Date Published: April 11, 2019
Publisher: Public Library of Science
Author(s): Jason A. Pfister, Chi Ma, Santosh R. D’Mello, Jaya Padmanabhan.
SIRT1, a NAD+-dependent deacetylase, protects neurons in a variety of in vitro and in vivo models of neurodegenerative disease. We have previously described a neuroprotective effect by SIRT1 independent of its catalytic activity. To confirm this conclusion we tested a panel of SIRT1 deletion mutant constructs, designated Δ1–Δ10, in cerebellar granule neurons induced to undergo apoptosis by low potassium treatment. We find that deletions of its N-terminal, those lacking portions of the catalytic domain, as well as one that lacks the ESA (Essential for SIRT1 Activity) motif, are as protective as wild-type SIRT1. In contrast, deletion of the region spanning residues 542–609, construct Δ8, substantially reduced the neuroprotective activity of SIRT1. As observed with LK-induced apoptosis, all SIRT1 constructs except Δ8 protect neurons against mutant huntingtin toxicity. Although its own catalytic activity is not required, neuroprotection by SIRT1 is abolished by inhibitors of Class I HDACs as well as by knockdown of endogenous HDAC1. We find that SIRT1 interacts with HDAC1 and this interaction is greatly increased by deleting regions of SIRT1 necessary for its catalytic activity. However, SIRT1-mediated protection is not dependent on HDAC1 deacetylase activity. Although other studies have described that catalytic activity of SIRT1 mediates is neuroprotective effect, our study suggests that in cerebellar granule neurons its deacetylase activity is not important and that HDAC1 contributes to the neuroprotective effect of SIRT1.
Histone deacetylases (HDACs) are the catalytic subunits of multiprotein complexes that deacetylate specific lysines in the tail residues of histones, resulting in the compaction of chromatin into a transcriptionally repressed state (reviewed in [1, 2]). Although best studied for their effects on histones and transcriptional activity, it is now known that they regulate the acetylation status of a number of other non-histone proteins, suggesting complex functions for HDACs [1, 2]. Mammals express eighteen HDACs that make up two distinct groups, classical HDACs (HDAC1–11) that are Zn-dependent, and the sirtuins (SIRT1–7), which require NAD+ for activation. It is well established that classical HDACs regulate neurodegeneration with some members of the family promoting neurodegeneration whereas others protect against it [3, 4]. Similarly, sirtuins have also been described to have both protective and neurotoxic effects [5–8].
It is generally believed that the many cellular functions of SIRT1 are mediated through its NAD+-dependent catalytic activity. But whether SIRT1 can function independent of its catalytic activity has not been rigorously examined. One well-described activity of SIRT1 is its ability to protect neurons both in cell culture and animal models of neurodegeneration, particularly when overexpressed [7, 9–11, 19]. We previously described that two separate point-mutant forms of SIRT1 lacking in catalytic activity, H355A and H363Y, were just as protective as wild-type SIRT1 in CGNs induced to undergo apoptosis by withdrawal of depolarizing stimuli . Additionally, neither of two structurally distinct pharmacological SIRT1 inhibitors reduced the ability of ectopically expressed SIRT1 to protect neurons . To rigorously confirm that SIRT1 could indeed protect independently of its catalytic activity, we have now analyzed additional mutants lacking relatively large portions of the catalytic domain as well as regions outside of it. Consistent with our previous data using point-mutants, the SIRT1 deletion mutants lacking portions of the catalytic domains were also neuroprotective (Fig 1B). The most important region for SIRT1’s deacetylase activity is the ESA motif mapped to residues 631–655 within SIRT1’s C-terminal, which interacts with and functions as an “on switch” for the deacetylase core . We find that expression of the Δ9 construct, which lacks the ESA, still protects neurons (Fig 1B). In addition to the C-terminal, other studies have described that the N-terminal of SIRT1 is necessary for its catalytic activity [36–39]. Within the N-terminal lie two motifs, Motif A (residues 1–52) and Motif B (residues 163–221). These motifs operate independently of one another and may function through association with each other in trans leading to SIRT1 dimerization, or in cis to enhance substrate binding and SIRT1 catalytic activity as a monomer . Motif B also contains a region that is targeted by small molecule activators of SIRT1 catalytic activity [52, 53]. We previously reported that a SIRT1 construct lacking its first 81 amino acids, and thus Motif A, was just as protective as the full-length protein . Here we show that deletion of either Motif A or B with constructs Δ1 and Δ3, respectively, did not interfere with protection (Fig 1B). Collectively, along with our previous study , our results now conclusively demonstrate that SIRT1 can protect neurons against apoptosis through a mechanism that is independent of its catalytic activity. A similar conclusion was reached by Singh et al. who described that a catalytically-dead point-mutant form, of SIRT1, H363Y, was protective in a cell culture model of Parkinson’s disease . Other studies using other non-neuronal cell types have described that some other properties and functions of SIRT1 are also independent of its catalytic activity [23–25].