Viewpoint Cite This: ACS Chem. Neurosci. XXXX, XXX, XXX−XXX
pubs.acs.org/chemneuro
Alzheimer’s Disease, Dendritic Spines, and Calcineurin Inhibitors: A New Approach? Melissa A. O’Neal, Nancy R. Stallings, and James S. Malter* Department of Pathology, O’Donnell Brain Institute, UT Southwestern Medical Center, Dallas, Texas 75390-9072, United States ABSTRACT: Therapeutics to effectively treat Alzheimer’s disease (AD) are lacking. In vitro, animal and human studies have implicated the excessive activation of the protein phosphatase calcineurin (CN) as an early step in the pathogenesis of AD. We discuss recent data showing that the prolyl isomerase Pin1 is suppressed by CN-mediated dephosphorylation induced by Aβ42 signaling. Pin1 loss directly leads to the reductions in dendritic spines and synapses characteristic of early AD pathology. Pin1 activity, and synapse and dendritic spine numbers are rescued by FK506, a highly specific and United States Food and Drug Administration approved CN inhibitor. Solid organ transplant recipients chronically treated with FK506 showed much lower AD incidence than expected. As such, we suggest prospective clinical trials to determine if systemic FK506 can normalize CN activity in the brain, preserve Pin1 function and support synaptic health in early AD. KEYWORDS: Alzheimer’s disease, calcineurin, dendritic spines, Pin1, FK506
W
abundantly expressed in postsynaptic dendritic spines and plays an important role in the maintenance and plasticity of spines, and the acquisition of learning, memory, and long-term potentiation (LTP), among other functions. A growing body of experimental work in cell culture as well as AD-model mice have shown that CN is activated in neurons through signaling induced by soluble, multimeric Aβ42. Blockade of CN with FK506 (tacrolimus) restored LTP, normalized behavior in ADmodel mice, and prevented synapse loss associated with Aβ42 overproduction or mutant tau overexpression.3 Solid organ transplant recipients are chronically immunosuppressed with FK506 (among other drugs) to prevent graft rejection, providing a unique research population to assess AD prevalence in humans on this class of drug. A large, single-center retrospective study of patients on chronic FK506 showed a much lower incidence of Alzheimer’s disease than in an agematched control population.4 While this work requires confirmation, the data exists at many transplant centers to do so. Interestingly, immunosuppression with the mTOR inhibitor rapamycin had no effect on AD development,4 further implicating CN as a critical signaling node. We have recently identified a downstream target of CN signaling involved in the maintenance of dendritic spines.5 These data demonstrated that Pin1, a peptidyl-prolyl isomerase with unique specificity for phosphorylated Ser-Pro or Thr-Pro dipeptides, was absolutely required to maintain mature dendritic spines. Pin1 has previously been implicated in AD due to its regulation of amyloid precursor protein (APP) cleavage as well as hyperphosphorylation of tau. Pin1 is one of many proteins that are lost in developing AD, suggesting a potential positive role in synaptic health and dendritic spine numbers. Unexpectedly, germ-line Pin1 KO mice show the opposite phenotype with increased spine density and enhanced
hile modest synapse loss is a common result of normal, cognitive aging, diffuse, large-scale dendritic spine and synapse loss is a pathologic and potentially catastrophic event in Alzheimer’s disease (AD), Parkinson’s disease, dementia with lewy bodies, Fragile-X syndrome, amyotrophic lateral sclerosis (ALS), Huntington’s disease, and multiple sclerosis.1,2 In AD, spine and synapse loss precedes the appearance of Aβ42 plaques, phospho-tau tangles, and neurodegeneration. As dendritic spines are considered the repository of memory, the disappearance of these critical structures likely accounts for the memory, learning, and behavioral deficits that are characteristic of early AD or so-called Mild Cognitive Impairment (MCI). As spine loss can accelerate neuronal degeneration, therapeutics aimed at synapse maintenance may potentially reduce cell death as well. Therapies for patients with neurodegenerative disorders generally do not modulate the primary pathology although Carbidopa/L-dopa are a critical exception. Currently, United States Food and Drug Administration (FDA) approved AD therapeutics target cholinergic (Aricep) and/or glutamatergic signaling (Memantine) and can improve memory and daily, cognitive function. Over the past 15 years, drugs/antibodies targeting Aβ42 production (β and γ secretase inhibitors) or accumulation (infusion of anti-Aβ42 antibodies or de novo production after vaccination with Aβ42) have been developed and tested. While Aβ42 levels have been positively affected, at least in subjects with familial AD, there have been dangerous side effects and/or insignificant clinical improvement. The repeated failure of antibody/immunization therapeutics has raised the possibility that reducing Aβ42 may not alter AD pathology. While the broad targeting of pathways (β and γ secretase inhibitors) with a multiplicity of effects beyond Aβ42 production was always problematic, the failure of immunebased approaches may reflect trial subjects with more significant and advanced disease than suspected. Clearly new approaches are needed in the fight against AD. One enzyme that may play a role in AD is calcineurin (CN), a protein phosphatase widely expressed in the brain. It is © XXXX American Chemical Society
Received: May 1, 2018 Accepted: May 4, 2018
A
DOI: 10.1021/acschemneuro.8b00213 ACS Chem. Neurosci. XXXX, XXX, XXX−XXX
Viewpoint
ACS Chemical Neuroscience
research has potentially identified a new approach to therapy for early stage neurodegenerative diseases.
long-term potentiation (LTP). Therefore, we determined spine phenotype after postnatal ablation of Pin1 in neurons and adult animals. Six days after injection of AAV-Cre into the hippocampus of Pin1fl/fl mice there was a consistent 35−40% reduction of dendritic spines. Genetic loss of Pin1 from neuronal cultures analyzed at DIV21−24 or wild type (WT) neurons treated briefly with Pin1 inhibitors showed similar spine losses as seen in Pin1 floxed mice in vivo. Notably, treatment of WT neurons with soluble, multimeric Aβ42 for 1 h also caused reproducibly significant, mature spine losses whose magnitude was equivalent to that seen in Pin1 floxed neurons. The transduction of exogenous Pin1 normalized dendritic spine numbers in KO neurons or WT cells treated with Aβ42, demonstrating Pin1 was downstream of Aβ42 signaling and both necessary and sufficient for healthy spines. These results suggested that Aβ42 signaling acutely affected Pin1 isomerase activity in WT cells, leading to the observed spine losses. In order to model these events, we employed synaptoneurosomes, a highly enriched preparation of ∼1 μm membrane encased vesicles, containing connected and functional pre- and post- synaptic structures. Soluble, multimeric Aβ42 rapidly (minutes) activated CN and completely inhibited Pin1 isomerase activity that was entirely blocked by FK506. FK506 also prevented Aβ42 mediated dendritic spine loss in cultured cells but only did so in Pin1 WT cells. These results suggested Aβ42 signaling led to CN activation with subsequent Pin1 blockade and dendritic spine loss. Mass spectrometry analysis of Pin1 immuno-precipitates (IPs) showed interactions with CN that were confirmed by IPs/immunoblot. Pin1 IPs analyzed for post-translational modifications revealed Aβ42 signaling induced, dephosphorylation of Pin1 at multiple sites including Ser111, a location not previously identified. Ser111 was particularly interesting due to its proximity to the active site Cys113. Pin1-S111 mutated to Ala was inactive and functioned as a dominant-negative in both synaptoneurosomes and neuronal cultures while phospho-mimetic Glu mutants were constitutively active, fully rescued spine numbers and antagonized Aβ42mediated signaling. Therefore, we conclude that Aβ42 signaling activates calcineurin which directly dephosphorylates Pin1, rendering it inactive. In the absence of Pin1 activity (as after acute or chronic exposure to Aβ42), APP processing favors the β−γ secretase pathway, enhancing the production of Aβ42 and creating a feed-forward loop. While premature, these and prior results suggest CN should be considered as a target to treat early AD, especially given the availability of FK506, an FDA approved drug given to hundreds to thousands of transplant recipients over the past 25 years. Clearly, repurposing an FDA-approved compound could save years and millions of dollars compared to bringing a new agent to market. In order to move toward that goal, additional retrospective analysis of AD incidence in solid organ transplant populations and ideally prospective trials in early AD patients or those likely to develop AD are needed. Critical questions include can a subimmunosuppressive dose of FK506 protect against soluble Aβ42 signaling? What are the dosing requirements for maintaining dendritic spines and what degree of CN suppression is optimal? Will spine health protect against later neurodegeneration? Will drug therapy be lifelong or could it be stopped once a critical threshold has been passed? Obviously, there are many questions remaining, but given the lack of other options to treat this devastating disease, an open mind to new possibilities seems warranted. We are excited our and other’s
■
AUTHOR INFORMATION
Corresponding Author
*E-mail:
[email protected]. ORCID
Nancy R. Stallings: 0000-0001-6132-1586 Notes
The authors declare no competing financial interest.
■
REFERENCES
(1) Henstridge, C. M., Pickett, E., and Spires-Jones, T. L. (2016) Synaptic pathology: A shared mechanism in neurological disease. Ageing Res. Rev. 28, 72−84. (2) Herms, J., and Dorostkar, M. M. (2016) Dendritic Spine Pathology in Neurodegenerative Diseases. Annu. Rev. Pathol.: Mech. Dis. 11, 221−250. (3) Shah, S. Z., Hussain, T., Zhao, D., and Yang, L. (2017) A central role for calcineurin in protein misfolding neurodegenerative diseases. Cell. Mol. Life Sci. 74, 1061−1074. (4) Taglialatela, G., Rastellini, C., and Cicalese, L. (2015) Reduced Incidence of Dementia in Solid Organ Transplant Patients Treated with Calcineurin Inhibitors. J. Alzheimer's Dis. 47, 329−333. (5) Stallings, N. R., et al. (2018) Pin1 mediates Aβ42-induced dendritic spine loss. Sci. Signaling 11, eaap8734.
B
DOI: 10.1021/acschemneuro.8b00213 ACS Chem. Neurosci. XXXX, XXX, XXX−XXX
