Sirt5 deacylation activities show differential sensitivities to nicotinamide inhibition

Abstract

Sirtuins are protein deacylases regulating metabolism and aging processes, and the seven human isoforms are considered attractive therapeutic targets. Sirtuins transfer acyl groups from lysine sidechains to ADP-ribose, formed from the cosubstrate NAD(+) by release of nicotinamide, which in turn is assumed to be a general Sirtuin inhibitor. Studies on Sirtuin regulation have been hampered, however, by shortcomings of available assays. Here, we describe a mass spectrometry-based, quantitative deacylation assay not requiring any substrate labeling. Using this assay, we show that the deacetylation activity of human Sirt5 features an unusual insensitivity to nicotinamide inhibition. In contrast, we find similar values for Sirt5 and Sirt3 for the intrinsic NAD(+) affinity as well as the apparent NAD(+) affinity in presence of peptide. Structure comparison and mutagenesis identify an Arg neighboring to the Sirt5 nicotinamide binding pocket as a mediator of nicotinamide resistance, and statistical sequence analyses along with testing further Sirtuins reveal a network of coevolved residues likely defining a nicotinamide-insensitive Sirtuin deacetylase family. The same Arg was recently reported to render Sirt5 a preferential desuccinylase, and we find that this Sirt5 activity is highly sensitive to nicotinamide inhibition. Analysis of Sirt5 structures and activity data suggest that an Arg/succinate interaction is the molecular basis of the differential nicotinamide sensitivities of the two Sirt5 activities. Our results thus indicate a Sirtuin subfamily with nicotinamide-insensitive deacetylase activity and suggest that the molecular features determining nicotinamide sensitivity overlap with those dominating deacylation specificity, possibly suggesting that other subfamily members might also prefer other acylations than acetylations.

Figure 1. Development of a label-free, quantitative mass spectrometry-based deacylation assay.

(A) Different amounts of acetylated (○) and deacetylated (•) CPS1-Lys527 peptide are plotted against their respective mass spectrometry signal areas. Interpolations (lines) show the linear correlations between peptide amounts and detected signals, and the slightly different slopes for the two peptide species. (B) Ratios of the injected amounts of deacetylated and acetylated peptide plotted against ratios of the measured log₁₀ signal areas (•). Equation and correlation for the linear interpolation (line) are indicated. (C) Scheme for the mass spectrometry-based deacylation assay. Percent deacetylation is calculated by normalizing the product area to the total signal area, and deacetylation rates are determined through analysis of aliquots taken after different incubation times.

Figure 2. The mass spectrometry-based deacetylation assay reveals an unusual low Sirt5 sensitivity for nicotinamide inhibition.

(A) Dose-dependent nicotinamide inhibition of the deacetylation of an ACS2 or CPS1 peptide, respectively, by Sirt3 and Sirt5. % activity was determined through relative quantification of reaction product by mass spectrometry and normalization to the respective non inhibited activity (set to 100%). Error bars represent standard errors for three independent measurements. (B) Dissociation constants for the interaction of NAD⁺ with Sirt3 and Sirt5, respectively, in presence of different nicotinamide concentrations. K_d values were determined by microscale thermophoresis measurements. Error bars represent standard errors of three independent measurements. NAM, nicotinamide.

Figure 3. Structure comparison and identification of a Sirtuin sequence motive indicating nicotinamide insensitive deacetylation activity.

(A) Structure comparison of Sirt3 (red) and Sirt5 (blue), overlaid with a Sir2Tm/nicotinamide complex (grey). Nicotinamide and coupled residues (see below) in Sirt3 and Sirt5 are displayed as sticks and colored by atom type for Sirt5 and for nicotinamide. (B) Statistical coupling analysis scores (lower panel) identify residues apparently coevolving (score cutoff used: 1.5) with Sirt5-Arg105: Thr69 and Tyr102 (highlighted in panel A), Trp77, Arg217, and Trp222. Amino acids in other Sirtuin sequences and the corresponding Sirtuin class are shown on top of the scores. (C) Inhibition Sirt5-Arg105Leu by nicotinamide. Activities, determined using our mass spectrometry assay, were normalized against activity in absence of nicotinamide. Data for wildtype Sirt3 and Sirt5 from Fig. 2A are shown here for comparison. (D) Comparison of IC₅₀ values, determined using the mass spectrometry assay, for nicotinamide inhibition of class I and III Sirtuins and of the Sirt5-Arg105Leu variant. Error bars represent the standard error of the fit. NAM, nicotinamide.

Figure 4. Nicotinamide effects on Sirt5 activity depend on the type of acyl group.

(A) Comparison of Sirt5 activities against acetylated and succinylated CPS1-Lys527 peptide, determinded by mass spectrometry and normalized using the desuccinylation activity. Error bars represent standard errors of linear fits to time-series experiments. (B) Nicotinamide-dependent inhibition of Sirt5 activity against acetylated and succinylated CPS1-Lys527 peptide, determined using mass spectrometry and normalized against activity in absence of nicotinamide. The Sirt3/ACS2 and Sirt5/CPS1 curves from Fig. 2A were added for comparison. (C) Cartoon representation of the Sirt5/succinylated substrate/NAD⁺ complex structure. The ligands and Arg105 are shown as sticks colored according to atom type (ligands) or in blue (Arg105). A modeled alternative sidechain conformation for Arg105 (rotamer 2) is shown in red. CPS1succ, succinylated CPS1-Lys527 peptide; NAM, nicotinamide.

Figure 5. Role of Sir2Af2-Met70 in deacylation specificity and nicotinamide sensitivity.

(A) Substrate concentration dependent activity of Sir2Af2 against acetylated (•) and succinylated (Δ) Prx1-Lys197 peptides determined using the mass spectrometry assay. Error bars represent standard errors of linear fits to time-series measurements. (B) Comparison of Sir2Af2 activities against acetylated and succinylated Prx1-Lys197 peptide, determined using mass spectrometry and normalized against wt deacetylation activity. Error bars represent standard errors of linear fits to time-series experiments. (C+D) Nicotinamide-dependent inhibition of Sir2Af2 wt (•) and Met70Arg variant (○) activity against (C) succinylated and (D) acetylated Prx1-Lys197 peptide, determined using mass spectrometry and normalized against activity in absence of nicotinamide. The effect of exchanging Met70 to Arg is indicated by arrows. Prx1succ, succinylated Prx1-Lys197 peptide; NAM, nicotinamide.