On proteins as repeatable marks on a wall, the potassium channel as a piece of graffiti too load-bearing to overwrite, and the point at which a misfolded protein carries the metaphor past anywhere the wall can follow.
Proteins are the working molecules of the cell: long chains of amino acids, assembled one link at a time by the ribosome according to instructions copied from DNA, then set loose to catalyse reactions, build structure, ferry substances across membranes, and carry signals from one part of the cell to another. What makes them worth reading as graffiti, a theme I have previously explored, is that a protein is never made once. The same genetic template is read over and over, by different ribosomes at different times, producing many separate instances of essentially the same mark, each one an individual object shaped by its own local conditions yet recognisably the same design as every other copy on the wall. That repeatability without exact uniformity is closer to how a tag actually behaves across a city than to any single static picture. It's that dynamic, repeated, slightly variable production, which is not just a resemblance of shape, that makes the analogy worth pursuing rather than merely decorative.
A protein begins as a single, unbroken line. It is one continuous chain of amino acids, added end to end by the ribosome in one uninterrupted pass. The line folds into a specific three-dimensional shape dictated entirely by the sequence of residues along it, much as a spray-can line, laid down in one gestural sweep, loops and curls according to the rhythm of the hand that drew it. Pushed further, though, it risks implying two clean, separate steps, draw the line, then let it fold. Biology doesn't really keep the two processes apart as folding usually begins while the chain is still being made, not once it's finished. Indeed, many proteins only reach their correct shape with help from dedicated chaperone proteins, closer to a second hand steadying the first than a line curling unaided. The image is best kept to what it demonstrates cleanly, that sequence determines shape. We shouldn't stretch the metaphor into a literal account of how the shape is reached.
Some images survive by their own excellence, whereas highly conserved, long-lasting proteins survive by structural entanglement with what is alongside them. Here we might get a sense of a slight weakening of the graffiti metaphor, as the wall lives on when the image is erased, whereas a cell would die if such a protein were changed. The potassium channel's selectivity filter is as clean an example as biology offers of exactly this kind of entanglement, a design held in place not by its excellence but by everything around it that depends on it staying fixed. Carbonyl oxygens spaced with atomic precision mimic the shell of water that normally surrounds a potassium ion, letting K⁺ through while excluding the smaller sodium ion, which isn't quite the right size to be coordinated the same way and so can't afford to shed its own hydrating water to pass. Move one of those oxygens by less than the width of an atom and the whole mechanism stops discriminating at all. This is exactly why this sequence has gone almost unchanged from bacteria to humans, not the best design in any aesthetic sense, but a piece of graffiti so structurally load-bearing that reworking it would collapse the wall around it. Against that stand the proteins that do get overwritten, for reasons that carry no such judgement: a tag can be repainted because the context of the wall has changed, as when a fast-evolving immune receptor is redrawn to keep pace with a shifting pathogen; because the same hand duplicated its own mark and left one copy free to drift while the other stayed fixed; or because the wall's own survival depends on refusing to let any one design become permanent, as with the deliberate reshuffling of antibody genes.
Every mark discussed so far, be it the channel's filter, the fast-evolving receptor, the duplicated and drifting copy, depends on an agent standing apart from the mark itself. This would be a ribosome reading a template, a genome being copied or a hand holding the can. A prion needs none of this. A misfolded protein, touching a correctly folded copy of itself, simply forces it into the same shape, and that shape then spreads to the next copy, and the next, with no template, no tool, and no artist anywhere in the chain. This is propagation by contact alone. Read against the K⁺ channel, prized precisely because it took so much scaffolding to build and hold in place across a billion years, this is the moment the graffiti wall stops being able to follow the story at all, and the reader is carried past it into something the wall was never built to picture.