Figure 3. SIRT6 binds DNA with no intermediates.

(A–B) Gel retardation assay of 32 P-5' end-labeled single-strand DNAs and sticky ended dsDNAs as a function of increasing concentrations of SIRT6-His (ssDNA, Kd = 1.48 ± 0.52; sticky dsDNA, Kd = 3.59 ± 0.17). (C) Suggested model of SIRT6 binding to ssDNA as a monomer or open ssDNA ends of dsDNA as a dimer. (D) Ability of SIRT6-Flag to bind to the DNA of circular, blunt-ended and sticky-ended cleaved plasmids. The bar chart depicts averages for three replicate experiments (error bars show SEM), after logarithmic transformation. (E) SIRT6-Flag DNA-binding ability for an open-ended +plasmid +/- NAD. Data are averages from four experiments, with error bars representing SEM (after logarithmic transformation). (F, G) Dimerization of SIRT6 at the LacO site, represented by the recruitment of SIRT6-Cherry by SIRT6-LacR-GFP (n = 181, p<0.005) or GFP-LacR (n = 104). Data are averages from four experiments, with error bars representing SEM.

Figure 3—figure supplement 1. SIRT6 binds DNA with no intermediates.

(A) Gel retardation assay of 32 P-5′ end-labeled dsDNAs with blunt ends as a function of increasing concentrations of SIRT6-His. (B) Gel retardation assay of 32 P-5′ end-labeled RNAs as a function of increasing concentrations of SIRT6-His. Kd could not have been calculated with a 95% confidence interval (CI). (C) DNA-binding ability of SIRT6-Flag, MRE11-Flag and NBS1-Flag for circular and linear plasmids, assessed by a DNA-binding assay. Data are means for 3–9 experiments (with error bars representing SEMs) after logarithmic transformation. (D) SIRT6-Flag DNA-binding ability for plasmids with a 3′ over-hang and a 5′ over-hang assessed by a DNA-binding assay. Log of averages of three experiments (with error bars representing SEMs). (E) Relative amount of NAD+ remaining after consumption by SIRT6 in the presence of DNA alone or with an acetylated peptide (H3K56ac). (F, G) Protection of sticky-ended and blunt-ended plasmids by BSA, SIRT6 and MRE11 against Exonuclease 1 in different time points. Data in panel (G) are means (with error bars showing +/- SEM) from three experiments. (H) Statistical analysis of the end protection: amounts of uncleaved DNA remaining in the presence of SIRT6 or MRE11 compared to BSA (as a control) after 0, 10 or 20 min (*, p <0.05; **, p <0.005; ***, p <0.0005).

Figure 3—figure supplement 2. SIRT6 binds DNA with no intermediates.

(A) SEC-MALS analysis of SIRT6 or of SIRT6 with a dsDNA 10-bp oligo with 3-bp overhanging ends on both sides. For SIRT6: Peak1 protein mass (calculated by UV) = 66.7 ± 3.3 kDa, protein mass (calculated by RI) = 70.7 ± 3.5 kDa; Peak2 protein mass (calculated by UV) = 114.3 ± 5.7 kDa, protein mass (calculated by RI) = 103.8 ± 4.9 kDa; Peak3 protein mass (calculated by UV) = 609.7 ± 14.6 kDa, protein mass (calculated by RI) = 833.6 ± 16.4 kDa. For SIRT6 +DNA: Peak1 protein mass (calculated by UV) = 52.5 ± 3.1 kDa, protein mass (calculated by RI) = 55.7 ± 3.2 kDa; Peak2 protein mass (calculated by UV) = 99.4 ± 3.8 kDa, protein mass (calculated by RI) = 94.6 ± 3.5 kDa; Peak3 protein mass (calculated by UV) = extremely high, protein mass (calculated by RI) = 9224 ± 258 kDa. (B) X-ray scattering profile (right) and the distance distribution function (left) of SIRT6 (blue) and SIRT6 bound to dsDNA (red). (C) Overall parameters for small angle X-ray scattering of SIRT6 alone and of SIRT6 bound to dsDNA determined from the distance distribution function P(r). Rg is the radius of gyration, and Dmax is the maximum dimension of the particle. (D) SAXS structure of SIRT6 (grey surface). Ab initio models were reconstructed from SAXS data using the computer program DAMMIN (Svergun, 1999) and were averaged by the computer program DAMAVER (Volkov and Svergun, 2003). The crystal structure of SIRT6 tetramer (grey spheres) was extracted from the crystal structure of SIRT6 (pdb id: 3PKI) and compared with the obtained SAXS model in PyMOL (http://www.pymol.org).