Prior to structural analysis by EM we characterised recombinant Geminin for its DNA

Transcription

1 Supplementary Data Functional characterisation of recombinant human Geminin Prior to structural analysis by EM we characterised recombinant Geminin for its DNA replication inhibition activity in the cell-free DNA replication assay 1. To test the activity of recombinant Geminin, in vitro DNA replication reactions were performed in the presence of recombinant his-tagged human Geminin (H 6 -hsgeminin). The purified recombinant Geminin was homogeneous and migrated as a single peak corresponding to a molecular weight of ~240 kda, indicating either an unusual shape of the protein and/or assembly into oligomers. G1 nuclei incubated in a physiological buffer supporting DNA elongation resulted in 2.2 % of nuclei synthesising DNA, in keeping with a small amount of S phase contaminants 1 (Supplementary Fig. 1a,i online). In contrast, 22% of G1 nuclei synthesised DNA upon incubation in S phase cytosol, representing true initiation in vitro (Supplementary Fig. 1a,ii online). Addition of Geminin to the reaction resulted in 6.7% of the nuclei replicating, indicating that 66% of replication-competent nuclei failed to synthesise DNA in vitro (Supplementary Fig. 1a,iii). In control reactions, DNA elongation was not affected by addition of recombinant Geminin to S phase nuclei with 64% of them continuing to replicate (data not shown). Taken together, these data demonstrate that recombinant Geminin is a potent inhibitor of the initiation of DNA replication in vitro, but does not arrest DNA elongation from origins that have fired in vivo prior to preparation of the nuclear templates from intact cells. We next analysed chromatin-bound protein fractions from the reactions identical to those described above. There was no Mcm2 protein bound to chromatin in fractions prepared from G1 nuclei incubated in elongation buffer (Supplementary Fig. 1b online). Incubation of G1 nuclei in S phase cytosol resulted in recruitment of Mcm2 to chromatin, indicative of pre-rc assembly in vitro (Supplementary Fig. 1b online). Addition of Geminin to co-incubations of G1 nuclei in

2 S phase cytosol banned the replication license by blocking loading of Mcm2 onto chromatin (Supplementary Fig. 1b online). Importantly, this block to pre-rc assembly coincides with binding of recombinant Geminin to chromatin (Supplementary Fig. 1b online). This suggests that the replication inhibitory activity of Geminin may also be coupled to its potential to bind DNA. These data demonstrate that bacterially expressed recombinant Geminin subjected to structural analysis is fully functional in repressing origin licensing and blocking initiation of DNA replication. Properties of an amino-terminal truncated form of human Geminin To distinguish between different structural domains in our model, we analysed a truncated form of Geminin with deletion of the amino terminal part of the protein, which is predicted to form an independent structural and functional domain (Supplementary Fig. 3a online). In Xenopus it has been demonstrated that the first 80 amino acid residues of the frog ortholog are not required for Geminin s replication inhibition activity, suggesting that the truncated protein folds into the active form 2,3. We therefore engineered a Geminin- Nt truncated derivative of human Geminin lacking the first 80 aa residues. Prior to analysis by EM, the Geminin- Nt recombinant protein was tested for its biochemical activity in the cell-free replication assay and showed activity (inhibition of 65.5% of competent nuclei) almost identical to that of the full-length protein (Supplementary Fig.1a,iii online), indicating that overall folding of either the truncated protein or domains involved in replication control have not been affected by the deletion. Next we analysed Geminin- Nt by EM at conditions identical to those employed for fulllength Geminin. Raw images of Geminin- Nt appeared as small, thread-like elongated particles. In contrast to particles of the full-length protein, the head-like bulk at the end was absent. Approximately 300 molecular images were selected and subjected to multivariate statistical 2

3 analysis 4,5. Five typical raw images of Geminin- Nt and one of the class averages are shown in Supplementary Fig. 4d online. All particles of Geminin- Nt are uniformly shaped as a short filament of ~25Å in diameter and ~170Å in length, thinner and slightly longer than we expected. Since the molecular images did not show the head-like domain at one end of the filaments, we conclude that the head-like part in the structure represents the amino terminal part of Geminin. The smaller diameter of Geminin- Nt particles in comparison to particles of the full-length protein is indicative of dimeric rather than tetrameric organisation of the molecule. In keeping with this observation, crosslinking experiments show that the major crosslinked species of Geminin- Nt is a protein band migrating as a dimer (Supplementary Fig. 4b online). This is in keeping with crystallographic data on a peptide representing the coiled coil domain of Geminin 6. Domain structure modelling of human Geminin Since the structure of the coiled coil domain is unavailable at present, we modelled the potential Geminin dimer with SwissModel structural homology modelling software using the 1GK6 atomic structure as a template 7. This structure, originally used by Thepaut and co-authors for the molecular replacement method 6, was identified as a potential match for the Geminin sequence by the Threader 3 program 8. The resulting structural model is a leucine/isoleucine zipper 9,10 that is coordinated into a parallel dimer of α-helices, with almost all hydrophobic aa residues forming the interface between monomers (Supplementary Fig. 4c online). The diameter of the dimer is ~22Å, consistent with our findings for Geminin- Nt particles. The length of the 37 aa dimer was measured as ~60Å. Considering that Geminin- Nt is 132 residues long, it is plausible to suggest that the overall length of the putative long helical structural organisation could be close to ~170Å as determined for Geminin- Nt. The extended length could be explained by the absence of 3

4 dimer-dimer interactions within the Geminin molecule and thus straightening of the Geminin- Nt dimer. References 1. Stoeber, K., Mills, A.D., Kubota, Y., Krude, T., Romanowski, P., Marheineke, K., Laskey, R.A. & Williams, G.H. Cdc6 protein causes premature entry into S phase in a mammalian cell-free system. EMBO J. 17, (1998). 2. McGarry, T.J. & Kirschner, M.W. Geminin, an inhibitor of DNA replication, is degraded during mitosis. Cell 93, (1998). 3. Kroll, K.L., Salic, A.N., Evans, L.M. & Kirschner, M.W. Geminin, a neuralizing molecule that demarcates the future neural plate at the onset of gastrulation. Development 125, (1998). 4. van Heel, M., Harauz, G., Orlova, E.V., Schmidt, R. & Schatz, M. A new generation of the IMAGIC image processing system. J. Struct. Biol. 116, (1996). 5. Serysheva, I.I., Orlova, E.V., Chiu, W., Sherman, M.B, Hamilton, S.L. & van Heel, M. Electron cryomicroscopy and angular reconstitution used to visualize the skeletal muscle calcium release channel. Nat Struct Biol. 2, (1995). 6. Thepaut, M., Hoh, F., Dumas, C., Calas, B., Strub, M.P. & Padilla A. Crystallization and preliminary X-ray crystallographic analysis of human Geminin coiled-coil domain. Biochim Biophys Acta. 1599, (2002). 7. Herrmann, H., Strelkov, S.V., Feja, B., Rogers, K.R., Brettel, M., Lustig, A., Haner, M., Parry, D.A, Steinert, P.M., Burkhard, P. & Aebi U. The intermediate filament protein 4

5 consensus motif of helix 2B: its atomic structure and contribution to assembly. J Mol Biol. 298, (2000). 8. Jones, D.T., Taylor, W.R. & Thornton, J.M. A new approach to protein fold recognition. Nature 358, (1992). 9. O'Shea, E.K., Rutkowski, R. & Kim, P.S. Evidence that the leucine zipper is a coiled coil. Science 243, (1989). 10. Struhl, K. Helix-turn-helix, zinc-finger, and leucine-zipper motifs for eukaryotic transcriptional regulatory proteins. Trends Biochem Sci. 14, (1989). 5