A paper published January 29 in Cell Reports claims that among the earliest changes in sporadic Alzheimer’s disease (SAD) are genetic changes that lead to premature neurogenesis. Scientists led by Bruce Yankner, Harvard Medical School, found that neural progenitor cells derived from fibroblasts of patients with late-onset AD have nuclei bereft of the repressor element 1-silencing transcription factor. REST normally reins in neuronal differentiation. The result? Unchecked differentiation and depletion of the neural progenitor pool.

  • Neural progenitors from AD patients lack the nuclear repressor REST.
  • Without it, neural differentiation accelerates, prematurely depleting the stem cell pool.
  • ApoE4 alone mimics this phenotype.

What’s more, the same phenotype arises in progenitor cells engineered to express the AD risk allele ApoE4. The authors saw abnormalities in the nuclear membranes of these differentiated cells, implying that its disruption is also part of early AD. Together, the results point to a new set of cellular changes in sporadic AD pathogenesis.

“This suggests that the changes to the genome may be happening at an early time point,” said Bess Frost, University of Texas Health, San Antonio, who was not involved in the study. “I’m excited because it provides even more evidence that there’s this problem with overall genomic architecture in the context of Alzheimer’s disease.”

Yankner had previously implicated REST in aging, when his lab reported that REST is highly expressed in aging and correlates with better cognitive function even in the face of AD pathology (Mar 2014 news). The transcriptional repressor turns down expression of genes for cell death and Alzheimer’s pathology, and turns up genes that protect against oxidative damage. Conversely, postmortem neuronal nuclei from AD dementia patients lacked REST, and the protein was down by half in people who had died with mild cognitive impairment.  

“Now in sporadic Alzheimer’s disease neurons [derived from iPSCs, Yankner] was able to recapitulate this, which really validates his previous observations,” said Li-Huei Tsai, Massachusetts Institute of Technology, Cambridge. Tsai is a co-author on the paper, though she contributed mostly technical support. “The pathways underlying this phenotype in this cell culture system may be very fundamental to the disease mechanism and REST could be contributing, though I wouldn’t be surprised if there were other factors involved in this process of accelerated neurodifferentiation and reduced progenitor cell renewal,” she said.

Restoring the Brakes. Neural progenitors (green) from sporadic AD patients (left) lack REST in the nucleus (blue) and hurtle toward differentiation into neurons (red). Overexpressing REST in these cells reins in differentiation (right). [Courtesy of Meyer et al., 2019.]

Co-first authors Katharina Meyer and Heather Feldman went looking for fundamental cellular changes in SAD by examining induced pluripotent stem cells (iPSCs) from patients. Past studies on iPSCs from AD patients primarily focused on mutations that cause early onset disease (Muratore et al., 2014; Sproul et al., 2014Yagi et al., 2011). While some studies have included iPSCs from sporadic patients, those haven’t settled on a common phenotype or central mechanism (Israel et al., 2012; Kondo et al., 2013). 

For this study, researchers collected dermal fibroblasts of five people with sporadic AD and six age-matched controls, reverted the cells to a pluripotent state, and then derived neural progenitors. This type of cell gives rise to most of the differentiated cells of the human brain, including astrocytes, oligodendrocytes, and mature neurons.

Right away, the researchers saw differences between the transcription profiles of neural progenitors from SAD and controls. The former expressed more genes related to neurogenesis and neuronal differentiation. Their tau expression was upregulated, and ApoE expression was down. A greater proportion of SAD cells differentiated into neurons; fewer proliferated to keep the pool of neural progenitors going.

Were there differences among mature neurons? To find out, the authors differentiated the neuronal progenitors into neurons for six weeks. During that time, SAD cells matured faster, some in less than half the time of control cells. Action potentials appeared in the SAD cells weeks earlier than in control cells, occurring twice as often and reaching higher amplitudes, suggesting they were hyperexcitable. SAD cells expressed more genes for synapse formation, axon guidance, neurotransmitter receptors, and ion channels.

What drives these expression differences? A search of the ENCODE chromatin immunoprecipitation (ChIP) database revealed that this particular suite of genes is normally governed by the transcriptional repressor REST, a kind of brake on their expression. Without REST, these genes go into hyper-expression mode. The gene analysis also turned up two components of the polycomb repressive complex 2 (PRC2), EZH2 and SUZ12, which interact with REST to help suppress expression.

REST bound less to its target genes in SAD cells. While these contained as much REST mRNA and protein as control cells, the normally nuclear protein seemed stuck in the cytoplasm. However, when Meyer overexpressed REST in SAD progenitor cells, they differentiated less and expressed normal levels of REST target genes.

Given that ApoE4 is the biggest genetic risk factor for AD, the authors wondered whether its expression would bring on this phenotype in iPSCs derived from people without AD. The authors used CRISPR-Cas9 to convert a homozygous ApoE3 iPSC control line to a homozygous ApoE4 one. This drove up expression of neural differentiation genes and of tau in neural progenitors. As in the SAD cells, REST strayed from the nucleus into the cytoplasm. Conversely, editing an ApoE4 iPSC line to express only ApoE3 led to lower expression of those genes. Together, the results suggest that neural progenitors in people with SAD are pushed to differentiate early, and ApoE4 alone can bring on this phenotype.

“In adults, if you have accelerated differentiation of neural progenitors, you’ll eventually exhaust the stem cell pool,” Yankner explained. “By the time someone is aging, we would hypothesize that their ability to generative new neurons would be diminished.” A previous study suggested that accelerated differentiation disrupts memory in mice (Akers et al., 2014).  “It could be that these neurons don’t function as well,” said Yankner.

If REST is wandering out into the cytoplasm, could there be a problem with nucleocytoplasmic transport? Under a confocal microscope, the authors observed that SAD cells took on an unusual shape, with folds and circular structures within the nuclear lamina (see image below). These abnormal indentations resemble ones recently linked to tau pathology, for example in iPSC-derived neurons from frontotemporal dementia patients, and in AD postmortem brain (Jan 2019 news; Frost et al., 2016Sep 2018 news). Scientists also reported abnormal nucleocytoplasmic transport in transgenic mouse models of AD. No one has previously seen the circular structures Yankner observed in the current study. He plans to do three-dimensional imaging of these nuclei to visualize them more fully.

Weird Invaginations. Nuclei from control neural progenitors (top) appear round, their nuclear lamina smooth (green). In SAD NPs (bottom), circular structures (left) and invaginations (right) appear within the lamina. [Image courtesy of Meyer et al., 2019]

“There is this converging set of observations in the area of neurodegenerative diseases that there’s a fundamental problem with the nuclear membrane/lamina, and the regulation of nucleocytoplasmic transport, which may contribute to the neurodegenerative process,” Yankner told Alzforum.

“These are interesting results,” said Robert Rissman, University of California, San Diego. “REST is an important protein and regulator of many different genes. I think this is a potential mechanism in AD neuropathology.” Rissman previously reported reduced REST in neuronal exosomes found in the blood of people progressing from mild cognitive impairment to AD dementia (Winston et al., 2016). It would be worth checking if REST levels vary in exosomes released from the iPSCs in Yankner’s study, given that they reflect an earlier stage of disease, he added.—Gwyneth Dickey Zakaib


  1. This new report from the Yankner lab provides convincing evidence that a dysregulation of neuronal gene transcription networks underlies events in brain that lead eventually to Alzheimer’s disease. Their work, looking not only at neurons derived from sporadic AD cases but also neurons genetically edited to be homozygous for the e4 allele of the major risk gene APOE, strongly points to the involvement of the transcription factor Repressor element 1-silencing transcription (REST). REST is known as a master negative regulator of neurogenesis and is also a key regulator of the aging brain’s stress response. Meyer at al., find that the nuclear translocation of REST and its ability to bind chromatin are disrupted in sporadic AD and APOE e4 neurons.

    In an earlier paper from the group (Lu et al., 2014), REST levels were shown to be reduced in the AD brain. Importantly, in individuals who exhibited AD pathology, nuclear REST was significantly elevated in cognitively preserved, as compared to cognitively impaired. This suggests that dysregulation of brain’s stress response in the presence of AD pathology causes cognitive decline rather than AD pathology alone. This has therapeutic relevance for AD and other aging-associated neurodegenerative diseases. This finding prompted us to investigate if REST protein could act as a readily accessible peripheral biomarker for AD. We have demonstrated evidence that REST may be reduced in AD and risk factors for AD (Ashton et al., 2017). 

    The question then is, why? Why is REST reduced in brain and why is it excluded from the nucleus where it performs its role as a repressive transcription factor? In the 2014 paper, REST levels are regulated by Wnt signaling and could be reduced by the Wnt antagonist Dickkopf-1 (Dkk1). The authors also showed that activating Wnt promoted the nuclear translocation of REST. This speaks to our work led by Dr. Richard Killick on the role of Dickkopf-1 in AD. It is plausible that the reduced levels of REST and its exclusion from the nucleus are the result of effects on Wnt signaling pathways through the ability of Aβ to induce Dkk1. Killick also described (October 2018 news) how Aβ synaptotoxicity drives further Aβ production via effects on Wnt signaling, again through the upregulation of Dkk1, which could be broken by the ROCK kinase inhibitor, fasudil. It would be interesting to now determine if fasudil might also reverse the effects on REST reported by the Yankner lab, i.e., its exclusion from the nucleus and downstream effects on gene transcription, and also its reduced levels in AD brain.

Make a Comment

To make a comment you must login or register.


News Citations

  1. No REST for Weary Neurons: Protective Factor Stems Cognitive Decline
  2. Invasion of the Microtubules: Mutant Tau Deforms Neuronal Nuclei
  3. Tau Stymies Transport Through Neuron’s Nucleus

Paper Citations

  1. . The familial Alzheimer's disease APPV717I mutation alters APP processing and Tau expression in iPSC-derived neurons. Hum Mol Genet. 2014 Jul 1;23(13):3523-36. Epub 2014 Feb 12 PubMed.
  2. . Characterization and molecular profiling of PSEN1 familial Alzheimer's disease iPSC-derived neural progenitors. PLoS One. 2014;9(1):e84547. Epub 2014 Jan 8 PubMed.
  3. . Modeling familial Alzheimer's disease with induced pluripotent stem cells. Hum Mol Genet. 2011 Dec 1;20(23):4530-9. PubMed.
  4. . Probing sporadic and familial Alzheimer's disease using induced pluripotent stem cells. Nature. 2012 Jan 25;482(7384):216-20. PubMed.
  5. . Modeling Alzheimer's disease with iPSCs reveals stress phenotypes associated with intracellular Aβ and differential drug responsiveness. Cell Stem Cell. 2013 Apr 4;12(4):487-96. Epub 2013 Feb 21 PubMed.
  6. . Hippocampal neurogenesis regulates forgetting during adulthood and infancy. Science. 2014 May 9;344(6184):598-602. PubMed.
  7. . Lamin Dysfunction Mediates Neurodegeneration in Tauopathies. Curr Biol. 2016 Jan 11;26(1):129-36. Epub 2015 Dec 24 PubMed.
  8. . Prediction of conversion from mild cognitive impairment to dementia with neuronally derived blood exosome protein profile. Alzheimers Dement (Amst). 2016;3:63-72. Epub 2016 May 7 PubMed.

Further Reading

Primary Papers

  1. . REST and Neural Gene Network Dysregulation in iPSC Models of Alzheimer's Disease. Cell Rep. 2019 Jan 29;26(5):1112-1127.e9. PubMed.