Research ArticleBiochemistry

Crystal structure of the hinge domain of Smchd1 reveals its dimerization mode and nucleic acid–binding residues

See allHide authors and affiliations

Science Signaling  16 Jun 2020:
Vol. 13, Issue 636, eaaz5599
DOI: 10.1126/scisignal.aaz5599
  • Fig. 1 The overall structure of the hinge domain of SMCHD1 reveals differences compared to canonical SMC proteins.

    (A and B) Schematic diagrams of SMCHD1 (A) and the canonical SMC (B) protein domain architecture. SMCHD1 and canonical SMC proteins dimerize through their hinge domain, which is composed of subdomain I (slate and cyan) and subdomain II (green). The canonical SMC dimer has long intramolecular antiparallel coiled coil arms that are composed of N- (blue) and C-terminal (pink) residues flanking the hinge domain, and the ABC-type ATPase domain constituted of N- and C-terminal subunits. The Smchd1 dimer contains shorter intermolecular coiled-coils that are connected to its N-terminal region (gray) encapsulating a GHKL-type ATPase domain (red). (C and D) The crystal structure of the Smchd1 hinge domain (C) and a canonical SMC hinge [PDB: 1GXK;(D)] as shown from above the torus in cartoon representation. The component protomers are demarked by a dashed border. The color scheme is the same as in Fig. 1A but with protomer 1 in purple (subdomain I) and orange (subdomain II) for contrast. (E and F) Diagrams depicting the topology of the Smchd1 monomer [residues 1710 to 1884; (E)] and the canonical SMC hinge domain from T. maritima (F). (G and H) Cartoons comparing the hinge domain monomers of Smchd1 (G) and T. maritima (H). (I) Superimposition of the Smchd1 and TmSMC hinge domains [from (G) and (H), respectively]. Secondary structure is annotated sequentially for Smchd1 and according to the original TmSMC description (19); however, short helical segments (consisting of less than two hydrogen bonds) have been omitted from the renderings.

  • Fig. 2 Multiple sequence alignment of the hinge domains of the Smchd1 hinge domain and canonical SMC proteins.

    Protein sequences of Mus musculus (Mm) Smchd1 and canonical SMC proteins—including MmSMC1 and SMC3, Homo sapiens (Hs) SMC2 and SMC4, and T. maritima (Tm) SMC—were aligned using the program MultAlin (50). α Helices (H) and β strands (S) are colored in red and yellow, respectively, the 310 helices are shaded in purple, and the loop connecting subdomains I and II is colored in blue. The secondary structures of mouse Smchd1 and TmSMC as defined by DALI are annotated above and below the aligned sequences, respectively. The image was generated using ESPript3.0 (51). Gray text indicates regions of the SMCHD1 structure that could not be built because of unresolved electron density.

  • Fig. 3 Dimerization of the Smchd1 hinge domain occurs via canonical and noncanonical interfaces.

    (A) Magnified view of the dimerization interface of the Smchd1 hinge domain as highlighted with subdomain I in slate and subdomain II in green and the side view at 90° as indicated. Key residues coordinating the dimer interface are shown as sticks with numbers indicated and highlighted according to their subdomain. Electrostatic interactions between residues are indicated by dashed lines. (B to G) Small-angle x-ray analysis of WT [B; replotted from our earlier report; (9)] and the representative dimer interface mutants Y1765A (C), V1774G (D), R1848A (E), K1873A (F), and G1872A/G1875A/G1876A (G). Top: Scattering intensity profiles are shown where the background-subtracted SAXS data are indicated as black circles representing mean intensity I(q) as a function of momentum transfer q in Å−1. Guinier plots for qRg ≤ 1.3 are shown as insets, where linearity indicates that high–molecular weight and polydisperse particles do not measurably contribute to the scattering. The radius of gyration and initial scattering intensity I(0) were approximated using the Guinier equation with PRIMUS (43), giving values as indicated in table S1. Bottom: Pair distribution functions, P(r) plots were calculated using GNOM (44). The Rg and maximum particle dimension Dmax calculated from the P(r) analysis are as reported in table S1. Data in (C) to (E) were collected with a PILATUS 1M detector; a PILATUS 2M detector was used to collect data for the G1872A/G1875A/G1876A mutant.

  • Fig. 4 The nucleic acid–binding activity of the Smchd1 hinge homodimer is potentiated by two positively charged interaction sites.

    (A) Electrostatic surface potential representations of the Smchd1 hinge homodimer viewed down the side of the torus. Positively charged residues are shown as sticks in dark blue. Lys1718, Arg1719, and Arg1771 at the top side of the dimer are grouped as cluster 1 (black). Lys1789, Arg1790, Arg1796, and Lys1799 located at the bottom side of the dimer are grouped as cluster 2 (cyan). Residues in close proximity to the pathogenic mutation Arg1867, including Arg1869, Lys1873, and Lys1880, are grouped as cluster 3 (blue). (B) Cartoon representation of the Smchd1 homodimer in the same orientation as that shown in (A) highlighting residues that interact with nucleic acid on one protomer. The thin red line indicates the possible position of nucleic acid binding as deduced from these experiments. (C to G) DNA binding fluorescence polarization (mP) of 12.5 nM 6-FAM–labeled, 20-base pair single-stranded DNA by wild-type (WT) and clusters 1 (C), cluster 2 (D), cluster 3 (E), dimer interface (F), and the torus (G) mutants of the Smchd1 hinge dimer at indicated protein concentrations. Data points are plotted for technical duplicates for three (WT and torus mutants) or two (all other mutants) independent experiments fitted to a hyperbolic binding curve. Kd values that were calculated from the fitted curves are reported in Table 2. (H to J) The Smchd1 hinge dimer has two clusters of positively charged residues implicated in nucleic acid binding. The first interaction site, comprising Arg1790, Arg1796, and Lys1799 from cluster 2 (cyan), is positioned on the surface of the dimer where the intersubdomain linker resides. The second interaction site is located at the central channel of the hinge dimer on the opposite side of the dimer comprising Arg1867 and Lys1873 from cluster 3 (blue). Residue Arg1848 (colored in magenta in the structural figures), which is also positioned at the central channel, is located between interaction sites 1 and 2 and is also implicated in nucleic acid binding. Panels (H) and (J) show the same view of the hinge domain dimer to show the torus residues selected for mutation (J) and cluster 2 residues (H). Distances between the residues located at the two nucleic acid interaction sites are marked beside corresponding dashed lines in (H) and (I).

  • Fig. 5 The mutations R1848A or R1867G or compromising the integrity of the hinge domain alter the nuclear localization pattern of Smchd1.

    (A to F) Immunofluorescence microscopy analysis of (A) control, nonsilencing shRNA–transduced, or shRNA-mediated SMCHD1 knockdown (KD) 293 cells in the absence (B) or presence of WT (C) or mutant (D and E) full-length Smchd1 or a variant missing the hinge domain (F). Maximum intensity projection images are shown as representative of n > 150 nuclei positive for Smchd1 overexpression per sample. Data are representative of three independent experiments. All images were obtained with identical settings between controls and all transfected cells to enable comparison between the images provided in the figure. Scale bars, 20 μm. DAPI and SMCHD1 staining and merged channels are displayed in fig. S7, and repeat experiments are shown in fig. S8.

  • Table 1 X-ray crystallography data collection and refinement statistics.

    Statistics for the highest-resolution shell are shown in parentheses.

    Structural parameters
    Smchd1 SeMet P4Smchd1 Hg R32
    Wavelength
    Resolution range48.45–3.3 (3.418–3.3)46.61–4.2 (4.35–4.2)
    Space groupP41212R32:H
    Unit cell dimensions123.549 123.549
    232.869 90 90 90
    154.058 154.058
    244.088 90 90 120
    Total reflections493800 (48634)92383 (9255)
    Unique reflections27895 (2722)8369 (821)
    Multiplicity17.7 (17.9)11.0 (11.3)
    Completeness (%)99.87 (99.71)99.38 (99.15)
    Mean I/sigma(I)14.81 (2.03)8.38 (1.67)
    Wilson B factor107.71171.93
    Rmeas0.2053 (1.956)0.2137 (1.75)
    Rpim0.04826 (0.461)0.06437 (0.5202)
    Reflections used in
    refinement
    27880 (2720)8326 (814)
    Reflections used for
    Rfree
    1396 (143)406 (40)
    CC1/20.999 (0.674)0.998 (0.812)
    Rwork0.2142 (0.2871)0.2973 (0.3700)
    Rfree0.2414 (0.3245)0.3098 (0.3368)
    Macromolecules84194104
    RMS (bonds)0.0130.016
    RMS (angles)1.741.75
    Ramachandran favored
    (%)
    94.3291.83
    Ramachandran
    allowed (%)
    4.926.42
    Clash score6.6617.35
    Average B factor117.5155.39
  • Table 2 Summary of melting temperature (Tm) and dissociation constant (Kd) values for the WT Smchd1 hinge and for those of the indicated mutants.

    N.D., not determined.

    Property of Smchd1 hingeTm (°C) at 0.05; 0.1 mg/ml*Kd (μM)
    WT53.6 ± 0.1; 52.6 ± 0.41.5 ± 0.1
    R1867G42.4 ± 0.1; 42.4 ± 0.16.2 ± 0.7
    Dimer interface mutants
      D1749A46.1 ± 0.1; 45.7 ± 0.1N.D.
      R1762A47.1 ± 0.1; 46.5 ± 0.13.2 ± 0.2
      Y1765A45.5 ± 0.1; 45.0 ± 0.11.1 ± 0.1
      V1774G35.2 ± 0.1; 34.7 ± 0.1N.D.
      D1842A52.0 ± 0.2; 51.3 ± 0.11.2 ± 0.1
      R1848A42.5 ± 0.1; 42.5 ± 0.111.4 ± 1.4
      G1872A46.1 ± 0.1; 45.7 ± 0.11.0 ± 0.1
      G1872A, G1875A, G1876A39.1 ± 0.1; 37.4 ± 0.1N.D.
      K1873A68.1 ± 0.1; 67.7 ± 0.18.1 ± 1.2
      F1874A41.6 ± 0.3; 41.2 ± 0.11.5 ± 0.1
    Cluster 1 mutants
      K1718A51.8 ± 0.2; 50.9 ± 0.11.4 ± 0.1
      R1719A52.6 ± 0.1; 51.5 ± 0.11.8 ± 0.1
      R1771A53.4 ± 0.2; 52.0 ± 0.22.4 ± 0.2
    Cluster 2 mutants
      K1789A52.0 ± 0.1; 51.4 ± 0.12.3 ± 0.1
      R1790A55.6 ± 0.1; 54.1 ± 0.57.6 ± 0.8
      R1796A54.2 ± 0.2; 52.5 ± 0.14.3 ± 0.3
      K1799A52.1 ± 0.2; 51.5 ± 0.13.4 ± 0.3
    Cluster 3 mutants
      R1869A61.9 ± 0.1; 61.6 ± 0.123.6 ± 4.6
      K1873A68.1 ± 0.1; 67.7 ± 0.18.1 ± 1.2
      K1880A44.9 ± 0.1; 44.6 ± 0.11.4 ± 0.1
    Torus mutants
      S1870MN.D.1.2 ± 0.1
      S1870NN.D.1.4 ± 0.1
      S1870QN.D.2.0 ± 0.1
      H1856WN.D.0.8 ± 0.1

    *Tm values are reported as ±SEM for the fit calculated at protein concentrations of 0.05 and 1 mg/ml. The curves for quadruplicate measurements at each concentration are displayed in fig. S6. The Tm for the wild-type hinge domain is reported as means ± SD of four independent experiments.

    Kd values were estimated from the fits shown in Fig.4 and are reported as ±SEM for the fit.

    ‡K1873A is categorized as both a dimer interface mutant and a cluster 3 mutant.

    Supplementary Materials

    • stke.sciencemag.org/cgi/content/full/13/636/eaaz5599/DC1

      Fig. S1. The Smchd1 rhombohedral crystal structure used to solve the phase problem by single anomalous dispersion and electron density around Arg1762 in the tetragonal crystal structure.

      Fig. S2. Coiled-coils flanking the core Smchd1 hinge domain are observed as two unassigned helices in the asymmetric unit, despite being present in intact MS analysis of SMCHD1 hinge dimer crystals.

      Fig. S3. Crystal structures of SMC proteins with coiled-coils.

      Fig. S4. Dimer interfaces of Smchd1 and canonical SMC hinge domains.

      Fig. S5. Size exclusion chromatograms of Smchd1 hinge mutants.

      Fig. S6. Tm values of WT and mutant Smchd1 hinge domains measured by DSF.

      Fig. S7. The mutations R1848A and R1867G and deletion or compromise of hinge domain integrity alter the nuclear localization pattern of Smchd1.

      Fig. S8. The mutations R1848A and R1867G perturb local interactions and abrogate formation of Smchd1 nuclear foci.

      Table S1. Data collection and scattering parameters for SAXS analysis.

    • This PDF file includes:

      • Fig. S1. The Smchd1 rhombohedral crystal structure used to solve the phase problem by single anomalous dispersion and electron density around Arg1762 in the tetragonal crystal structure.
      • Fig. S2. Coiled-coils flanking the core Smchd1 hinge domain are observed as two unassigned helices in the asymmetric unit, despite being present in intact MS analysis of SMCHD1 hinge dimer crystals.
      • Fig. S3. Crystal structures of SMC proteins with coiled-coils.
      • Fig. S4. Dimer interfaces of Smchd1 and canonical SMC hinge domains.
      • Fig. S5. Size exclusion chromatograms of Smchd1 hinge mutants.
      • Fig. S6. Tm values of WT and mutant Smchd1 hinge domains measured by DSF.
      • Fig. S7. The mutations R1848A and R1867G and deletion or compromise of hinge domain integrity alter the nuclear localization pattern of Smchd1.
      • Fig. S8. The mutations R1848A and R1867G perturb local interactions and abrogate formation of Smchd1 nuclear foci.
      • Table S1. Data collection and scattering parameters for SAXS analysis.

      [Download PDF]

    Stay Connected to Science Signaling

    Navigate This Article