Through the interaction of the SRA domain with the targeting domain (TS) of Dnmt1 (Leonhardt et al. 1992; Achour et al. 2007), ICBP90 facilitates the loading of Dnmt1 to replicating regions (Bostick et al. 2007; Sharif et al. 2007).
However, the molecular mechanism that controls the activity of Dnmt1 and the recognition of hemi-methylated DNA remains largely unknown.
C.5.1 Dnmt1 – disruption of a protein dimer
Recombinant Dnmt1 was subjected to size exclusion chromatography (E.2.3.6) on a Superdex200 gel filtration column (GE Healthcare) followed by immunoblot analysis (E.2.3.3). Dnmt1 runs at an apparent molecular size of 400kDa which corresponds approximately to twice the molecular weight of Dnmt1 (183 kDA). The presence of 2M NaCl in the running buffer could not disrupt the intramolecular protein interaction, indicating a probable hydrophobic interaction domain (Figure 46).
Figure 46 Recombinant Dnmt1 – an apparent dimer
A.) Recombinant Dnmt1 was loaded on a Superdex200 gelfiltration column (GE Healthcare) in running buffer (200mM NaCl/2M NaCl, 20mM Tris, pH 8.0; 1mM MgCl2, 10% glycerol). Upper panel. Dnmt1 in running buffer supplemented with 200mM NaCl. Lower panel. Dnmt1 in running buffer with 2M NaCl. B.) Dnmt1 (30µg) was adjusted to the indicated SDS concentration and either incubated for 30min at 37°C or directly loaded on a Superose6 gelfiltration column (GE Healthcare) in running buffer (150mM NaCl, 20mM Tris, pH 7.5; 1mM EDTA, 10% glycerol) supplemented with the indicated concentrations of SDS. For both columns, 500µl fractions were collected, TCA precipitated and subjected to SDS-PAGE and Western Blot analysis (detection with D1-2G3 rat monoclonal antibody; E.1.6). Molecular weight standards and fractions are indicated
To further test this assumption, Dnmt1 was run on a Superose6 gelfiltration column in the presence of 0,01% SDS at 4°C and 0,1% SDS at room temperature. SDS seemed to partially disrupt the Dnmt1 dimer (Figure 46). Higher concentrations of SDS or incubation of the sample at 37°C for 30min prior to loading dramatically enhanced this effect. Not only the dimer peak vanished and Dnmt1 was shifted to fractions of lower molecular weight, but also multimerization could be observed. Tertiary structure and intramolecular interactions were disturbed allowing for intermolecular hydrophobic interactions to occur.
Taken together, these data indicate a Dnmt1 dimer, whose interaction might be mediated by a hydrophobic part of multiple parts of the protein.
C.5.2 Dnmt1 interacts with the TS domain
The human TS domain was expressed as MBP fusion construct and purified as described (E.2.8.7). Dnmt1 was purified from baculovirus infected Sf21 cells (E.2.7.2). Individual proteins together with BSA or Dnmt1 and hTS in a 1:6 molar ratio were incubated for 1h at 30°C and loaded on a gelfiltration column. Immunoblot analysis of collected fractions showed that Dnmt1 eluted in a fraction corresponding to a molecular mass of about 400 kDa (Figure 47, A). This is approximately two-fold the molecular weight of Dnmt1 (183 kDa) indicating the
presence of a dimeric complex. The hTS domain eluted with a distinct peak at about 65 kDa corresponding to a TS dimer and higher molecular weight complexes in the size range above 500 kDA indicating multimerization of the TS domain (Figure 47).
Dnmt1 incubated with hTS showed partial disruption of both Dnmt1 and hTS dimers. Dnmt1 was shifted to higher molecular weight driven by multimerization together with hTS and a downshift of multiple fractions of lower molecular weight suggesting Dnmt1-hTS heterodimers. Signals of Dnmt1 in fractions smaller than its own molecular weight can be explained by an altered structural conformation together with the hTS resulting in a different running behavior. In contrast, the hTS domain showed an altered distribution over the fractions of the column, directly inverse to the observation of Dnmt1. Incubation of TS with Dnmt1 led to a downshift of multimerized TS domains and a peak broadening in the low molecular weight region.
Accordingly, the hTS domain which by itself can form a stable dimer, seems to be the determinant of dimerization of Dnmt1.
In order to further test the interaction properties of Dnmt1 with the TS domain, a co-immunoprecipitation was performed. Dnmt1 was bound to proteinG sepharose coupled with Dnmt1 specific antibody 2E8 (E.1.6) and incubated with an excess of hTS. As indicated in Figure 47 (B), Dnmt1 is able to interact with and precipitate the TS domain.
In a recent publication, K. Fellinger (Fellinger et al. 2009) provides evidence by means of biochemical characterization that Dnmt1 forms a stable dimer. The parallel orientation of the Dnmt1 dimer is mediated by the interaction of the TS domain.
Figure 47 Dnmt1 interacts with hTS domain – gelfiltration analysis
A.) Molecular interactions between Dnmt1 and the TS domain were analyzed by gelfiltration on a Superdex200 column in 20mM Tris, pH 7.5, 150mM NaCl, 1mM EDTA, 1mM DTT. Dnmt1 and the TS domain were analyzed after mixing both proteins for 1 h at 30°C in a 1:6 ratio. For reference, individual proteins were incubated together with BSA. The proteins were visualized by Western blot with the 2C1 rat monoclonal antibody recognizing both, Dnmt1 and the TS domain (E.2.8.7, E.2.7.2). The proteins (Dnmt1 : d and hTS : t) and the migration of the size standards are indicated. B.) Dnmt1 co-immunoprecipitates the TS domain. MBP-TS was cleaved with TEV Protease (lane 2) and incubated with the Dnmt1 specific antibody 2E8 coupled to proteinG beads. 2% of the flowthrough and 50% of the beads were loaded (lanes 3 and 4). Recombinant Dnmt1 (lane 6) was incubated with a 50 fold excess of the hTS domain for 1h at 30°C. The reaction was incubated with the 2E8 antibody coupled to beads and 10% of the flowthrough and 50% of the beads were loaded on a 6% SDS-PAGE (lane 7 and 8).
Proteins were stained with Coomassie Blue and the position of Dnmt1, the antibody heavy chain (HC), the Maltose Binding Protein (MBP) and the human TS domain are indicated