|Home | About | Journals | Submit | Contact Us | Français|
SIRT1 is one of seven mammalian sirtuin (silent information regulator 2-related) proteins that harbor NAD+-dependent protein deacetylase activity and is implicated in multiple metabolic and age-associated pathways and disorders. The sirtuin proteins contain a central region of high sequence conservation that is required for catalytic activity, but more variable N- and C-terminal regions have been proposed to mediate protein specific activities. Here we show that the conserved catalytic core domain of SIRT1 has very low catalytic activity toward several known protein substrates, but that regions N- and C-terminal to the catalytic core potentiate catalytic efficiency by between 12- and 45-fold, with the N-terminal domain contributing predominantly to catalytic rate, relatively independent of the nature of the acetyl-lysine protein substrate, and the C-terminal domain contributing significantly to the Km for NAD+. We show that the N- and C-terminal regions stimulate SIRT1 deacetylase activity intramolecularly and that the C-terminal region stably associates with the catalytic core domain to form a SIRT1 holoenzyme. We also demonstrate that the C-terminal region of SIRT1 can influence the inhibitory activity of some sirtuin inhibitors that are known to function through the catalytic core domain. Together, these studies highlight the unique properties of the SIRT1 member of the sirtuin proteins and have implications for the development of SIRT1-specific regulatory molecules.
Sirtuins comprise a family of NAD+-dependent protein deacetylase enzymes that catalyze the removal of an acetyl moiety from the ϵ-amino group of lysine residues within protein targets (1, 2) to yield the deacetylated protein product, nicotinamide, and 2′-O-acetyl-ADP-ribose (3, 4). The founding member of this protein family, Saccharomyces cerevisiae Sir2p, one of five yeast sirtuin proteins including HST1–4, is a limiting factor in yeast aging, because deletion of the SIR2 gene results in reduced replicative lifespan (5), and additional copies of SIR2 result in increased yeast replicative lifespan (6). The sirtuin protein family is broadly conserved in all three domains of life (7), and increased sirtuin expression in higher eukaryotes has been reported to lead to increased lifespan in worms (8) and flies (9), and increased longevity caused by a calorie-restricted diet in some of these organisms is sirtuin-dependent (9). Humans have seven sirtuin proteins (SIRT1–7) (10), several of which have been shown to have distinct biological properties (11). The SIRT1 protein has been the most thoroughly studied of the human sirtuins and has been shown to play roles in many biological processes including cell survival, apoptosis, stress resistance, fat storage, insulin production, glucose homeostasis, and lipid homeostasis through direct deacetylation or regulation of its many known in vivo targets including transcription factors such as p53, Forkhead box class O, and peroxisome proliferator-activated receptor γ and histones such as H3 (K9 and K14) and H4 (K16) (reviewed in Ref. 12). SIRT1 has also been implicated to play a role in a number of age-related human diseases including diabetes, cancer, and inflammation (reviewed in Ref. 13).
Sequence alignment of the sirtuin proteins indicates that they contain an ~275-amino acid conserved catalytic core domain (7, 14) (Fig. 1A and supplemental Fig. S1). Consistent with the sequence conservation among the sirtuin proteins, the catalytic core domains of known sirtuin structures show a high degree of structural superposition (supplemental Fig. S1). Several x-ray crystal structures of sirtuin proteins from bacteria (CobB and Sir2Tm), yeast (Sir2 and Hst2), archaea (Sir2-Af1 and Sir2-Af2), and human (SIRT2, SIRT3, and SIRT5) in several different liganded forms reveal a structurally conserved catalytic core domain that adopts an elongated shape containing a large and structurally homologous Rossmann fold domain, characteristic of NAD+/NADH-binding proteins; a more structurally diverse, smaller, zinc-binding domain; and several loops connecting the two domains (reviewed in Ref. 15). These loops form a pronounced, extended cleft between the large and small domains where the NAD+ and acetyl-lysine-containing peptide substrates enter from opposite sides and bind to the enzyme. The amino acids involved in catalysis and the reactive groups of both bound substrate molecules are buried within a protein tunnel in the cleft between the two domains, the region of the enzyme that contains the highest sequence conservation within the sirtuin enzymes (supplemental Fig. S1).
In addition to the sequence and structural conservation of the catalytic core domain, many eukaryotic sirtuin proteins contain N- and C-terminal flanking regions that are variable in length and sequence and have been proposed to play protein-specific regulatory roles (Fig. 1A). For example, the N- and C-terminal segments of yeast Hst2 have been shown to mediate homotrimer formation that plays an autoinhibitory role in the deacetylase activity of Hst2 (16), whereas the N-terminal region was also shown to contain a nuclear export signal (17). Of the seven human sirtuin proteins, the 747-residue SIRT1 protein contains the most extended N- and C-terminal segments that flank a catalytic core domain within approximately residues 240–500, with the function of these N- and C-terminal segments unknown (Fig. 1A and supplemental Fig. S1).
In this study, we tested the hypothesis that, like yeast Hst2, the N- and C-terminal segments of SIRT1 play autoregulatory functions to modulate SIRT1 deacetylase activity. Consistent with this hypothesis, we find that a SIRT1 protein construct harboring the conserved catalytic core domain contains very low deacetylase activity, whereas the addition of N- and C-terminal segments of SIRT1 can potentiate the catalytic efficiency of the core region by between 12–45-fold, with the N-terminal domain contributing predominantly to catalytic rate and the C-terminal domain contributing significantly to the Km for NAD+. We also show that the N- and C-terminal regions stimulate SIRT1 deacetylase activity intramolecularly to form a SIRT1 holoenzyme and that the C-terminal domain can influence the activity of SIRT1 inhibitors.
A plasmid expressing full-length human SIRT1 was a generous gift from David Sinclair (Harvard Medical School). DNA fragments encoding various SIRT1 constructs (residues 214–583, 160–583, 160–665, and 584–665) were cloned into the first multiple cloning site of the pETDuet-1 vector containing the gene for yeast SMT3, a ubiquitin-like protein of the SUMO family. BL21-CodonPlus (DE3 RIL) cells harboring the SIRT1 expression constructs were grown at 37 °C in LB medium and induced at log phase with 0.5 mm isopropyl β-d-thiogalactopyranoside at 18 °C overnight. The cells were harvested and lysed by sonication in a sonication buffer containing 50 mm Tris, pH 7.5, 200 mm NaCl, 5 mm imidazole, 10 mm β-mercaptoethanol and 5% glycerol with the addition of 0.1 mg/ml PMSF. The cell lysate was centrifuged to remove cell debris, and the protein supernatant was loaded onto a nickel-nitrilotriacetic acid-agarose column and washed extensively with sonication buffer supplemented with 30 mm imidazole, before eluting the protein with a 30–600 mm imidazole gradient in sonication buffer. The His-SUMO tag was removed by incubating the SIRT1 protein with ULP1 (ubiquitin-like-specific protease 1) at 4 °C overnight and then passing the overnight incubation solution through a fresh nickel-nitrilotriacetic acid column equilibrated in sonication buffer plus 30 mm imidazole. The NaCl concentration of the collected flow through was lowered to ~100 mm by diluting with sonication buffer containing 50 mm NaCl and loading the solution on a HiTrap Q ion exchange column. SIRT1 protein elution was carried out with a gradient of 100–800 mm NaCl in sonication buffer. Peak SIRT1 protein fractions were further purified on Superdex-200 in a buffer containing 20 mm HEPES, pH 7.5, 100 mm NaCl, 5% glycerol, and 10 mm DTT to yield >90% pure SIRT1 protein (as judged by SDS-PAGE). SIRT1 protein was concentrated to ~10 mg/ml using a microconcentrator and either used immediately or stored at −80 °C after flash freezing prior to use. A 6-liter purification typically yielded 1–3 mg of purified protein depending on the specific SIRT1 protein construct.
The steady state parameters and catalytic efficiency of deacetylase activity of SIRT1 protein constructs were determined using a radioactive assay that employs [14C]NAD+, labeled on the carbonyl group of the nicotinamide moiety. This assay measures the SIRT1-dependent hydrolysis of the [14C]nicotinamide group, which is isolated from unreacted [14C]NAD+ by extraction using ethyl acetate and quantified by scintillation counting as described elsewhere (2). The peptide substrates used in this study were purchased from Genscript and were centered around the SIRT1 targets
H3(4–14)K9Ac (QTARKK(Ac)STGGK), H4(7–25)K16Ac (GKGLGKGGAK(Ac)RHRKVLRDN), and p53(372–389)K382Ac (KKGQSTSRHKK(Ac)LMFKTEG). Radioactive assays were carried out for 15 min at 30 °C (in the linear range of enzyme activity) in a total reaction volume of 100 μl in a buffer containing 50 mm glycine, pH 9.0, 0.5 mm DTT, 5 mm tetrasodium pyrophosphate, and 0.1 mg/ml BSA. The reaction also contained 100 nm protein, a saturating concentration of NAD+ of 2 mm (or near saturating for the CC and N-CC SIRT1 constructs that have relatively high Km values for NAD+) while varying peptide concentrations (0–500 μm) (see Table 1) or saturating peptide substrate (500 μm) with varying NAD+ concentrations (0–800 μm) (see Table 2). The reactions were quenched by adding 67 μl of 0.1 m sodium borate, pH 8.0, into the reaction buffer before the cleaved [14C]nicotinamide product of the reaction was extracted with ethyl acetate and quantified by scintillation counting. Background control reactions were performed in the absence of enzyme. All of the reactions were performed in triplicate, and the data were directly fit to the Michaelis-Menten equation in GraphPad Prism.
IC50 values for SIRT1 inhibitors (Ex-527, suramin and Ro 31-8220) were determined at room temperature in a 100-μl reaction volume in the presence of 100 nm SIRT1 protein, 300 μm [14C]NAD+, 1 mm of H3(4–14)K9Ac peptide substrate, and varying concentrations of the respective inhibitors (76 nm to 167 μm of Ex-527, 6–110 μm of suramin, and 0.1–1000 μm of Ro 31×8220 stored in Me2SO) in the same reaction buffer noted above, and the reaction was quenched and analyzed as noted above. The IC50 value of nicotinamide was determined using the Fluor-de-Lys SIRT1 fluorometric drug discovery assay kit (Enzo Life Sciences) according to the manual. All of the reactions were performed in triplicate, and the data were fit to a sigmoidal curve using GraphPad Prism.
Analytical ultracentrifugation of the SIRT1(160–665) construct was performed at 4 °C using absorbance optics with a Beckman Optima XL-I analytical ultracentrifuge, using a four-hole rotor. The partial specific volume, buffer density, and viscosity were estimated using Sednterp (18). Analysis was performed using six-channel centerpieces with quartz windows, spinning at 18,000, 22,000, and 26,000 rpm and at protein concentrations of 1.3, 0.9, and 0.6 mg/ml in 20 mm HEPES, pH 7.5, 100 mm NaCl, 5% glycerol (v/v), and 5 mm β-mercaptoethanol. A global fit of the data, using the program Heteroanalysis, fit best to a single species SIRT1(160–665) monomer. The quality of the fit was assessed from examination of root mean square deviation.
We successfully prepared recombinant full-length SIRT1 protein in bacteria but found that the protein was highly prone to aggregation and degradation and that its catalytic activity was highly variable (data not shown). To prepare a more stable and catalytically active SIRT1 protein construct, we used a sequence alignment of sirtuin proteins to aid in the design of SIRT1 protein constructs containing the catalytically active region (supplemental Fig. S1). This alignment suggested that the catalytic core domain of SIRT1 mapped to within residues 240–510. The preparation of several different SIRT1 protein constructs containing this region resulted in the identification of three SIRT1 protein constructs that could be expressed to high levels in bacteria and purified to >95% homogeneity and isolated as a monodisperse, nonaggregated species on gel filtration chromatography. These protein constructs were SIRT1(214–583) (herein called SIRT1-CC for containing the catalytic core domain), SIRT1(160–583) (herein called SIRT1-N-CC for containing the catalytic core domain plus an extended N-terminal region), and SIRT1(160–665) (herein called SIRT1-N-CC-C for containing the catalytic core domain plus extended N- and C-terminal regions).
The deacetylase activity of these three SIRT1 protein constructs were analyzed using saturation kinetics to determine steady state parameters (kcat and Km) and catalytic efficiency (kcat/Km). For these initial studies, varying concentrations of a peptide substrate centered around histone H3(4–14)K9Ac was used with saturating levels of NAD+. These studies revealed that the shortest of the SIRT1 protein constructs harboring the catalytic core domain, SIRT1-CC, exhibits steady state parameters of Km = 132.6 μm, kcat = 3.4 min−1 and a catalytic efficiency of 25,900 min−1 m−1 (Fig. 2A). Surprisingly, the SIRT1 protein constructs that contained extended N- and C-terminal segments showed significantly greater activity with SIRT1-N-CC and SIRT1-N-CC-C constructs exhibiting catalytic efficiencies of 313,000 min−1 m−1 and 648,000 min−1 m−1, respectively (Fig. 2). This elevation in catalytic efficiency of ~12-fold for the N-terminally extended protein construct and of ~25-fold for the N- and C-terminally extended construct suggests that regions N- and C-terminal to the conserved catalytic core domain of SIRT1 potentiate deacetylase activity on a histone H3(4–14)K9Ac-containing substrate.
To determine whether the C-terminal region of SIRT1 potentiates the activity of the catalytic core domain by directly binding to the catalytic core domain or through some indirect mechanism, we tested the ability of a SIRT1 fragment containing the isolated extended C-terminal region, SIRT1(584–665) (SIRT1-C) to potentiate the activity of SIRT1-N-CC in trans. As can be seen in Fig. 3 and Tables 1 and and2,2, a 10-fold molar excess of SIRT1-C added to SIRT1-N-CC increases the catalytic efficiency of this domain by ~1.5–5-fold, depending on the nature of the substrate, and very close to the activity of the intact SIRT1-N-CC-C protein construct. Importantly, the addition of a 10-fold molar excess of SIRT1-C to SIRT1-N-CC-C did not further increase the activity of SIRT1-N-CC-C (Fig. 3, A and B). Taken together, these data are consistent with a direct interaction between the catalytic core domain of SIRT1 with the C-terminal segment and possibly also the N-terminal segment to potentiate the catalytic activity of SIRT1.
Because the data described above is consistent with an interaction of the N- and C-terminal segments of SIRT1 with the catalytic core domain to potentiate catalytic activity through the same or possibly another molecule of SIRT1 (for example through dimer formation), we subjected SIRT1-N-CC-C to sedimentation equilibrium studies using analytical ultracentrifugation to establish the oligomerization state of SIRT1. We carried out these experiments at three protein concentrations (1.3, 0.9, and 0.6 mg/ml) and three centrifugation speeds (18,000, 22,000, and 26,000 rpm) (Fig. 3C). A global analysis of this data fit extremely well to a single species monomer model with a molecular mass of 53.3 kDa (actual mass = 57.0 kDa) and root mean square deviation of 0.0105 Abs280. Taken together, these data suggest that the N- and C-terminal segments of SIRT1 potentiate catalytic activity by an intramolecular mechanism.
We were curious to determine whether the SIRT1 catalytic potentiation observed against the H3(4–14)K9Ac peptide substrate might also be observed for other SIRT1 substrates. To address this, we assayed the SIRT1 constructs SIRT1-CC, SIRT1-N-CC, and SIRT1-N-CC-C against peptides containing H4(7–25)K16Ac and p53(372–389)K382Ac, two other known SIRT1 substrates. As can be seen in Fig. 4A, deacetylation of the H4(7–25)K16Ac substrate shows ~6.7- and ~14.4-fold enhancements in catalytic efficiency of the SIRT1-N-CC and SIRT1-N-CC-C constructs relative to the SIRT1-CC construct, respectively. This is comparable with the ~12.1- and 25-fold enhancement that was observed for the corresponding protein constructs against the H3(4–14)K9Ac substrate (Table 1). Similar levels of potentiation by N- and C-terminal regions was also observed on p53(372–389)K382Ac substrate (9.6- and 11.8-fold, respectively). Taken together, these studies reveal that catalytic potentiation by the N- and C-terminal segments that flank the SIRT1 catalytic core domain is relatively independent of the nature of the acetyl-lysine protein substrate and therefore likely does not participate in protein specific acetyl-lysine recognition.
An analysis of the activity of SIRT1 constructs on several acetyl-lysine-containing protein substrates clearly shows that the potentiating effects of the N- and C-terminal segments on SIRT1 catalysis is largely driven by an enhancement of kcat of ~6–8-fold for the N-terminal segment and of another ~2–3-fold for the C-terminal segment, with relative minor effects on Km for acetyl-lysine substrate binding (Table 1). To explore whether the N- and/or C-terminal segments of SIRT1 contributed to the Km for NAD+ catalysis, we carried out similar saturation kinetics under conditions of saturating acetyl-lysine substrate peptide and varying concentrations of NAD+ substrate (Fig. 5). We employed both the H4(7–25)K16Ac and p53(372–389)K382Ac substrates for these studies. We observed that the catalytic efficiency is enhanced ~5.7- and ~44.8-fold for SIRT1-N-CC and SIRT1-N-CC-C relative to SIRT1-CC, respectively, against the H4(7–25)K16Ac substrate and similarly ~4.6- and ~26.7-fold, respectively, against the p53(372–389)K382Ac substrate. Interestingly, this degree of catalytic potentiation is ~3-fold higher than that observed when varying the acetyl-lysine protein substrate (compare Tables 1 and and2).2). A comparison of the steady state parameters reveals that this increase on catalytic potentiation on the NAD+ substrate over the acetyl-lysine protein substrate is largely due to a 3–5-fold increase in the Km for NAD+ of the SIRT1-N-CC construct relative to SIRT1-N-CC-C. This observation suggests that the C-terminal segment of SIRT1 also contributes to NAD+ binding for catalytic potentiation. We note that in the absence of the SIRT1 C-terminal segment, relatively high Km values for NAD+ are obtained for SIRT1-CC and SIRT1-N-CC (489–888 μm) so that 2 mm NAD+ is at a subsaturating concentration for these protein constructs. This suggests that the apparent kcat values obtained for these protein constructs at 2 mm NAD+ (Table 1) are marginally smaller than the actual values (because the enzymes only have 70–80% of their maximal activity), and therefore the reported potentiation by the C-terminal segment reported in Table 1 is a slight overestimate. Taking the data together, the extended N- and C-terminal segments of SIRT1 both contribute to potentiating the catalytic efficiency of the SIRT1 catalytic domain. The N- and C-terminal segments contribute to an elevation in kcat, with the N-terminal segment playing a significant role. In contrast, the C-terminal segment of SIRT1 plays an additional role in lowing the Km for NAD+.
Several sirtuin inhibitors have been described in the literature (reviewed in Ref. 19) including suramin (20) and EX-527 (21), the most potent SIRT1 inhibitor with an IC50 value in the mid nanomolar range. Less potent sirtuin inhibitors, such as Ro 31-8220, an adenosine mimic (22), and nicotinamide, a product and noncompetitive inhibitor of sirtuins (23, 24), have also been described. Although the structure of suramin bound to SIRT5 suggests that the inhibitor binds competitively with both NAD+ and acetyl-lysine to the catalytic domain (25), docking studies with Ex-527 suggests that it binds competitively with only NAD+ (26). Ro 31-8220 and nicotinamide are also proposed to bind parts of the conserved NAD+ binding pocket. We were curious to determine whether the N- and C-terminal segments of SIRT1 could influence the potency of sirtuin inhibitors. To do this, we assayed the four inhibitors suramin, Ex-527, Ro 31-8220, and nicotinamide against the SIRT1-N-CC and SIRT1-N-CC-C protein constructs using the H3(4–14)K9Ac peptide substrate. Because of the low activity of the SIRT1-CC construct, we could not obtain reliable inhibition data using this protein construct. The IC50 curves were generated for all inhibitors revealing that although Ex-527, Ro 31-8220 and nicotinamide inhibited with comparable IC50 values for both the SIRT1-N-CC and SIRT1-N-CC-C protein constructs (Fig. 6), suramin showed more than a 10-fold greater potency against SIRT1-N-CC (IC50 = 0.52 μm) relative to SIRT1-N-CC-C (IC50 = 7.8 μm) (Fig. 6). These results demonstrate that the N- and C-terminal segments of SIRT1 can have an influence on the potency of some sirtuin inhibitors.
Here we show that the SIRT1 protein deacetylase contains nonconserved N- and C-terminal segments that potentiate the catalytic activity of a central conserved catalytic region. The degree of potentiation is relatively independent of the nature of the acetyl-lysine-containing protein substrate, with the N-terminal segment contributing predominantly to the catalytic rate and the C-terminal domain contributing significantly to the Km for NAD+. The contributions of the N- and C-terminal segments of the SIRT1 catalytic core domain is schematized in Fig. 7. We also show that the N- and C-terminal segments interact with the catalytic core domain through an intramolecular mechanism to form a SIRT1 holoenzyme and that they can influence the inhibitory activity of some sirtuin inhibitors that are known to function through the catalytic core domain. The observation that the suramin sirtuin inhibitor has a higher IC50 value for SIRT1-N-CC-C over SIRT1-N-CC (by more than 10-fold) further suggests that the C-terminal segment of SIRT1 might bind along part of the interface that suramin makes with sirtuins (25).
These studies demonstrate that the SIRT1-N-CC and SIRT1-N-CC-C constructs show a range of potentiation between ~12- and 45-fold relative to SIRT1-CC, respectively, which suggests that either the N-terminal region plays a more important role in potentiation or that the N- and C-terminal regions have a cooperative effect on potentiation. We were unfortunately not able to prepare a soluble SIRT1 construct to test between these two possibilities, although the fact that we were not able to prepare a stable SIRT1-CC-C construct does suggest that the C-terminal segment requires the N-terminal segment for protein stability arguing for an interaction between the two domains that might lead to a cooperative potentiation of SIRT1 catalytic activity (Fig. 7).
A sequence alignment of the N- and C-terminal catalytic potentiating regions of SIRT1 orthologs shows a high degree of conservation between residues 188–228 at the N terminus and residues 621–664 in the C terminus (Fig. 1, B and C), implicating these specific regions to be important for catalytic potentiation. Interestingly, secondary structure predictions of these regions suggest that although the N-terminal segment is highly helical, the C-terminal region is largely unstructured. It is possible that the C-terminal segment might adopt structure when bound to the central catalytic core domain and/or the N-terminal segment. In this study, we have shown that the C-terminal segment can potentiate the activity of the SIRT1-N-CC construct in trans. It would be interesting to determine whether isolated peptides containing these conserved segments might function as modulators (activators or inhibitors) of SIRT1 activity. If this was the case, then small molecule mimics of these regions might be particularly selective SIRT1 modulators.
Because of the association of SIRT1 with several human diseases, there has been considerable effort in developing small molecule sirtuin modulators. Toward this end, several compounds have also been identified and characterized as small molecule sirtuin inhibitors. In addition to suramin (27), Ro 31-8220 (22), and Ex-527 and its analogs (28), the compounds sirtinol (29), splitomycin (30), cambinol (31), tenovin (32), and surfactin (33) have also been described as sirtuin inhibitors. It is not clear where most of these compounds bind to the sirtuin enzymes or how they exert their inhibitory effect, although the observation that several of these inhibitors have different potencies against different members of the sirtuin family suggest that they do not exclusively target the conserved catalytic core domain. There have also been reports of sirtuin activators including resveratrol (27) and a family of compounds unrelated to resveratrol (34), although several other reports suggest that the SIRT1 activation observed for these compounds is an artifact of the assay method used (28, 35–37). The studies presented here suggest new avenues for the development of SIRT1-specific modulators that take advantage of the SIRT1-specific N- and C-terminal catalytic potentiating segments. In addition to preparing molecules that might inhibit the ability of the N- and C-terminal segments to potentiate catalytic activity, it may be possible to design molecules to promote this potentiation to serve as SIRT1-specific activating compounds. For example, molecules that disrupt the interaction of the C-terminal SIRT1(584–665) segment with the conserved catalytic core domain might inhibit SIRT1 activity, and molecules that promote this interaction might serve as SIRT1 activators.
Other sirtuin proteins, such as yeast Hst2 have also been shown to employ nonconserved N- and C-terminal segments to modulate catalytic activity (16). It is therefore possible that the N- and C-terminal segments of other sirtuins might play a particularly important and specifying role in modulating sirtuin-specific function and an understanding of the molecular basis for this in other sirtuin proteins might lead to other sirtuin-specific modulators that might have therapeutic applications.
We acknowledge the use of the Wistar Proteomics Core facility for the work reported here.
*This work was supported, in whole or in part, by National Institutes of Health Grants GM060293 and CA107107 (to R. M.) and CA010815 (to the Wistar Proteomics Core facility).
This article contains supplemental Fig. S1.