Evidence tying SorLA to Alzheimer’s has grown since James Lah and colleagues at Emory University, Atlanta, reported its reduction in AD neurons (see Scherzer et al., 2004). Just last month, this group added new data that its expression is down in cases of mild cognitive impairment from the Religious Orders Study, as well (see Sager et al., 2007). Also this year, SORL1, the SorLA gene, climbed to seventh place in the AlzGene top 10 list (see AlzGene entry), as genetic studies led by Peter St. George-Hyslop, University of Toronto (see Rogaeva et al., 2007), and Richard Mayeux, Columbia University, New York (see Lee et al., 2007), turned up associations between SORL1 polymorphisms and AD (see ARF related news story). But the biology of SorLA has remained unclear. To explore it, Willnow and colleagues tested a variety of SorLA constructs for their ability to alter APP localization and processing.

First author Vanessa Schmidt and colleagues first examined how wild-type SorLA affects APP compartmentalization in Chinese hamster ovary and SH-SY5Y cells. When they coexpressed the sorting receptor with APP, the amount of mature, O-glycoslyated APP increased, a sign it might be accumulating in the Golgi because glycosylation occurs there. Live imaging confirmed this: in cells expressing APP tagged with green-fluorescent protein, APP rapidly exited the trans-Golgi network—60 percent of it cleared in 12 minutes. But when the sorting receptor was coexpressed with the tagged APP, transit of the green fluorescence slowed to a trickle. The findings indicate that SorLA can dramatically slow APP’s transit through the Golgi network.

To see how closely SorLA and APP compartmentalization are linked, the researchers next introduced a two-lysine motif (KKLN) to the carboxy terminal of SorLA to act as an ER retention signal. Accordingly, that is where they found most of the protein. Interestingly, when SorLA(KKLN) was coexpressed with APP, the latter also accumulated in the ER. This shift was accompanied by a 65 percent reduction in sAPPα, the APP α-secretase cleavage product, and a 95 percent reduction in Aβ40 production, indicating how strongly SorLA localization can affect APP cleavage and, with that, the upstream sources of AD pathology.

SorLA, of course, does not work alone. It carries cytoplasmic domain motifs that bind, for example, GGAs (for Golgi-localized, γ ear-containing, ADP ribosylation factor-binding proteins) and PACS-1 (phosphofurin acidic cluster sorting protein-1). Both are adaptor proteins involved in trafficking cargo to and from the trans-Golgi network. To ask how important these potential interactions might be, Schmidt and colleagues made SorLA mutants to abolish GGA binding (SorLAgga), PACS-1 binding (SorLAacidic), or both (SorLAΔcd). They found that while all these mutants could still bind APP, none of them were capable of retaining its mature form in the Golgi. This confirmed that wild-type SorLA has the unique property to accumulate mature APP molecules in the Golgi, most likely by extending the transit time of the precursor through this compartment, write the authors.

Compared to wild-type SorLA, the mutant forms tended to end up on the cell surface. Additionally, SorLAgga was enriched in endosomes. Such redistribution, if it occurred in vivo, could have important consequences for APP processing. The researchers found that overexpressing SorLAgga led to a significant reduction in Aβ40 production and an increase in sAPPα, indicating a switch from amyloidogenic to non-amyloidogenic processing. In contrast, SorLAacidic or SorLAΔcd had the opposite effect—there was a massive increase in Aβ40, while sAPPα remained unchanged. Aβ42 is normally below detection limits in CHO cells expressing APP (with or without SorLAwt), but this form became readily detectable when the cells expressed SorLAs lacking either the C-terminal or the acidic motif. All told, the data illustrate how SorLA’s localization dramatically influences APP compartmentalization and processing.

Do these laboratory SorLA mutations have any physiological correlates? “One can envision a situation where sequence variations that affect adaptor binding or overall expression levels of SorLA in individuals may have dramatic consequences for APP processing kinetics and onset of sporadic AD,” the authors write. In this regard, Brad Hyman, Massachusetts General Hospital, Boston, and collaborators showed that overexpression of GGA-1 can decrease Aβ production (see von Arnim et al., 2006). Jochen Walter at the University of Bonn, Germany, reported that GGA1 is preferentially expressed in neurons in the brain and is also found in activated microglia surrounding plaques (see Wahle et al., 2006). The role of GGA-1 is complicated by its involvement with β-secretase, as well (see ARF related news story and Tesco et al., 2007). As for SorLA levels, the mutations that have been so far linked to AD are not within the coding region of the gene, suggesting that they may influence expression.—Tom Fagan.

Reference:
Schmidt V, Sporbert A, Rohe M, Reimer T, Rehm A, Andersen OM, Willnow TE. SorLA/LR11 regulates processing of amyloid precursor protein via interaction with adaptors GGA and PACs-1. J. Biol. Chem. 2007, September 12. Abstract