Independent research teams led by Worley and coauthor Dietmar Kuhl of the University Medical Center Hamburg-Eppendorf, Germany, identified Arc (aka Arg3.1) in 1995 (Lyford et al., 1995; Link et al., 1995), and today’s Cell paper grew out of a longstanding quest to understand how Arc and other immediate early genes work at synapses to confer plasticity. Arc is dramatically upregulated in neurons engaged in information processing (Guzowski et al., 1999; Guzowski et al., 2000). It associates with endosomes to regulate AMPA receptor trafficking (Chowdhury et al., 2006), and helps these receptors maintain synaptic strength in a process known as homeostatic scaling (see Shepherd et al., 2006 and ARF related news story). Fitting with these roles, mice lacking Arc cannot form long-term memories (Plath et al., 2006).

Conceptually, this line of work converged with AD research when researchers found that increasing synaptic activity drove up Aβ levels in hippocampal slices of AD transgenic mice (Kamenetz et al., 2003) and in the brain interstitial fluid of living mice (Cirrito et al., 2005), and that the latter depended largely on endocytosis (see ARF related news story on Cirrito et al., 2008).

First author Jing Wu and colleagues wondered whether activity-dependent Aβ generation would occur differently in Arc knockout mice. “That was really our starting point,” Worley said. As it turns out, the process was not just different—it was gone. When seizures or drugs were administered to Arc-deficient mice to send hippocampal neurons into overdrive, Aβ levels did not budge.

To explore the mechanism, the researchers put fluorescently tagged Arc into mouse hippocampal neurons and watched the protein make its way through the cell. By immunohistochemistry, they saw Arc colocalize with presenilin-1 (PS1) in dendritic endosomes. Anti-Arc immunoprecipitation of mouse brain extract pulled down the PS1 N-terminal fragment—but not its C-terminus or other γ-secretase components or BACE1—confirming the Arc/PS1 interaction and mapping it to the N-terminal end.

Combined with electron microscopy analysis of rat hippocampus, “the cell biology data make the case that the Arc/PS1 interaction promotes Aβ generation by recruiting γ-secretase to recycling endosomes that process APP,” Worley told ARF. Consistent with this idea, crossing APPswe/PS1deltaE9 transgenic mice onto an Arc-deficient background reduced soluble Aβ40 and plaque load.

Is any of this physiologically relevant? Looking at human brain extracts, the researchers found twice as much Arc in the medial frontal cortex of AD patients (n = 26) compared to that of age-matched controls (n = 14). No Arc differences showed up in the visual cortex, a brain area spared in AD.

All told, the paper suggests that Arc is essential for activity-driven Aβ generation. “Connecting the activity dependence of Aβ formation to the mechanisms controlling plasticity is interesting and should provide new directions of research,” commented Roberto Malinow of the University of California, San Diego. However, he and others said further work will be needed to tease out where Arc is active. “Arc is primarily a post-synaptic protein, whereas Aβ appears to be released both axonally and dendritically,” said John Cirrito at Washington University School of Medicine in St. Louis, Missouri. “Whether Arc impacts both those compartments, or just acts dendritically, remains in question.”

Another question for future studies, Cirrito said, is whether Arc has any relationship with PICALM or BIN1—endocytic factors that were linked to sporadic AD risk in recent genomewide association studies (ARF related news story on Naj et al., 2011 and Hollingworth et al., 2011). Arc does not appear in the AlzGene database of AD risk genes.

Brewing a New Model for Aβ Toxicity
The importance of PICALM and other endocytic trafficking proteins did surface in the Science paper on a yeast model for Aβ toxicity. Though yeast cells are a far cry from the highly specialized neurons involved in AD, they have the basic vesicular trafficking compartments common to eukaryotic cells, senior investigator Lindquist noted. Her yeast models have proven useful for research on other protein aggregation disorders including Parkinson’s and Huntington’s (see ARF related news story on Cooper et al., 2006; ARF related news story on Gitler et al., 2009; ARF related news story on Outeiro and Lindquist, 2003 and Willingham et al., 2003). Yeast prion proteins fold into oligomers shaped like amyloid-β and other pathogenic peptides (Serio et al., 2000; Kayed et al., 2003; Shorter and Lindquist, 2004). “We had good reason to believe these conformational transitions are really an ancient problem,” Lindquist told ARF.

To model Aβ toxicity in yeast, first author Sebastian Treusch and colleagues directed Aβ1-42 to the endoplasmic reticulum sans ER retention signal, then let the peptide ride out of the cell along the secretory pathway. The yeast cell wall keeps the Aβ from diffusing out, and instead allows it to interact with the plasma membrane and transit back into the cell through endocytic compartments. Using stable Aβ1-42 transformants to screen a near-complete yeast genomic library containing some 6,000 genes, the researchers came up with 23 suppressors and 17 enhancers of Aβ toxicity. The former accelerated yeast growth, whereas the latter slowed it. Of the 40 modifiers, 12 had clear human homologs, three with roles in clathrin-mediated endocytosis. One of these is the yeast version of PICALM, one of the top late-onset AD risk genes. The other two have human homologs that interact with BIN1 and CD2AP, which also came up as top hits in recent AD GWAS. Treusch validated all three endocytosis genes in worm glutamatergic neurons and rat cortical neurons, where they modified Aβ toxicity in the same direction as in the yeast.

The research “provides a great system for understanding why PICALM is linked to AD, perhaps pointing to a currently unknown molecular mechanism underlying Aβ toxicity,” noted Sam Gandy of Mount Sinai Medical Center in New York City. As for whether yeast are too simple to model mammalian neurons, he said, “that is a caveat for any model system (yeast, fly, worm, mouse) until you figure out the molecular pathway.” Lindquist’s lab has received funding from the National Institutes of Health, Bethesda, Maryland, to use their yeast system to screen for Aβ toxicity modifiers in the NIH’s ChemIDplus library of 350,000 chemical compounds.—Esther Landhuis.

References:
Wu J, Petralia RS, Kurushima H, Patel H, Jung M, Volk L, Chowdhury S, Shepherd JD, Dehoff M, Li Y, Kuhl D, Huganir RL, Price DL, Scannevin R, Troncoso JC, Wong PC, Worley PF. Arc/Arg3.1 Regulates an Endosomal Pathway Essential for Activity-Dependent Beta-Amyloid Generation. 28 Oct 2011;147:615-628. Abstract

Treusch S, Hamamichi S, Goodman JL, Matlack KE, Chung CY, Baru V, Shulman JM, Parrado A, Bevis BJ, Valastyan JS, Han H, Lindhagen-Persson M, Reiman EM, Evans DA, Bennett DA, Olofsson A, DeJager PL, Tanzi RE, Caldwell KA, Caldwell GA, Lindquist S. Functional Links Between Abeta Toxicity, Endocytic Trafficking, and Alzheimer’s Disease Risk Factors in Yeast. Science. 27 Oct 2011. Abstract