Researchers in Junying Yuan’s laboratory at Harvard Medical School report that they have identified a small molecule that interferes with a protein phosphatase in such a way that it enables cultured cells to withstand endoplasmic reticulum stress brought on by misfolded proteins. The study will appear in tomorrow’s issue of Science.

The work is not directly about Alzheimer disease, and the compound itself is not a drug candidate (see author Q&A below). However, it could open up an avenue toward cell-protective agents that target the endoplasmic reticulum. In that sense, it offers a new tack to current efforts in academia and industry to find neuroprotective agents that target shared pathways of cellular distress rather than the specific cause of a single disease. These efforts represent a new wave of research after earlier attempts to develop antiapoptotic drugs had failed.

The study is notable from a broader drug-discovery perspective, as well, because phosphatase enzymes have proven notoriously difficult to exploit as drug targets. Their protein structure is such that they are fairly promiscuous in their choice of substrate. Consequently, most available inhibitors either are not selective, i.e., they inhibit all different substrates from which a given phosphatase removes a phosphate group, or they are not specific, i.e., they inhibit many different phosphatases all at once. Therefore, phosphatase inhibitors tend to be too toxic for use in humans. By contrast, salubrinal was non-toxic at the doses where its cell protection activity reaches its peak, Boyce and colleagues report. The reason for this may be that the compound inhibits only the PP1 phosphatase, and only about four of its substrates.

ER stress arises in viral infections, and also when misfolded proteins accumulate in the ER, as happens in many neurodegenerative diseases. The cell tries to save itself by activating the unfolded protein response (UPR), a set of pathways that turns down overall protein synthesis while turning up the production of a few particular proteins. When this response becomes stymied, persistent ER stress leads to apoptosis. First author Michael Boyce, working with Yuan and colleagues at Harvard and three other institutions in the U.S. and China, studied this phenomenon by screening compound libraries for small molecules that protect rat PC12 cells from this kind of cell death. They called their best candidate salubrinal and discovered that it was not a general apoptosis inhibitor but was specific to insults that stress the ER.

Then Boyce and colleagues studied where in the varied pathways of the UPR salubrinal struck. They found that it left alone the UPR branches that change gene transcription but affected only post-transcriptional modifications. There, it kept active a mediator known to decrease global translation and increase translation of stress-induced mRNAs. Its name is a mouthful: Eukaryotic translation initiation factor 2 subunit α (eIF2α). Specifically, salubrinal induced the phosphorylation (i.e., activation) of eIF2α and did so, surprisingly, not by activating one of its four known kinases, but by inhibiting its requisite phosphatase. Exactly how salubrinal does this remains uncertain, but the authors report that it appears to interfere with the protein complex containing the actual phosphatase PP1 and its cofactor GADD34.

When further experiments revealed that salubrinal was both more specific and more selective than previous phosphatase inhibitors, Boyce and colleagues turned to herpes simplex infection for a first-pass evaluation of its potential in disease. They report that their compound inhibits both eIF2α dephosphorylation mediated by the virus and viral replication. Salubrinal also reduced the viral titer in eye swabs of a mouse HSV cornea infection model.

This study shows that “selective small-molecule inhibition of eIF2α dephosphorylation effectively protects cells from ER stress,” the authors write. The neurodegeneration community could pick up from there and test if this compound has any effect in some of the existing cell-based and animal model systems.—Gabrielle Strobel

Q&A with Junying Yuan.

Q: These days, active compounds tend to get checked for their commercial potential before they are openly published with their chemical structure, as salubrinal is. Could salubrinal be a candidate therapeutic molecule? If no, why not?
A: We don't propose that salubrinal is a drug candidate (or even a lead molecule), and we haven't looked into any of the parameters that good drugs need to have, such as its pharmacokinetic or pharmacodynamic properties in animals. However, we think that since salubrinal can affect ER stress-related and virus-related dephosphorylation pathways in a novel way, it provides an exciting proof-of-principle that future drugs might be able to take advantage of these pathways, as well. Salubrinal itself will probably be most useful as a research tool.

Q: An EC50 of 15 micromolar is generally considered not good enough for drug development. Are you working with follow-up compounds that bring this number into the nanomolar range?
A: Our collaborators at the Shanghai Institute of Organic Chemistry have synthesized a number of derivates of salubrinal that we've tested for things like stronger protection from apoptosis or lower EC50 values. So far, we haven't seen a dramatic improvement along those lines (with about 60 compounds tested). It's worth noting, though, that rescuing a cell from intense ER stress caused by strong poisons like tunicamycin is very difficult, so salubrinal is unique in this regard, even though its EC50 isn't as low as we might like.

Q: Many neurodegenerative diseases are thought to be protein conformation/misfolding diseases. Would cytoprotective compounds that alleviate the downstream effects of protein misfolding be general treatments for several diseases at once, or would they have to be specific to the given misfolding protein, i.e., Ht, ataxin, Aβ, α-synuclein, tau?
A: That was certainly part of the original rationale for starting our project, and many other labs have published evidence suggesting that correcting protein misfolding might be a productive therapeutic strategy in diverse neurodegenerative disorders. Hopefully, compounds like salubrinal that manipulate protein folding and processing pathways will contribute to the development of future drugs targeting these processes.

Q: There is ongoing research on ER stress and the UPR in Alzheimer research, but no clear picture yet on where it fits into AD pathogenesis, especially in vivo. Where would salubrinal fit in?
A: We haven't looked into whether salubrinal might be useful in Alzheimer's model systems yet, but it would certainly be interesting and easy to test. There are some unpublished data for a potentially neuroprotective role of salubrinal from our collaborator’s lab, but we are not in a position to reveal them.

Q: Are any of the pathways you experimented with—eIF2α, PP1/GADD34,CReP—active in stressed neurons? Are they expressed in adult brain?
A: All these proteins are expressed in the adult brain, and many studies in the literature have indicated that ER stress pathways (including eIF2α phosphorylation) occurs in response to brain injury, such as acute ischemia. What's not yet clear is whether the induction of these pathways is a protective or destructive response to the injury, but we hope reagents like salubrinal might help answer this question.

Q: What are the next experiments to define the link to age-related neurodegeneration?
A: It would be certainly interesting to look into the possibility if Aβ-induced neuronal cell death can be protected by salubrinal.

Q: You describe initial results whereby topical salubrinal treatment reduced virus titer in eye swabs of eight mice whose corneas were infected with HSV. Your paper mentions potential uses of drugs like salubrinal in neurodegeneration in general terms. Can you expand on that?
A: As you mentioned before, protein misfolding and/or aggregation occurs in many neurodegenerative disorders and can disrupt multiple cellular organelles and processes, including the ER and secretory system. Knowing that, we hope that compounds like salubrinal could be used to manipulate these pathways and perhaps protect cells from the toxic effects of protein misfolding. Even though we don't consider salubrinal a drug candidate, we think its novel mode of action might point in a useful direction for new drugs to be developed.

Q: Have you tried salubrinal in any assays with a neurodegeneration readout?
A: Not yet.

Q: Does it cross the blood-brain barrier? Could it be infused into AD/PD/HT models? Applied to slice cultures?
A: We haven't checked any of these things yet.

Q: Just how selective and safe is salubrinal? Your paper shows it does inhibit several PP1 substrates other than eIF2α, after all.
A: We think it's likely that salubrinal affects other dephosphorylation events in the cell besides eIF2α, but probably not many. It seems clear that it doesn't affect all or even most PP1-mediated dephosphorylations. (Evidence for that is presented in the paper.) Also, the compound seems non-toxic to most cells at concentrations significantly higher than what's needed for cytoprotection, which at least suggests that inhibiting eIF2α dephosphorylation to a useful extent isn't necessarily toxic. One mystery we'd love to solve is to figure out exactly how, mechanistically, salubrinal is able to inhibit only a small subset of eIF2α dephosphorylations, but we don't know that yet.


  1. This is a very interesting paper. Boyce and colleagues showed that pharmacological inhibition of dephosphorylation of eukaryotic initiator factor 2α increases its activity, thus protecting against the effects of ER stress. They also demonstrated that this effect slows down HSV replication.

    However, these important findings do not seem to have a direct application in Alzheimer disease. ER stress, and the consequent UPR, are not implicated in β amyloid production, or in APP processing, as shown by my and other's groups (Siman et al., 2001; Piccini et al., 2004). Instead, inhibition of the effects of ER stress may be potentially beneficial in neurodegenerative disorders characterized by intracellular toxic aggregates, such as Parkinson disease and tauopathies, in which ER stress may contribute to the creation of misfolded peptides.


    . Endoplasmic reticulum stress-induced cysteine protease activation in cortical neurons: effect of an Alzheimer's disease-linked presenilin-1 knock-in mutation. J Biol Chem. 2001 Nov 30;276(48):44736-43. PubMed.

    . Fibroblasts from FAD-linked presenilin 1 mutations display a normal unfolded protein response but overproduce Abeta42 in response to tunicamycin. Neurobiol Dis. 2004 Mar;15(2):380-6. PubMed.

Make a Comment

To make a comment you must login or register.


Other Citations

  1. see author Q&A below

Further Reading


  1. . ER stress and the unfolded protein response. Mutat Res. 2005 Jan 6;569(1-2):29-63. PubMed.
  2. . Valproate protects cells from ER stress-induced lipid accumulation and apoptosis by inhibiting glycogen synthase kinase-3. J Cell Sci. 2005 Jan 1;118(Pt 1):89-99. PubMed.
  3. . Intracellularly generated amyloid-beta peptide counteracts the antiapoptotic function of its precursor protein and primes proapoptotic pathways for activation by other insults in neuroblastoma cells. J Neurochem. 2004 Dec;91(6):1260-74. PubMed.
  4. . [Neurodegeneration caused by ER stress?--the pathogenetic mechanisms underlying AR-JP]. Nihon Yakurigaku Zasshi. 2004 Dec;124(6):375-82. PubMed.
  5. . Heat shock protein 70 participates in the neuroprotective response to intracellularly expressed beta-amyloid in neurons. J Neurosci. 2004 Feb 18;24(7):1700-6. PubMed.
  6. . Degradation of misfolded proteins prevents ER-derived oxidative stress and cell death. Mol Cell. 2004 Sep 10;15(5):767-76. PubMed.

Primary Papers

  1. . A selective inhibitor of eIF2alpha dephosphorylation protects cells from ER stress. Science. 2005 Feb 11;307(5711):935-9. PubMed.