01 Feb 2023

A new study offers up yet more evidence that the endosome may be ground zero for Alzheimer’s disease pathogenesis. In Proceedings of the National Academy of Sciences on January 18, scientists led by Olav Andersen of Aarhus University in Denmark and Gregory Petsko of Brigham and Women’s Hospital in Boston report that SORL1—an endosomal recycling protein with strong genetic ties to AD—forms dimers within the endosome. This molecular liaison, which can happen via either of two domains at opposite ends of the protein, attracts the retromer, a multi-subunit complex that ushers various proteins out of endosomal compartments. Together, SORL1 and the retromer then whisk APP out of the endosome, quelling production of amyloidogenic Aβ peptides. That SORL1 may simultaneously dimerize at opposite ends hints that it forms polymers within the endosomal membrane, stabilizing retromer complexes that facilitate protein recycling, the authors propose. The findings could also explain why risk variants are scattered across the SORL1 gene.

SORL1 forms dimers on the endosomal membrane.
These enhance retromer function, APP recycling.
Modeling predicts SORL1 forms polymers, stabilizing retromer complexes.

“This study provides new and important information about how SORL1 can form dimeric structures important for recruiting APP into membrane tubules for endosomal transport,” wrote Brett Collins of the University of Queensland in Brisbane, Australia.

“These new insights could pave the way for designing either molecular or genetic tools that target these pathogenic pathways in Alzheimer’s disease.”

“With perturbed retromer function already closely associated with onset of AD, this thought-provoking study will be of significant interest to the AD community and, more broadly, those interested in the field of retromer biology and endosomal cargo sorting,” wrote Peter Cullen of the University of Bristol, U.K.

When paired with the retromer complex, sortilin-related receptor 1 (SORL1), aka SORLA, binds and transports proteins from the endosome to other cellular locales such as the trans-Golgi network or plasma membrane (Jan 2022 news). The gene is an AD hotspot, housing variants believed to cause autosomal dominant AD as well as those that up risk for the disease (Jan 2007 news; Jun 2018 news). These harmful variants are distributed throughout the gene. In the protein’s large lollipop-like ectodomain, VPS10p and YWTD domains top a stalk of 11 complement-type receptor domains and six fibronectin type III (3Fn) domains. A small endodomain sits in the cytosol (image below). While the CR region is known to bind cargo including APP, the diminutive cytoplasmic tail is tasked with snagging the retromer (Jun 2012 news). The functions of the VPS10p and 3Fn domains are unclear, although in other proteins, these domains can dimerize. Recent studies suggest that SORL1 can dimerize as well (Zhang et al., 2022). Might these domains play a part in that?

**SORL1 Domains.** The lion’s share of the large SORL1 protein comprises the ectodomain, which consists of several domains that play roles in cargo binding and, as reported here, in dimerization. The cytoplasmic tail (bottom) binds the retromer complex. *[Courtesy of Jensen et al., PNAS, 2023]*

To investigate, first author Anne Jensen and colleagues launched a barrage of molecular and biochemical techniques, including the construction of a SORL1 “mini-receptor” comprising the 3Fn, transmembrane, and cytoplasmic tail domains. They found that the mini-receptor could dimerize, and that the 3Fn domains facilitated this pairing, both in vitro and within cells. Next, using a proximity ligation assay, which can detect close contacts between tagged proteins via fluorescence, the researchers found that SORL1 only forms dimers when located on the endosomal membrane. Importantly, these dimers co-localized with the retromer complex, which gathered on the cytoplasmic face of endosomes wherever SORL1 pairs were spotted. Indeed, they found that SORL1 dimers in the endosome existed almost entirely in complex with retromer.

How might SORL1 dimerization influence the recycling and processing of APP? To find out, the researchers overexpressed APP and either full-length SORL1 or the 3Fn-containing mini-receptor, in human neuroblastoma cells. Full-length SORL1 led to a dip in both amyloidogenic and non-amyloidogenic processing of APP. The scientists believe this stems from removal of APP from the endosomal compartment into the trans Golgi network, where it becomes waylaid or routed to the lysosomes for degradation. Transport to the TGN also keeps APP away from amyloidogenic, β-secretase cleavage in the endosomes, and from non-amyloidogenic, α-secretase cleavage in the plasma membrane. Surprisingly, overexpression of the mini-receptor, which can dimerize with itself and also with bind endogenous SORL1, increased non-amyloidogenic processing, but decreased amyloidogenic processing, thus reducing levels of Aβ42. The effects of the mini-receptor are reminiscent of those of previously reported retromer chaperones that route APP back to the plasma membrane (Apr 2014 news; Jan 2020 news).

“These results have multiple therapeutic implications,” wrote Samuel Gandy of Icahn School of Medicine at Mount Sinai in New York (comment below). “In addition to the chemical retromer stabilizers that this group have reported, it is now possible to consider an alternative strategy in which gene therapy might be used to deliver the SorL1 mini-receptor to the brains of AD patients.”

Strikingly, Jensen and colleagues found that in addition to linking up via its 3Fn domains, SORL1 dimerizes via VPS10p, a domain reported to bind Aβ peptides (Feb 2014 news). Using computational modeling, she found that SORL1 could theoretically use both of these domains to form polymers on the endosomal membrane (image below). When coupled with retromers, such SORL1 polymers could drive the formation of endosomal tubules—projections of the membrane formed by networks of retromer complexes that ultimately pinch off, delivering endosomal cargo to other compartments. Andersen emphasized that they have yet to obtain direct physical evidence of SORL1 polymers.

Polymeric Possibilities. By simultaneously dimerizing via its VPS10p (red asterisk) and its 3Fn (blue asterisk) domains, SORL1 could theoretically form polymers on the endosomal membrane, facilitating polymerization of retromer complexes on the cytoplasmic face, which in turn form endosomal tubules. [Courtesy of Jensen et al., PNAS, 2023.]

Still, he interprets the findings as clear evidence of the intimate collaboration between SORL1 and retromer in driving endosomal recycling, and suggests that SORL1 dimerization is a critical mechanistic step of this pathway. Moreover, Andersen emphasized that SORL1 cargo proteins, including APP, glutamate receptors, and TREKB, also form homodimers, suggesting that dimerization could be a critical regulatory feature of endosomal recycling.

Jessica Young of the University of Washington in Seattle noted that both the 3Fn and VPS10p domains harbor pathogenic variants in SORL1. “This paper is strongly suggestive that these variants may cause AD by disrupting endosomal recycling,” she wrote. “This work paves the way for new targets for AD that focus on the SORL1-retromer interaction as a tractable unit of the overall pathway for therapeutic design.” Young said that future studies should focus on how this unit works to recycle other proteins that that are implicated in AD-related neurodegeneration, including glutamate receptors and neurotrophin receptors.

Taking a broad perspective, Ralph Nixon of New York University in Orangeburg noted that defects in two reciprocal forces—entry and exit of proteins into endosomes—contribute to the bulging of these compartments observed in the early stages of AD pathogenesis (Cataldo et al., 2000). Nixon reported that accumulation of β-CTF, a product of β-secretase cleavage of APP that occurs within the endosome, activates Rab5, a GTPase that accelerates endocytosis (Jul 2015 news; Jan 2021 news). To his mind, the current study suggests that defects in SORL1 function not only slow endosomal egress, but also accelerate endosomal entry by way of increasing amyloidogenic processing of APP, and thus stoking levels of β-CTF and Rab5 activation. “The emerging biology integrates a recycling deficit, as shown here, with a cargo entry acceleration mediated by β-CTF/Rab5,” Nixon said.—Jessica Shugart

Comments

- Brett Collins
  The University of Queensland
- Posted: 01 Feb 2023
The sortilin-related receptor (SORL1) is a large neuronal protein significantly implicated in the pathogenesis of Alzheimer’s disease. One of its well-studied roles is to transport the amyloid precursor protein (APP) through cellular compartments called endosomes, and when this is disrupted, cells produce increased levels of the toxic, APP-derived peptide Aβ. This study by Jensen and colleagues provides new and important information about how SORL1 can form dimeric structures important for recruiting APP into membrane tubules for endosomal transport. This occurs through another endosomal protein complex linked to Alzheimer’s called retromer. These new insights could pave the way for designing either molecular or genetic tools that target these pathogenic pathways in Alzheimer’s disease.

View all comments by Brett Collins

- Peter Cullen
  University of Bristol
- Posted: 01 Feb 2023
SORLA is an integral membrane protein linked to early and late-onset AD. In this study, Jensen and colleagues used a variety of approaches, including GFP nanotraps, bimolecular fluorescence complementation assays, and molecular modelling of the SORLA structure, to define how the VPS10P domain and the fibronectin-type III repeats within the ectodomain drive SORLA dimerization.

As an integral membrane protein, SORLA undergoes sorting and transport through the endosomal network. Here, it is orientated such that its ectodomain is facing into the endosomal lumen and its short carboxy terminal “tail” is exposed to the cytoplasm. The authors propose that SORLA dimerization serves to influence the assembly of the retromer coat on the cytoplasmic facing endosomal leaflet to facilitate the retromer-mediated sorting of SORLA and its associated protein APP through the endosomal network. Indeed, they go one step further to show that disruption of SORLA dimerization leads to an increase in amyloidogenic processing of APP, thereby linking the observations back to AD.

With perturbed retromer function already closely associated with onset of AD, this thought-provoking study will be of significant interest to the AD community and, more broadly, those interested in the field of retromer biology and endosomal cargo sorting.

View all comments by Peter Cullen

- Sam Gandy
  Icahn School of Medicine at Mount Sinai
- Posted: 01 Feb 2023
The VPS10 family of receptors is comprised of sortilin, SorL1, SorCS1, SorCS2, and SorCS3, and each of these plays one or more roles in the sorting and processing of integral membrane proteins as they move through the central vacuolar pathway. Many of these molecules have also been implicated in the pathogenesis of major neurodegenerative disorders. SorL1 (SorLA) is associated with Alzheimer’s disease (AD), while sortilin is associated with frontotemporal lobar degeneration (FTLD). SorL1 was first identified as a receptor for apolipoprotein E, and the implication that SorL1 played a role in AD was established by the identification of variants that segregated with AD in genome-wide association studies (Rogaeva et al., 2007). These VPS10 family members also interact with the VPS35 component of the retromer, a molecule also implicated in the pathogenesis of AD by Small et al. (2005). The underlying mechanism of action of VPS10 proteins is to control the partitioning of transmembrane cargoes between the trans Golgi network (TGN) and the endosomal system. The Alzheimer’s amyloid precursor protein (APP) is one such transmembrane protein cargo for transport by SorL1 and the VPS35 retromer. However, what has been missing until now is the molecular mechanism underlying how SorL1 and VPS35 interact.

This new paper by Jensen et al. promises to change that. Using a SorL1 fragment that includes the fibronectin-type-III, transmembrane, and cytoplasmic domains (a fragment known as the SorL1 mini-receptor), these authors demonstrate directly that these domains of SorL1 can dimerize, and that this is the key event in enhancing Vps35-retromer-dependent sorting of APP, including specification of the residence time of APP in the amyloidogenic pathway. In intact cells, SorL1 dimers, such as those generated by Jensen and colleagues, are selectively localized in the endosomal compartment, while monomers tend to be more abundant in the TGN.

Several groups had previously demonstrated that the overexpression of full-length SorL1 decreased APP trafficking in the secretory pathway, while also assisting APP in recycling out of the endosome. To investigate the consequences of SorL1 dimerization, Jensen et al. transfected N2a cells with various combinations of human APP, together with SorL1, in the absence or presence of the SorL1 mini-receptor. In line with previous results, SorL1 alone decreased both amyloidogenic and non-amyloidogenic processing. However, when both SorL1 and the mini-receptor were present, Jensen observed decreased processing of APP through both pathways, with a surprising elevation in the level of the non-amyloidogenic APP carboxyl-terminal fragment. They concluded that this combination represented a phenocopy of the retromer stabilization that some of these same investigators had induced with small molecules in a previous study (Mecozzi et al., 2014).

To visualize how these phenomena were occurring at the molecular level, Jensen turned to the available crystal structure of the SorL1 VPS10 and found that several alternative structures were predicted: one model held that local dimers could be held together by the fibronectin-type-III domain or the Vps10 domain (or both), while the other model predicted that dimerization sites from different molecules led to formation of a large polymer of SorL1 molecules that coat endosomal tubules. To determine which of these models was more likely, Jensen turned to the AlphaFold protein folding prediction platform. The original modeling of SorL1 structure turned out to be flawed, so AlphaFold was applied selectively to the SorL1 ectodomain. In the newer iteration of AlphaFold, Jensen determined that the more likely model was the latter (i.e., the one that predicted polymerization of SorL1 and the retromer on endosomal tubules).

These results have multiple therapeutic implications. In addition to the chemical retromer stabilizers that this group has reported, it is now possible to consider an alternative strategy in which gene therapy might be used to deliver the SorL1 mini-receptor to the brains of AD patients. This would expand the list of potential AD gene therapy strategies beyond the current use of adenoviral-associated viruses to deliver APOE ε2 DNA to treat symptomatic APOE ε4 homozygotes (Kaplitt et al., 2022).

References:

Kaplitt M, Leopold P, Noch E, Ivanidze J, Chazen L, Crystal R, Kaminsky S, Bowe H, Wang M, Ballon D, Dyke J, Sondhi D, Gandy S, Giannantoni-Ibelli G, Barth JA. A Phase 1, Open-Label, 52-Week, Multicenter Study to Evaluate the Safety and Biochemical Efﬁcacy of AAV Gene Therapy (LX1001) in Patients with APOE4 Homozygote Alzheimer’s Disease. Interim Data. JPAD 2022; 9(1): S33. JPAD CTAD Abstracts

Mecozzi VJ, Berman DE, Simoes S, Vetanovetz C, Awal MR, Patel VM, Schneider RT, Petsko GA, Ringe D, Small SA. Pharmacological chaperones stabilize retromer to limit APP processing. Nat Chem Biol. 2014 Jun;10(6):443-9. Epub 2014 Apr 20 PubMed.

Rogaeva E, Meng Y, Lee JH, Gu Y, Kawarai T, Zou F, Katayama T, Baldwin CT, Cheng R, Hasegawa H, Chen F, Shibata N, Lunetta KL, Pardossi-Piquard R, Bohm C, Wakutani Y, Cupples LA, Cuenco KT, Green RC, Pinessi L, Rainero I, Sorbi S, Bruni A, Duara R, Friedland RP, Inzelberg R, Hampe W, Bujo H, Song YQ, Andersen OM, Willnow TE, Graff-Radford N, Petersen RC, Dickson D, Der SD, Fraser PE, Schmitt-Ulms G, Younkin S, Mayeux R, Farrer LA, St George-Hyslop P. The neuronal sortilin-related receptor SORL1 is genetically associated with Alzheimer disease. Nat Genet. 2007 Feb;39(2):168-77. PubMed.

Small SA, Kent K, Pierce A, Leung C, Kang MS, Okada H, Honig L, Vonsattel JP, Kim TW. Model-guided microarray implicates the retromer complex in Alzheimer's disease. Ann Neurol. 2005 Dec;58(6):909-19. PubMed.

View all comments by Sam Gandy

- Jessica Young
  University of Washington
- Posted: 01 Feb 2023
Endosomal dysfunction is a consistent and early cellular phenotype in AD. The gene SORL1 codes for an endosomal receptor (SORLA) that is strongly associated with increased AD risk. In this study, Jensen and colleagues further elucidate the functional interactions of SORL1 with the multi-protein sorting complex retromer. Using elegant biochemical studies and structural analysis, the group shows how SORLA, a very large protein, can fold to fit into an endosomal tubule. Furthermore, they identify key domains that can dimerize and stabilize retromer, thereby enhancing retromer-related endosomal recycling of APP. This has a beneficial effect of reducing amyloidogenic cleavage by APP. This work is reminiscent of retromer chaperones, small molecules that stabilize the cargo-recognition core of retromer and enhance its function, in particular reducing APP cleavage and tau phosphorylation (Mecozzi et al., 2014; Young et al., 2018).

Here, for the first time, the authors identify the distinct domains of SORL1 that can dimerize and promote stabilization of the SORL1-retromer complex—the 3Fn domains and the VPS10 domains. As each of these has been shown to harbor pathogenic variants in SORL1, this is strongly suggestive that these variants may cause AD by disrupting endosomal recycling. The endosomal system is highly complex and thinking how to design therapeutic options for this pathway is challenging. However, this work paves the way for new targets for AD that focus on the SORL1-retromer interaction as a tractable unit of the overall pathway for therapeutic design. Important future studies should focus on how this unit works to recycle other proteins that that are implicated in AD-related neurodegeneration, including glutamate receptors and neurotrophin receptors, both of which are essential for neuronal health and function.

References:

Mecozzi VJ, Berman DE, Simoes S, Vetanovetz C, Awal MR, Patel VM, Schneider RT, Petsko GA, Ringe D, Small SA. Pharmacological chaperones stabilize retromer to limit APP processing. Nat Chem Biol. 2014 Jun;10(6):443-9. Epub 2014 Apr 20 PubMed.

Young JE, Fong LK, Frankowski H, Petsko GA, Small SA, Goldstein LS. Stabilizing the Retromer Complex in a Human Stem Cell Model of Alzheimer's Disease Reduces TAU Phosphorylation Independently of Amyloid Precursor Protein. Stem Cell Reports. 2018 Mar 13;10(3):1046-1058. Epub 2018 Mar 1 PubMed.

View all comments by Jessica Young