Title:Modelling autophagy selectivity by receptor clustering on peroxisomes

Abstract:When subcellular organelles are degraded by autophagy, typically some, but not all, of each targeted organelle type are degraded. Autophagy selectivity must not only select the correct type of organelle, but must discriminate between individual organelles of the same kind. In the context of peroxisomes, we use computational models to explore the hypothesis that physical clustering of autophagy receptor proteins on the surface of each organelle provides an appropriate all-or-none signal for degradation. The pexophagy receptor proteins NBR1 and p62 are well characterized, though only NBR1 is essential for pexophagy (Deosaran {\em et al.}, 2013). Extending earlier work by addressing the initial nucleation of NBR1 clusters on individual peroxisomes, we find that larger peroxisomes nucleate NBR1 clusters first and lose them due to competitive coarsening last, resulting in significant size-selectivity favouring large peroxisomes. This effect can explain the increased catalase signal that results from experimental siRNA inhibition of p62. We also consider receptor cluster formation on individual peroxisomes with enhanced levels of ubiquitin and find that size-selectivity, but not cluster formation, is suppressed. Thus, our selectivity mechanism does not hinder selection of individual damaged organelles via enhanced ubiquitination. NBR1 cluster formation provides a viable physical mechanism for all-or-none substrate selectivity in pexophagy.