Slipping your car into reverse while barreling down the highway is hardly good practice. Neither is simultaneously stomping on the gas pedal and the brake. But when it comes to driving the cell cycle, that might just be what presenilins, the catalytic components of γ-secretase, can do to cells. In the 22 October Journal of Neuroscience, researchers led by Bruce Lamb at Ohio’s Cleveland Clinic report that oligomers of Aβ, released by γ-secretase activity, drive cell cycle re-entry. And in the 29 October issue of the journal, researchers led by Hui Zheng at Baylor College of Medicine in Houston, Texas, report a γ-secretase-independent function for presenilin-1 (PS1) that puts the cycle on hold. Whether familial AD mutations in PS1 compromise that braking effect is unclear. “The take-home message from both groups is that neuronal cell cycle re-entry is an early event and can be triggered by multiple factors,” suggested Zheng. She also emphasized that while PS1-related cell cycle events may make neurons more vulnerable, they are not sufficient to induce neurodegeneration. “We tend to agree with that,” said Lamb. “There is some sort of second hit, or factor, that pushes neurons to degenerate. What those second hits might be is not clear.”
Being post-mitotic cells, neurons are not supposed to be in the business of cell division. Yet there is evidence to suggest that in AD and other neurodegenerative diseases, the neuronal cell cycle machinery is active (see related ARF live discussion). Whether this is a cause or effect of the ongoing pathology is not known, but indications are that aberrant cell cycling is an early event in the disease (see Yang et al., 2003 and ARF related news story). To investigate what drives cycling attempts in AD, Lamb and colleagues examined neuronal cell cycle re-entry in the B6-R1.40 transgenic model of AD, a yeast artificial chromosome-based model that expresses the Swedish amyloid-β precursor protein (APP) mutation driven by human promoters. In this mouse strain, cell cycle markers show up long before any signs of Aβ deposition, giving the researchers an opportunity to test if soluble oligomeric forms of Aβ may be a driving force behind aberrant cycling.
First author Nicholas Varvel analyzed mouse brain of varying ages and stages of Aβ deposition for markers such as cyclin A, cyclin D, and proliferating cell nuclear antigen. Consistent with previous results, the researchers found signs of DNA replication and cell cycle event (CCE) in layers II/III of the frontal cortex in six-month-old, but not four-month-old mice. About 45 percent of cells in this cortical region exhibited CCE markers at six months and about 11 percent were polyploid, indicating their DNA had replicated. By 12 months CCE markers had spread to cortical layers V/VI, where again about 50 percent of cells were CCE positive and 11 percent polyploid. The extra six months did not lead to increased CCE expression or polyploidy in layer II/III cells, however.
To correlate these cell cycle changes with Aβ production, Varvel and colleagues repeated the same measurements in the D2-R1.40 strain, which develops Aβ deposits later, at around 24 months. In these animals, CCE markers were also delayed. There was no sign of cell cycle re-entry at six months, and by 12 months only cortical layers II/III tested positive for CCE markers. “These studies demonstrate that genetic reduction in steady-state levels of Aβ delays the appearance of neuronal CCEs in all brain regions in the D2-R1.40 model,” write the authors. Further evidence for this conclusion came by crossing B6-R1.40 mice with BACE1 knockout mice. The crosses showed no CCE markers in cortical layers II/III at six months. “This suggests that at the very least CCE events are related to amyloidogenic processing, and probably some sort of Aβ species” said Lamb. “And since the deposits do not occur until later we postulated that it may be some sort of soluble Aβ species.”
To test this, the researchers turned to in-vitro studies. Varvel challenged primary neuronal cultures with either preparations of monomers and small polymers (trimers, and tetramers as judged by Western analysis) or oligomers (ranging from 25 to 98 kDa) with Aβ. Exposure to increasing concentrations of only the latter preparation led to increasing incorporation of the DNA marker BrdU, indicating DNA replication and cell cycle re-entry. The NU-2 monoclonal antibody, which binds to Aβ oligomers, prevented this effect (see Lambert et al., 2007). The findings suggest that soluble Aβ oligomers can induce cell cycle events long before Aβ deposits are formed, in keeping with a growing consensus that these soluble intermediates are an early toxic form of Aβ.
Zheng and colleagues came to their conclusions from a different angle; that is, they rescued presenilin double knockout mice. Ablating both PS1 and PS2 causes profound neuronal loss. To learn whether this is related to γ-secretase activity, first author Verena Kallhoff-Munoz and colleagues used various PS mutants to rescue conditional double knockouts (PS cDKO), whose PS2 is deleted on top of a conditional PS1-null background. As expected, wild-type PS completely rescued the neuronal loss phenotype, but curiously, a γ-secretase-deficient mutant (D257A) partially did, too, suggesting that other domains in PS1 contribute to neuronal viability. One candidate is the hydrophilic loop encoded by exon 10. Work from Eddie Koo’s lab at the University of California, San Diego, has linked this loop to cell cycle re-entry prevention in non-neuronal cells (see Soriano et al., 2001; Kang et al., 2002). If the loop plays a similar role in neurons, then losing PS1 may lift this block and contribute to neuronal loss via the cell cycle, the researchers reasoned.
In support of this idea, Kallhoff-Munoz and colleagues found aberrant CCE marker expression in the PS cDKO brain. This was rescued not only by wild-type PS1 but also by the D257A γ-secretase-deficient mutant. “That result simply says that regulation of the neuron cell cycle by presenilin is γ-secretase independent,” said Zheng. However, the PS1 ΔE10 mutant, which lacks the hydrophilic loop domain, had no effect on CCE markers, indicating that it is the PS hydrophilic loop that prevents cell cycle re-entry in neurons.
But does cell cycle re-entry explain the rampant neuronal loss seen in double PS knockouts? Apparently not, since the D257A mutant could rescue cell cycle marker phenotypes but not neurodegeneration. “This also tells us that activating the neuronal cell cycle alone does not lead to neurodegeneration,” said Zheng. The researchers confirmed this with a PS ΔE10 mutant knock-in strain, which had readily detectable CCE markers but no overt neuronal loss. “Those two findings simply state that presenilins regulate the cell cycle, and that this is γ-secretase independent and not a direct cause of neurodegeneration,” Zheng said. That is consistent with Lamb’s findings. “In APP transgenics, they saw a great deal of cell cycle events, but no neurodegeneration,” Zheng added.
Several questions remain. For example, how does Aβ lead to cell cycle re-entry? For that, common pathways affected by oligomers need more study. Even working that out leaves open the question of how this connects to neurodegeneration. Zheng and colleagues propose that cell cycle activation increases a neuron’s vulnerability to stress by other challenges. For example, they showed that PS1 ΔE10 mutant cells are more susceptible to hydrogen peroxide; they also found that this mutation leads to expression of p53 in the cytoplasm and that oxidative stress makes p53 move to the nucleus, where it can induce apoptosis.
The leap from these mouse and cell-based data to the AD brain has not been made, either, though in the case of familial AD (FAD) the connection may be more obvious. Recent evidence suggests that loss-of-function mutations in PS1 that cause FAD also cause loss of cell cycle control (see Malik et al., 2008).—Tom Fagan
- Cell Cycle Hypothesis Pedaling into Mainstream Acceptance? Results in Fly, Mouse Models Warrant a Second Look
- Yang Y, Mufson EJ, Herrup K. Neuronal cell death is preceded by cell cycle events at all stages of Alzheimer's disease. J Neurosci. 2003 Apr 1;23(7):2557-63. PubMed.
- Lambert MP, Velasco PT, Chang L, Viola KL, Fernandez S, Lacor PN, Khuon D, Gong Y, Bigio EH, Shaw P, De Felice FG, Krafft GA, Klein WL. Monoclonal antibodies that target pathological assemblies of Abeta. J Neurochem. 2007 Jan;100(1):23-35. PubMed.
- Soriano S, Kang DE, Fu M, Pestell R, Chevallier N, Zheng H, Koo EH. Presenilin 1 negatively regulates beta-catenin/T cell factor/lymphoid enhancer factor-1 signaling independently of beta-amyloid precursor protein and notch processing. J Cell Biol. 2001 Feb 19;152(4):785-94. PubMed.
- Kang DE, Soriano S, Xia X, Eberhart CG, De Strooper B, Zheng H, Koo EH. Presenilin couples the paired phosphorylation of beta-catenin independent of axin: implications for beta-catenin activation in tumorigenesis. Cell. 2002 Sep 20;110(6):751-62. PubMed.
- Malik B, Currais A, Andres A, Towlson C, Pitsi D, Nunes A, Niblock M, Cooper J, Hortobágyi T, Soriano S. Loss of neuronal cell cycle control as a mechanism of neurodegeneration in the presenilin-1 Alzheimer's disease brain. Cell Cycle. 2008 Mar 1;7(5):637-46. PubMed.
- Kallhoff-Munoz V, Hu L, Chen X, Pautler RG, Zheng H. Genetic dissection of gamma-secretase-dependent and -independent functions of presenilin in regulating neuronal cell cycle and cell death. J Neurosci. 2008 Oct 29;28(44):11421-31. PubMed.
- Varvel NH, Bhaskar K, Patil AR, Pimplikar SW, Herrup K, Lamb BT. Abeta oligomers induce neuronal cell cycle events in Alzheimer's disease. J Neurosci. 2008 Oct 22;28(43):10786-93. PubMed.