Slashdot Mirror
"Normal" Prions May Protect Myelin
thomst writes "Nature Neuroscience just published an online article about the function of 'normal' prions in protecting myelin, the substance that sheathes and protects sensory and motor nerves. The international study (which has 11 authors) concluded that 'normal' (i.e., not mis-folded) prions may form a protective coat around myelin. The researchers found that Prnp -/- mice (mice with the gene for prions knocked out) consistently developed progressive demyelination, inevitably leading to persistent polyneuropathy by 60 weeks of age. Their data suggest that damage to myelin sheaths cause normal prions to cleave, and the resulting prion fragments activate Schwann cells, which are known to play a part in myelin repair. This research might eventually lead to possible treatments for progressive polyneuropathies in humans, including those mediated by Creutzfeldt-Jakob disease, multiple sclerosis, Alzheimer's, and even diabetes."
First, some background. Most people won't know what a prion is, so I'll explain with a bit of a computer analogy.
Most proteins are like binaries, being executed by the universe. If you put another molecule next to it (usually called a 'substrate') then it will do things to that molecule, by changing its shape and moving its charge around. Enzymes are proteins that return to their original shape afterward. There are also some relatively inert proteins that don't change shape or do anything; they're just for structural purposes.
The mechanics of the cell are very flaky, however. In a computer, we can be sure that when we copy a program from disk (DNA) into memory (polypeptides, an actual molecule) and run it, we'll get an exact copy. In molecular biology, though, all of these processes are imperfect: sometimes we copy the wrong data into the data bus (transcription errors), sometimes we write the wrong thing into RAM (translation errors), and sometimes, since these are 3D structures that need to fold into a proper shape to work, we actually rearrange the bytes that get loaded into memory (there are lots of bits that say 'insert tab A into slot B', but they work off electrical charges). This is called misfolding, reasonably enough. Most of the time the cell can recognise a malformed protein and marks it for deletion with a molecule called ubiquitin. (It then gets sent to the bit bucket.) To make matters worse, proteins can get old and misfold on their own (this usually calls for another round of ubiquitin if the protein doesn't break down totally)
A prion is a very specific class of misfolded protein, which appears sufficiently normal to the cell that it can't decompose it, lives in the brain where the body's immune system can't obliterate it, and, most importantly, if it collides with other proteins of what it was supposed to look like, it will turn them into prions as well, somewhat like vampires, zombies, or your classic EXE-modifying computer virus. The effect is that the prion spreads exponentially, screwing up the machinery of the cell.
Now, a protein has to be really complex for this to be possible. Some proteins are really simple, like the humble microtubule, which just provides a conduit, and some are incredibly complex, like DNA polymerase, which reads the nucleotides on DNA and makes a duplicate. These proteins are usually highly conserved (that is, they look very similar in many species, because if they break, the organism dies, and evolution hits a dead end), and very, very important. As a result, when a prion forms, it comes at a great cost to the overall health of the organism. Worse, it's transmissible (though usually only by cannibalism, which is kind of funny in a scary sort of way.)
So, after all that, what am I complaining about? Well, the headline makes it sound like we've discovered a case of stable self-modifying code, but we haven't. The article just talks about a protein, PrP^C, which is known to cause a prion problem when broken. It's named "axonal prion protein" because, until this study was conducted, that's all we knew about it: if it broke, it was bad. Similarly, there are a bunch of genes called "oncogenes" because they cause cancer if they break, but they're actually really important; removing them generally prevents cell division completely. There is no such thing as a "normal" prion, at least not one introduced by this article. It just turns into a prion if it breaks.
But hey, I'm only an undergrad; what do I know?
Prions are dangerous mainly because proteins are more stable than nucleic acids, so sterilization techniques that are adequate against viruses and bacteria aren't effective against prion-based diseases like BSE and CJD.
Prions are proteins that have mis-folded. They stop folding on the same microsecond timescale as normal proteins, but most develop "amyloid folds" that (as you say) causes them to build up like trash in the body.
Prions certainly stop growing. Also, a big problem is that they form structures that are very stable. You seem to be describing cancer, which is effectively immortal, lacks the usual constraints on mitosis frequency, and breaks apart (metastasizes) to spread throughout the body.
A prion's actual method of infection is that the mis-folded protein induces other correctly-folded proteins in its vicinity to change to the mis-folded state.