You have probably seen the diagram: nucleosomes coiling into a 30-nanometre fibre.
That structure does form from chromatin purified in a test tube. It is not confirmed inside living cells.
ChromEMT imaging in 2017 found chromatin in cells to be an irregular chain of 5–24 nanometres. No regular 30-nanometre fibre appeared.
The whole hierarchical model — a fixed staircase of levels — is now unsteady, and the textbook picture may be an artefact of the laboratory.
This one is not settled.
Unfolding
Folding is not completion. It is deferral.
Function arises only in the folded state, and operation is possible only when the fold is ready to come apart. An active site is neither always closed nor always open.
The margin of a few hydrogen bonds. That is the width that makes all of it possible.
Coming undone is not a failure. It is the condition.
In short
- Transcription versus translation is whether the language changes: DNA→RNA is copying, RNA→protein is translation
- Protein folding and DNA winding are different phenomena. They share reversibility and nothing else
- One nucleus holds about 2 m of DNA — 10,000 km at 5 million times — inside a 30 m room
- A nucleosome is 147 base pairs wrapped 1.65 turns around a histone octamer
- Euchromatin and heterochromatin show that folding is selection, not compaction
- Folded stability is 5–15 kcal/mol: the value of a few hydrogen bonds
- Life did not choose reversibility; the lineages without it disappeared
- Lens crystallin is never replaced. When it unfolds, that is cataract
- The textbook 30 nm fibre is not found inside living cells
This piece is rebuilt for search readers from “Folding” of the serial essay The Great Corona Swindle. The original’s point of view is kept; the numbers and structures have been reset against primary sources.
Frequently asked
- What is the difference between transcription and translation?
- Whether the language changes. Transcription copies a DNA base sequence into an RNA base sequence — four letters into four letters, within the same nucleic acid language. Translation converts three-base codons of mRNA (64 of them) into 20 amino acids; the writing system itself changes. One is copying, the other is translating.
- Are protein folding and DNA folding the same thing?
- No. A protein's amino acid sequence determines its fold and it folds with no external instruction. DNA does not build its own structure; it is wound around an octamer of histone proteins. The first is spontaneous folding, the second is closer to winding. The only thing they share is that both can be undone.
- How strong is a folded protein?
- Far weaker than one would assume. For small globular proteins the free-energy difference between folded and unfolded states is typically 5–15 kcal/mol. A hydrogen bond between water molecules is about 5 kcal/mol, so the shape is held by the value of a few of them. This is why enzymes fail at high fever.
- Does everything that folds come undone?
- Mostly, but not always. Crystallin proteins in the eye's lens are effectively never replaced; radiocarbon dating showed their carbon dates from around the person's birth. They stay folded for decades, and when they do unfold and clump the result is cataract. Amyloid fibrils are likewise effectively irreversible.
References
- Luger, K., Mäder, A. W., Richmond, R. K., Sargent, D. F., Richmond, T. J. (1997). Crystal structure of the nucleosome core particle at 2.8 Å resolution. Nature, 389, 251–260. — 147 base pairs wrapped about 1.65 turns around the histone octamer, at atomic resolution.
- Kornberg, R. D. (1974). Chromatin structure: a repeating unit of histones and DNA. Science, 184(4139), 868–871. — The proposal that chromatin is built from a repeating unit — the nucleosome model.
- Ou, H. D. et al. (2017). ChromEMT: Visualizing 3D chromatin structure and compaction in interphase and mitotic cells. Science, 357(6349), eaag0025. — Chromatin in cells is a disordered 5–24 nm chain; no 30 nm fibre was observed.
- Rao, S. S. P. et al. (2014). A 3D map of the human genome at kilobase resolution reveals principles of chromatin looping. Cell, 159(7), 1665–1680. — Loops and domains formed by CTCF and cohesin.
- Ganji, M. et al. (2018). Real-time imaging of DNA loop extrusion by condensin. Science, 360(6384), 102–105. — Direct observation of condensin pulling DNA into loops.
- Dill, K. A. & MacCallum, J. L. (2012). The Protein-Folding Problem, 50 Years On. Science, 338(6110), 1042–1046. — The funnel model and the marginal stability of folded proteins.
- Lynnerup, N., Kjeldsen, H., Heegaard, S., Jacobsen, C., Heinemeier, J. (2008). Radiocarbon dating of the human eye lens crystallines reveal proteins without carbon turnover throughout life. PLoS ONE, 3(1), e1529. — Lens crystallins are never replaced — the exception that stays folded for decades.
- Anfinsen, C. B. (1973). Principles that Govern the Folding of Protein Chains. Science, 181(4096), 223–230. — Folding information resides in the sequence. Covered in the previous instalment.
Published September 26, 2026