Prion propagation is not just one-shot templating (a misfolded monomer touching a normal one). Once a small amyloid fibril (a "propagon") exists, it grows by capturing normal, natively-folded monomers onto its exposed ends and templating their conformation onto the aggregate — nucleated-conformational-conversion polymerization:
dL/dt = k+ · [monomer] · (number of growing ends)
Fragmentation rate per fibril ≈ k_frag · (L − 1)
New propagons created by a break = +1 per fragmentation event
Long fibrils are fragile. A chaperone system (in yeast, Hsp104; a related surveillance exists for mammalian PrP aggregates) periodically shears a fibril into two shorter ones. Each break creates a brand-new growing end — a new propagon — so fragmentation, not just growth, is what multiplies the infectious/heritable unit. This is the Cox–Tuite / Masel–Jansen–Nowak model for heritable yeast prions such as [PSI+], and the same growth-and-scission logic underlies mammalian PrPSc amplification assays (PMCA).
- Elongation rate k+ — how fast free monomer is captured onto fibril ends.
- Fragmentation rate k_frag — how often fibrils are cut; too low and propagons cannot keep pace with dilution.
- Cell-division interval — every division randomly partitions existing propagons roughly 50/50 between "daughters" (only one is rendered); this dilutes propagon number every generation.
- Inject Misfolded Seed — adds one short fibril, mimicking exposure to an infectious/misfolded template.
The outcome is a threshold effect: if k_frag is too low relative to the division rate, stochastic loss during partitioning can drive propagon count to zero — the population "cures" itself ([psi-]) even though monomer is still abundant. If k_frag is high enough, propagon number is replenished faster than dilution removes it, and the prion state is mitotically stable indefinitely.