Modern technology availability is dangerous
What used to cost years of study now costs an afternoon. The example that stuck with me carries no DNA and no RNA, survives standard sterilisation, and can take decades to kill.
Our access to knowledge and technological power has never been greater, and neither have the risks that come with it. Prions are the example I keep coming back to, less for what they do to a person than for how badly everything we use to notice things works on them.
A pathogen with no genome#
Prions are misfolded proteins. Not bacteria, not viruses, not alive in any sense anyone bothers to argue about. They carry no DNA and no RNA, so there are no instructions in them at all, and they copy themselves regardless.
- genetic materialnone
- copies byrefolds the protein it touches
- made ofone misfolded protein chain, and nothing else
Why now#
Understanding a field used to take years and a supervisor. Now it takes an afternoon. I have never built a quadcopter. If I started one tomorrow a model would walk me through the design, tell me which parts to order, and sit with me at two in the morning while I worked out why the thing refuses to arm. I think that is wonderful. It works exactly as well on whatever else somebody decides to read about.
The equipment went the same way. You used to have to build a lab. Now you order one online.
And the check that normally catches this does not apply. Order synthetic DNA and the supplier matches your sequence against a list of ones nobody should have. That check needs a sequence. A prion does not have one.
None of which means somebody can do this tomorrow. Deliberate prion production is not a demonstrated capability and I am not going to pretend to know how far off it is. What has changed is the lock on it. That used to be hardware costing millions and the few people trained to run it. Both got cheap.
The copy has no instructions in it#
Your neurons already make the protein. PrPC, folded correctly, sitting in the membrane doing something nobody has fully pinned down. A prion is that same protein (same gene, same amino acid sequence, letter for letter) folded wrong. When a wrong fold touches a right one, the right one takes on the wrong shape, and that contact is the whole replication mechanism.
No genome means no mutation, no error correction, and no need for either. It also means there is nothing to sequence.
Detecting a shape, not a sequence#
PCR amplifies DNA. RT-PCR amplifies RNA. Antibody tests bind foreign protein. A prion has no DNA, no RNA, and its protein is not foreign, because it is yours. The lab is not failing so much as asking three questions the sample has no answer to.
- PCRamplifies DNAblind — no DNA anywhere in it
- RT-PCRamplifies RNAblind — no RNA either
- antigen / ELISAantibody binds a foreign proteinblind — same sequence as your own protein — only folded differently
- RT-QuICfeeds it normal protein and watches for a cascadedetects — the mechanism is the assay — routine since ~2015
Then there is the fourth row. RT-QuIC feeds the sample normal protein and watches for a cascade, using the mechanism itself as the assay. It has been routine since around 2015 and it works: roughly 92% sensitivity on spinal fluid, specificity close to perfect.
So prions are not undetectable. They are detectable by a test somebody has to already suspect they need, and that turns out to be the slower problem.
It survives standard sterilisation#
Standard sterilisation is pressurised steam at 121 °C for 15 minutes, and it is what a hospital trusts to make an instrument safe. It takes bacteria and viruses down by six orders of magnitude or more, and it barely inconveniences a prion. The published protocol for suspected prion contamination is 134 °C for 18 minutes in 1 N sodium hydroxide. Plenty of places skip that and incinerate the instrument, because incineration is cheaper than being wrong.
Sunlight does not help either, for the same reason PCR does not: UV sterilises by breaking nucleic acid, and there is none to break. Freezing does nothing at all. Scrapie has come back on pasture 16 years after the last infected sheep left it.
The gap is the whole problem#
Incubation runs years to decades. Kuru went past fifty in documented cases, vCJD from British beef had a median around twelve, and CJD from contaminated growth hormone ran anywhere between five years and thirty-plus. An incubation measured in weeks could be contained. One measured in decades cannot.
The first person shows symptoms somewhere around year two and a half, the tenth around year five. Nobody sees a curve. They see one confused patient in their fifties, then another one in a different city a year and a half later, with nothing linking them but an event that was never written down. By the time the shape is visible the exposure is long over, and there is nothing left to trace it back to. The delay is arithmetic, not incompetence.
The asymmetry#
Understanding got fast. Detection did not, and mostly cannot: it still needs someone to be suspicious, a sample to exist, and a particular assay to get ordered on it. Everything above is one long example of that gap, and prions are the most extreme version of it I could find.
Nobody is ready#
This is the part that frightens me.
Governments are not ready for it. Surveillance counts cases, and for the first decade there would be no cases to count. Doctors are not ready for it. The test exists, but somebody has to suspect a prion strongly enough to order it, on a patient whose symptoms look like a dozen commoner things first. And none of us are ready for what comes after, because there is no cure, no vaccine, and nothing that changes the ending.
The same cheap reading that lets me build things I am not qualified to build is the reason I know any of this, and I have not worked out what follows from that. What is protecting us today is that nobody capable has thought it worth doing. I would like to call that a safeguard. It is luck, and I do not know how long luck holds.