The narrative on APP phosphorylation began in 1988 with the discovery that threonine 654 and serine 655 were potentially phospho-acceptor sites for protein kinase C (PKC) and Ca2+/calmodulin-dependent protein kinase II (CaMKII) (Gandy et al., 1988). Based on that lead, the effects of PKC activation on APP metabolism were discovered. The impact of acute PKC activation was stimulation of α-secretase cleavage and release of sAPPα (Buxbaum et al., 1990; Caporaso et al., 1992). Mutagenesis of the APP cytoplasmic tail showed that direct APP phosphorylation did not play an obvious role in phosphorylation-stimulated α-secretase cleavage (da Cruz e Silva et al., 1993), but both phosphorylation and α-secretase cleavage of APP have continued to receive substantial attention over the ensuing 17 years. “Regulated α-secretase cleavage” provided the first mechanism for decreasing Aβ generation (  Read more

The narrative on APP phosphorylation began in 1988 with the discovery that threonine 654 and serine 655 were potentially phospho-acceptor sites for protein kinase C (PKC) and Ca2+/calmodulin-dependent protein kinase II (CaMKII) (Gandy et al., 1988). Based on that lead, the effects of PKC activation on APP metabolism were discovered. The impact of acute PKC activation was stimulation of α-secretase cleavage and release of sAPPα (Buxbaum et al., 1990; Caporaso et al., 1992). Mutagenesis of the APP cytoplasmic tail showed that direct APP phosphorylation did not play an obvious role in phosphorylation-stimulated α-secretase cleavage (da Cruz e Silva et al., 1993), but both phosphorylation and α-secretase cleavage of APP have continued to receive substantial attention over the ensuing 17 years. “Regulated α-secretase cleavage” provided the first mechanism for decreasing Aβ generation (Buxbaum et al., 1993; Hung et al., 1993), and a host of hormones and neurotransmitters have been implicated as dynamic influences on Aβ generation via the regulated cleavage process (reviewed in Small and Gandy, 2006).

Two new papers bring these phenomena back to light and integrate both narratives with the protein-sorting biology surrounding the Vps10-domain-containing sortilins (Andersen et al., 2005; Rogaeva et al., 2007) and the Vps35 component of the retromer (Small et al., 2005). Vps stands for vacuolar protein sorting protein (based on homology with yeast counterparts), and the retromer is a vesicular structure responsible for retrograde transport of cargos from the endosome to the trans-Golgi network. In a related development, Willnow and colleagues reported at the International Conference on Alzheimer’s Disease in July 2010 that SorL1 modulation of APP metabolism required interaction of the FANSHY domain of the SorL1 tail with the Vps35 component of the retromer. Small had proposed that SorL1 acted as an adaptor, linking APP to Vps35 (Small and Gandy, 2006), but Willnow was the first to provide proof. Recently, we reported that the type 2 diabetes gene, SorCS1, encoded another Vps10 domain protein that modulated APP metabolism (Lane et al., 2010; and see ARF related news story). This not only provided a molecular mechanistic link between diabetes and AD, but other data we reported in that paper implicated SorL1 and Vps35 in the action of SorCS1 on APP metabolism. In short, SorL1 and SorCS1 are now both genetically and mechanistically linked to AD and both control APP metabolism via the Vps35 component of the retromer. Deficiency of retromer function (secondary to hypomorphic SorL1 or SorCS1) presumably extends the residence time for APP and secretases colocalized in endosomes, thereby enhancing Aβ generation.

Now, Vieira et al. provide yet another wrinkle. The interaction of APP with the retromer is apparently dependent upon the phosphorylation state of APP at serine 655. The phospho-mimetic APP S655E (glutamate substituted for serine) shows enhanced retrieval from the endosome and decreased Aβ generation, while the dephosphomimetic APP S655A undergoes extended endosomal residence and enhanced catabolism. The sorting of APP and the missorting of APP mutants were shown to be Vps35-dependent and, by implication, retromer-dependent. Downregulation of Vps35 appeared to abolish retromer-dependent APP sorting, as one would predict.

In summary, work from several labs, each pursuing somewhat different leads, has converged to establish several new concepts in the field: 1) The retromer is an essential component of the mechanism of SorL1-related and SorCS1-related AD; 2) A role for the phosphorylation status of APP Ser655 in modulating retrieval of APP from the endosome via Vps35 and the retromer illuminates a longstanding mystery about the possible physiological significance of this potential phospho-acceptor site; and 3) SorCS1/SorL1/Vps35 function is essential not only for APP sorting, but for some mechanism underlying type 2 diabetes. A role for these factors in specifying insulin sensitivity is predicted but is as yet unproven.

View all comments by Samuel Gandy

The study by Vieira and colleagues investigates a potential role for APP phosphorylation in regulation of retrograde transport from endosomes to the trans-Golgi network (TGN). Recently, this trafficking route received particular attention because of the association of amyloidogenic processing with late endosomal compartments.

Several components of the retrograde sorting pathways, including the adaptor complex retromer and the sorting receptor SORLA/SORL1, have been implicated in intracellular transport and processing of APP, and in AD-related pathologies in animal models and in humans. Taken together, these findings argue that altered trafficking of APP between endosomes (where APP processing occurs) and the Golgi may be a molecular mechanism underlying enhanced amyloidogenic processing in sporadic AD.

Now, this work by Vieira et al. identifies interesting new details of the regulatory processes that may control transport of APP between endosomes and the TGN. Using immunofluorescence microscopy, the authors investigate the intracellular trafficking of APP-GFP fusion...  Read more

The study by Vieira and colleagues investigates a potential role for APP phosphorylation in regulation of retrograde transport from endosomes to the trans-Golgi network (TGN). Recently, this trafficking route received particular attention because of the association of amyloidogenic processing with late endosomal compartments.

Several components of the retrograde sorting pathways, including the adaptor complex retromer and the sorting receptor SORLA/SORL1, have been implicated in intracellular transport and processing of APP, and in AD-related pathologies in animal models and in humans. Taken together, these findings argue that altered trafficking of APP between endosomes (where APP processing occurs) and the Golgi may be a molecular mechanism underlying enhanced amyloidogenic processing in sporadic AD.

Now, this work by Vieira et al. identifies interesting new details of the regulatory processes that may control transport of APP between endosomes and the TGN. Using immunofluorescence microscopy, the authors investigate the intracellular trafficking of APP-GFP fusion proteins in cultured cells. In particular, they demonstrate that a phosphomimetic mutant of APP at serine 655 (resembling constitutively phosphorylated APP) displays enhanced retrograde sorting and a tendency for decreased amyloidogenic processing. In contrast, a dephosphomimetic variant of APP at S655 shows decreased retrograde transport and accelerated lysosomal targeting. These findings argue that retrograde sorting of APP is influenced by phosphorylation, a mechanism proposed to involve the retromer and SORLA/SORL1.

In the future, it may be interesting to elaborate on the molecular details of this pathway, for example, by clarifying phosphorylation-dependent interaction of APP with retromer components and/or SORLA. Initial observations using co-immunoprecipitation are provided in this study. Also, it may be important to transfer findings in Cos7 cells to the in vivo situation in animal models where the effects of APP S665 phosphorylation on amyloidogenic processing may be investigated in the context of the brain.

View all comments by Vanessa Schmidt
View all comments by Thomas Willnow