Yeast Kex1p is a Golgi-associated membrane protein: deletions in a cytoplasmic targeting domain result in mislocalization to the vacuolar membrane

A number of models could explain the observation that the membrane-associated mutant forms of Kexlp are deh'vered to the vacuole while the soluble truncated forms of Kexl are exported to the cell surface (Fig. 9) . (1) The first model suggests that membrane-associated truncated forms of Kexlp are misfolded, and as such may be targeted, via an unknown mechanism, to the vacuole for degradation. This model seems unlikely as all of the truncated forms of Kexlp, both soluble and membrane associated, have similar total pro-tease activity to that of the wild-type protein, suggesting that at a minimum the catalytic domain of the mutants has folded to a conformation similar to that of wild type. In addition, these forms of Kexlp all exit the ER and reach the Golgi apparatus where they are both glycosylated and process killer toxin. The misfolding targeting model would be unusual in that it must explain the mislocalization of the membraneassociated mutant forms of Kexlp yet allow the soluble forms to be secreted. Also, such a garbage pathway for delivery of misfolded mutant forms of Kexlp cannot explain the vacuolar localization of highly expressed but presumably correctly folded Kexlp. (2) The second model proposes that the membrane-spanning domain of Kexlp contains a latent or cryptic targeting signal for the vacuole. Studies are currently under way to address this possibility. It should be noted that although Kexlp is homologous to the vacuolar protein CPY, the homology does not extend to include the vacuolar targeting signal found in proCPY. (3) The third model involves Kexlp-Hpa (-Bcl, -Hinc) being secreted to the plasma membrane by default, where it is then endocytosed and delivered to the vacuole. Such a "transient appearance" of Kexlp-Hpa (-Bcl, -Hint) at the cell surface would not be detected by the activity assay used or by indirect immunofluorescence. Soluble Kexlp-AS, having reached the cell surface, would be released into the medium and, hence, would not be endocytosed to the vacuole. (4) The final model proposes that the vacuole is the direct default destination for membraneassociated proteins that enter the secretory pathway, such that membrane proteins lacking positive targeting/retention signals would be delivered to the vacuole.