Reconstitution of human DNA licensing and the structural and functional analysis of key intermediates

Abstract

Human DNA licensing initiates replication fork assembly and DNA replication. This reaction promotes the loading of the hMCM2-7 complex on DNA, which represents the core of the replicative helicase that unwinds DNA during S-phase. Here, we report the reconstitution of human DNA licensing using purified proteins. We showed that the in vitro reaction is specific and results in the assembly of high-salt resistant hMCM2-7 double-hexamers. With ATPγS, an hORC1-5-hCDC6-hCDT1-hMCM2-7 (hOCCM) assembles independent of hORC6, but hORC6 enhances double-hexamer formation. We determined the hOCCM structure, which showed that hORC-hCDC6 recruits hMCM2-7 via five hMCM winged-helix domains. The structure highlights how hORC1 activates the hCDC6 ATPase and uncovered an unexpected role for hCDC6 ATPase in complex disassembly. We identified that hCDC6 binding to hORC1-5 stabilises hORC2-DNA interactions and supports hMCM3-dependent recruitment of hMCM2-7. Finally, the structure allowed us to locate cancer-associated mutations at the hCDC6-hMCM3 interface, which showed specific helicase loading defects.

Fig. 1. Reconstitution of human DNA licensing using purified proteins.

a Purified hORC1-5, hORC6 and hCDC6. b IP pull down with Strep-hORC1-5 as bait in the presence and absence of a 90 bp yeast ARS1 DNA containing the A, B1 and B2 elements. c Pre-RC assay performed under low salt conditions with hORC1-5, hCDC6 and hORC6 independently and in combination. d Purified hMCM2-7. e Purified full-length hCDT1 and hCDT1_ΔN (aa158-546). f Pre-RC assays using hCDT1 and hCDT1_ΔN, under low and high salt conditions in pre-RC buffer which contains 1 mM ATP. All SDS-PAGE gels are representative of three independent biological replicates. Source data are provided as a Source Data file.

Fig. 2. Comparative analysis of double-hexamer formation and Cdt1 interaction with MCM2-7 using purified yeast and human proteins.

Assembled pre-RC reactions were subjected to washes with a sodium chloride gradient ranging from 0 to 500 mM for (a) yeast and (b) human. SDS-PAGE gels are representative of two independent biological replicates. c Negative stain EM 2D class averages generated with CryoSPARC carried out on pre-RC assay reactions washed with 300 mM NaCl from yeast and human. The percentage numbers represent the proportion of the total particles in each class of either double hexamer (DH) or single hexamer (SH). hMCM2-7 DH formation was more variable than yMcm2-7 DH-formation, suggesting additional regulation mechanisms exist. Mass photometry on (d) yMCM2-7 and yCdt1 and (e) hMCM2-7 and hCDT1_ΔN, was carried out in solution with a two-fold excess of CDT1. Only yeast proteins form a complex in solution, as detected by an upshift (marked with an asterisk) in the yMCM2-7 mass peak that is proportional to the mass of yCdt1. f Pre-RC assay under low salt conditions was carried out in the absence and presence of hCDT1_ΔN, highlighting that initial hMCM2-7 recruitment occurs without hCDT1. SDS-PAGE gel is representative of three independent biological replicates. Source data are provided as a Source Data file.

Fig. 3. Specificity of human DNA licensing.

Specificity of the pre-RC assay was assessed under (a) low and (b) high salt conditions by sequentially omitting loading factors hORC1-5 (lane 1), hCDC6 (lane 2) or hCDT1 (lane 3) compared to a reaction containing all pre-RC proteins (lane 4). SDS-PAGE gels are representative of three independent biological replicates. c Pre-RC assay under low and high salt wash conditions was carried out with hB2-Lamin and yARS1 DNA. Highlighting that human DNA licensing can occur on non-human DNA. SDS-PAGE gel is representative of two independent biological replicates. Source data are provided as a Source Data file.

Fig. 4. The role of hORC6 in DNA licensing.

Specificity of the pre-RC assay was assessed in the presence (WT pre-RC) and absence of hORC6 (hORC6(-)) under (a) low and (b) high salt conditions. c Normalised hMCM2-7 signal in the presence and absence of hORC6 after high salt wash conditions (as in b). Mean plotted; individual data points are marked with black circles. n = 3 independent experiments, error bars represent standard deviation, statistical significance was calculated using paired, two-tailed t test, *P ≤ 0.05 (P = 0.0246). d Pre-RC assays were carried out under low salt conditions in the presence of ATP or ATPγS. e Two-step pre-RC assay to assess the timing of hORC6 function carried out in low salt conditions with ATP or ATPγS. The absence and presence of hORC6 is denoted by (–) and (+) respectively. Lane 1 represents the input for the second part of the reactions, shown in lanes 3–7. All SDS-PAGE gels are representative of three independent biological replicates. f Normalised hMCM2-7 signal following the two-step pre-RC assay under low salt conditions. Lane 4–7 from (e) are quantified. The absence and presence of hORC6 is denoted by (−) and (+) respectively. Mean plotted; individual data points are marked with black circles. n = 3 independent experiments, error bars represent standard deviation, statistical significance was calculated using RM one-way ANOVA with Tukey’s multiple comparisons test, ns—not significant, *P ≤ 0.05. P values as presented on graph from left to right: P = 0.0381, 0.0995, 0.0414, 0.0380. g An in-solution assay was developed to probe double hexamer formation in the presence and absence of hORC6 by negative stain EM. h The resulting analysis of double hexamer formation was determined based on 2D class averages resulting after 2 rounds of 2D classification from samples applied onto grids at the indicated timepoints. Mean plotted, individual data points are marked with black circles. Three independent experiments, error bars represent the standard deviation. Source data are provided as a Source Data file.

Fig. 5. The electron microscopy structure of the hOCCM.

a Negative stain class averages of a low-salt washed pre-RC reaction containing ATPyS. We observed the following pre-RC complexes: single hMCM2-7 hexamers, hOCCM and hMCM2-7 double hexamers. The green densities represent hMCM2-7 rings, shaded for orientation with light green highlighting the C-terminus and darker green N-terminus. hORC-hCDC6 rings have been shaded with pink. b Representative 2D class averages from cryo-EM of the hOCCM denoting top and side views (upper), and projections of the resulting 3D map onto the 2D class averages showing good fit (lower). c The segmented experimental density map is shown from different angles and coloured by protein subunit. d The hOCCM molecular model was built from the cryo-EM structure (PDB: 8RWV). e Structure comparison of hOCCM (in colours) with yOCCM (grey, PDB: 5V8F).

Fig. 6. hORC-hCDC6 in the hOCCM structure.

a Domain organisation of hORC1-5 and hCDC6 proteins, and b Organisation of the AAA+ and Winged Helix Domains of the DNA-bound hORC1-5-hCDC6 ring in the hOCCM. c Comparison of hORC1-5-hCDC6 ring in the hOCCM (coloured) with human hORC1-5-DNA structure (grey, PDB: 7JPS). d Reorganisation of hORC5 WHD in hOCCM when compared to hORC1-5-DNA structure (grey, PDB: 7JPS). e hORC2 (purple) and hORC3 (red) become reorganised to accommodate hCDC6 in the hOCCM structure compared to the hORC1-5-DNA only structure (grey, PDB: 7JPS). f No hORC6 densities were observed in the hOCCM, while yOrc6 was observed in yOCCM (blue, PDB: 5V8F) and DmOrc6 in the active DmOrc1-5 structure (green, PDB:7JK6). Human, yeast and Drosophila models were aligned by ORC3. Experimental densities were downloaded from the EMDB and aligned. The Orc6 densities from yeast and Drosophila were segmented by a radius of 4 Å and gaussian filtered by 1.5 Å (grey) to display experimental Orc6 position relative to the hOCCM density (coloured).

Fig. 7. The hORC1-hCDC6 interface becomes restructured in the hOCCM and hCDC6 ATPase is crucial for complex disassembly.

a Observed stabilisation of an hORC1 helix formed between residues 660–674, which is unstructured before hCDC6 (grey) binding as evidenced by docking hORC1 (PDB: 7JPS) in light grey and hORC1 of hOCCM in green. b Purified hORC1-5_WT and hORC1-5 ORC1 Walker B mutant (hORC1-5_E621Q), hCDC6_WT and hCDC6 Walker B mutant (hCDC6_E285Q). c The pre-RC assay was carried out under low and high salt wash conditions with hORC1-5_WT, hORC1-5_E621Q, hCDC6_WT or hCDC6_E285Q. SDS-PAGE gel is representative of three independent biological replicates. Source data are provided as a Source Data file.

Fig. 8. DNA structure and interactions in hOCCM.

a The DNA interacting with yORC in the yOCCM (PDB: 5V8F; grey) is bent by 10° respect to that of the hOCCM (orange). The hOCCM DNA is stretched with respect to the canonical B-form DNA (light grey) in the upper part of the DNA. b Detailed hORC1-5 and hCDC6 interactions with DNA. The upper deformed part of the DNA is contacted by the ISM, and the lower less deformed part by WHDs. The view is rotated 180° with respect to (a). c hORC4-DNA interactions. hORC4 (light blue) is positioned closer to the DNA in hOCCM than in yOCCM (PDB: 58VF; grey). The DNA is not well-resolved enough in hORC1-5 (PDB: 7JPS; light grey). d hORC4 sequence alignment. The ISM loop region is longer in yeast containing a helix insert, while higher eukaryotes contain a conserved alternative ISM sequence. e Alternative Orc4 WHD close interactions with DNA in the human (blue) and yeast (grey) proteins.

Fig. 9. Five hMCM2-7 WHDs make contact with hORC1-5-hCDC6.

a Five hMCM (grey) WHDs (coloured) form contacts with hORC1-5-hCDC6 in hOCCM, contrasting with only four inter-subunit interactions in yOCCM (PDB: 5V8F). The conserved interfaces are highlighted in brackets, and the unique interfaces are circled. b Pre-RC assay under low and high salt wash conditions was carried out with WT and hMCM2_ΔC (hMCM2-WHD truncated mutant) hMCM2-7. SDS-PAGE gel is representative of two independent biological replicates. c Comparison between yeast and human MCM2 domain organisation. hMCM2_ΔC lacks aa826-904, the WHD domain (red). d Purification of the hMCM5_ΔC truncation mutant results in sub-stoichiometric hMCM2-7 complexes compared to purified hMCM2-7_WT. e Structural comparison of the human and yeast MCM6-CDT1 interface and sequence alignment of the corresponding region. Source data are provided as a Source Data file.

Fig. 10. The hORC2-WHD in the hOCCM is reorganised upon DNA and hCDC6 binding.

a Structure of hORC2-WHD. The hORC2 (mauve) is stabilised in the open claw position by the presence of hCDC6 AAA+ domain (light grey), as compared with the closed claw conformations of hORC2-WHD in hORC2-5 (PDB: 7JPQ) and in DmOrc2 (PDB: 4XGC) which is not compatible to forming an interaction with DNA. b The orientation adopted by the hORC2-WHD (mauve) upon hCDC6 binding falls within the vicinity of the hMCM3-WHD (light blue) in the hOCCM structure. c The location of relevant COSMIC mutations (red) in hCDC6 (grey) in the hOCCM structure may locally disrupt interaction with hMCM3 via disruptions to inter- and intra-protein interactions. d Purified hCDC6_WT, hCDC6_F375C and hCDC6_C394F. e Pre-RC assay under low and high salt wash conditions was carried out with hCDC6_WT, hCDC6_F375C and hCDC6_C394F. SDS-PAGE gel is representative of four independent biological replicates. f Corresponding western blot analysis for hCDC6 (anti-hCDC6 mouse monoclonal, Santa Cruz, Cat# sc-9964) and hCDT1 (anti-hCDT1 mouse monoclonal, Santa Cruz, Cat# sc-365305) in the presence of hCDC6_WT, hCDC6_F375C and hCDC6_C394F. ATPγS arrests complex formation at the hOCCM stage and shows maximal hCDC6 and hCDT1 association. Western blots are representative of two independent biological replicates. Source data are provided as a Source Data file.