When intracellular trash disposals overflow with misfolded proteins, cells may dump the excess onto their neighbors’ doorsteps. In the June 13 Nature Cell Biology, researchers introduced a secretion pathway that spews misfolded cytoplasmic proteins into the extracellular abyss by way of a cellular stress response. A chaperone called USP19 escorts the waste into a newly described export lane, particularly when the proteasome becomes overwhelmed. In a first hint toward relevance to neurodegeneration, the researchers, led by Yihong Ye of the National Institutes of Health in Bethesda, Maryland, reported that the chaperone-turned-bouncer dispatched α-synuclein in this manner. It is not known if neurons do this, too, yet the potential contribution of this secretion system to the spread of amyloidogenic proteins excited researchers.
“The findings add a new pathway to the list of potential therapeutic targets for neurodegenerative disease,” commented Lary Walker of Emory University in Atlanta, who was not involved in the work. “The study reaffirms that cells can respond to adversity in unexpected ways, and so it seems likely that more unconventional proteostatic pathways await discovery.” (See full comment below.)
Cells expend a remarkable amount of energy disposing of misfolded proteins that threaten to gum up the works. The proteasome, which chomps up ubiquitinated proteins in the cytoplasm, is the best-known protein quality control system, while lysosomal degradation, autophagy, and chaperone-based unfolding and refolding also keep the cell tidy (see Wolff et al., 2014; Buchberger et al., 2010). Some misfolded proteins, including tau, reportedly leave the cell via tiny vesicular packages called exosomes; however, the researchers commented that export via exosomes tends to favor normal, folded proteins (see Saman et al., 2011; Lo Cicero et al., 2015).
Ye’s group was in the midst of investigating the role of USP19, the only deubiquitinating enzyme located in the ER membrane, in yet another protein quality-control pathway called ER-associated degradation (ERAD), when they stumbled across the new secretion system. In cells overexpressing the deubiquitinase, the researchers noticed an abundance of GFP, which was expressed as a cytosolic protein, floating in the cell medium. “This made me wonder how a cytosolic protein could be secreted,” Ye said.
First author Jin-Gu Lee and colleagues determined that HEK293T cells overexpressing USP19 secreted GFP in a manner dependent on USP19’s catalytic activity and localization to the ER membrane. GFP did not leave the cells because transfection made the cells leaky, as USP19 overexpression did not increase cellular permeability. Normal ER cargo such as clusterin, which get ferried through the ER to the Golgi and then the plasma membrane and contain signal sequences, were unaffected by USP19 overexpression, and blocking the normal ER secretion pathway with Brefeldin A did not affect USP19-mediated GFP secretion. These findings suggested to the researchers that the deubiquitinase was involved in an unconventional secretion pathway, which somehow targeted proteins in the cytoplasm rather than in the ER.
The researchers also noticed that the GFP secreted into the medium, which they detected using antibodies, had lost its fluorescence. Since fluorescence depends on the tight packaging of a fluorophore within the barrel-like structure of the protein, the researchers wondered if GFP had somehow lost its normal shape and that was why the cell dumped it. To investigate, they expressed a non-functional version of GFP (GFP1-10) that tends to misfold. The HEK cells secreted this wonky GFP even more than normal GFP, and this also depended upon USP19 catalytic activity.
What about proteins other than GFP? Out of a panel of highly expressed proteins, the researchers found that USP19 mediated the secretion of ubiquitin-like protein 4A (Ubl4A), but only when Ubl4A was overexpressed. Normally Ubl4A exists in a protein complex, so the researchers reasoned that the extra copies of lone Ubl4A were somehow targeted by the secretion pathway. At this point, they dubbed the new pathway “misfolding-associated protein secretion” (MAPS), owing to its preference for misfolded or unassembled proteins.
What about proteins implicated in neurodegenerative disease? Most of this paper—a massive 12-page study plus seven supplementary figures and videos—used the workhorse tools of GFP and HEK cells to nail down the basic biology of MAPS. In a first foray into neurodegeneration, the researchers found that α-synuclein, but not tau, was secreted in a USP19-dependent fashion. Several Parkinson’s disease- associated mutants of α-synuclein also hitched a ride out of the cell via MAPS. When the researchers knocked down USP19 using short interfering RNA, or wiped it out completely with CRISPR, they saw a dramatic drop in secretion of GFP1-10, Ubl4A, and α-synuclein from cells overexpressing these proteins.
To determine the relationship between MAPS and the proteasomal degradation pathway, the researchers treated cells with the proteasome inhibitor MG-132. They found a build-up of GFP1-10 and Ubl4A in the cell, suggesting that these proteins are normally degraded by the proteasome. In cells overexpressing USP19, proteasome inhibition diverted more GFP1-10 and Ubl4A into the MAPS pathway. Interestingly, when the researchers compared the ubiquitination state of these proteins inside and outside of the cell, they found that the secreted versions were lacking ubiquitin adornments, while those that remained in the cell were ubiquitinated. This suggested that USP19 removes ubiquitin as a way to target proteins into MAPS. Finally, the researchers found that the backup secretion system made cells hardier: A higher percentage of USP19-expressing cells survived transient exposure to the proteasome inhibitor than cells lacking USP19 expression. This indicated that the MAPS pathway provides critical relief during times when the proteasome cannot keep up with the cell’s demand for degradation.
How did USP19 mediate the secretion of MAPS substrates? Because USP19 is anchored to the cytosolic side of the ER membrane via a transmembrane domain at its C-terminus, the researchers investigated whether MAPS substrates also ended up there. They found that USP19 overexpression elevated the level of GFP1-10 in membrane fractions of cells, and confocal microscopy revealed comingling of USP19, GFP1-10, and the ER marker PDI.
All Aboard the MAPS Train. Misfolding-prone GFP1-10 (green) mingles with USP19 (red) at the ER membrane. The localization patterns of both proteins trace similar paths (bottom panels). [Courtesy of Lee et al., Nature Cell Biology 2016.]
After the ER, where to next? Live cell imaging revealed ER-associated vesicular bubbles periodically moving away from the ER domain. These so-called “MAPS vesicles” were mobile and expressed Rab9, much like late endosomes. Importantly, once released from the cell, the contents of these MAPS vesicles was free, rather than encapsulated in persistent vesicles, like exosomes, as the secreted GFP1-10 did not sediment via ultracentrifugation the way exosomes do.
Ye compared the MAPS model to a train. “Imagine the ER as the platform of a train station, whereas Rab9-positive late endosomes are cars on a train transporting passengers to the cell exterior,” he said. “USP19 is a chaperone that brings passengers to the platform.”
The idea that misfolded, toxic proteins spread between neurons in a prion-like fashion in neurodegenerative diseases has gathered steam, though it has also been called into question because mechanisms of transmission or toxicity still have to be proven (see Apr 2016 webinar).
“By identifying a previously unknown process by which cells secrete misfolded cytosolic proteins into the extracellular space, Lee and colleagues have taken a compelling step toward meeting this challenge,” Walker told Alzforum. Patrik Brundin of the Van Andel Research Institute in Grand Rapids, Michigan, agreed. “These are very exciting findings that potentially will impact research into novel therapeutic targets for Parkinson’s disease and related α-synucleinopathies,” he wrote. “While the authors have mostly focused their studies on non-neuronal cells, the findings are very important and provide insight into a key molecular mechanism of high relevance to Parkinson’s pathogenesis.” (See full comment below.)
Natura Myeku of Columbia University in New York voiced a concern shared by all commentators as well as the researchers themselves: “The study is to be complimented for its significance and quality results, but it remains to be seen if this mechanism is present in cultured primary neurons or in vivo, and under normal expression of USP19.” That α-synuclein, but not tau, is a MAPS substrate even though both proteins are implicated in proteopathic spread highlights the need to study this in the brain, Myeku added (see full comment below). Ye told Alzforum that he wants to collaborate with neuroscience labs to test the importance of MAPS in neurons and in neurodegenerative disease.
Marc Diamond of the University of Texas Southwestern Medical Center in Dallas, who studies tau propagation, commented that his lab may look into MAPS as a mechanism. About tau, he noted that Ye’s study did not investigate whether oligomers could be secreted via MAPS, noting that tau trimers, not misfolded monomers, are putative proteopathic seeds (see Apr 2015 news). Ye speculated that because MAPS substrates must somehow translocate from the cytoplasm into the ER, it seems unlikely that large aggregates could enter the secretion pathway.
How might MAPS secretion compare to protein release in exosomes? Last year, Tsuneya Ikezu’s group at Boston University reported that microglia transmitted tau between neurons via these tiny vesicular packages (see Oct 2015 news). Ikezu told Alzforum that while MAPS could be a bona fide pathway worthy of further investigation, the authors did not definitively rule out exosomal secretion of GFP1-10. While they saw no GFP1-10 in a would-be exosomal fraction, they had no positive control to prove that they could detect exosomes in the first place, Ikezu said. He added that the contribution of MAPS could vary between different cell types and between different isoforms of tau.
All researchers contacted by Alzforum agreed that replication and translation of the findings to neurons was a crucial next step. At the same time, they were enthusiastic about the potential importance of the pathway in neurodegeneration. Walker added that MAPS could potentially flag the earliest stages of neurodegeneration. “The transition from proteasomal degradation to MAPS might leave biochemical signatures that could serve as biomarkers during the long, clinically silent period that precedes obvious signs and symptoms of disease,” he said.—Jessica Shugart
- Tau Triple Threat: Do Trimers Make Bad Seeds?
- Deadly Delivery: Microglia May Traffic Tau Via Exosomes
- Wolff S, Weissman JS, Dillin A. Differential scales of protein quality control. Cell. 2014 Mar 27;157(1):52-64. PubMed.
- Buchberger A, Bukau B, Sommer T. Protein quality control in the cytosol and the endoplasmic reticulum: brothers in arms. Mol Cell. 2010 Oct 22;40(2):238-52. PubMed.
- Saman S, Kim W, Raya M, Visnick Y, Miro S, Jackson B, McKee AC, Alvarez VE, Lee NC, Hall GF. Exosome-associated tau is secreted in tauopathy models and is selectively phosphorylated in cerebrospinal fluid (CSF) in early Alzheimer's Disease. J Biol Chem. 2011 Nov 4; PubMed.
- Lo Cicero A, Stahl PD, Raposo G. Extracellular vesicles shuffling intercellular messages: for good or for bad. Curr Opin Cell Biol. 2015 Aug;35:69-77. Epub 2015 May 19 PubMed.
- Fontaine SN, Zheng D, Sabbagh JJ, Martin MD, Chaput D, Darling A, Trotter JH, Stothert AR, Nordhues BA, Lussier A, Baker J, Shelton L, Kahn M, Blair LJ, Stevens SM Jr, Dickey CA. DnaJ/Hsc70 chaperone complexes control the extracellular release of neurodegenerative-associated proteins. EMBO J. 2016 Jul 15;35(14):1537-49. Epub 2016 Jun 3 PubMed.
- Lee JG, Takahama S, Zhang G, Tomarev SI, Ye Y. Unconventional secretion of misfolded proteins promotes adaptation to proteasome dysfunction in mammalian cells. Nat Cell Biol. 2016 Jun 13; PubMed.