N-terminus of hMLH1 confers interaction of hMutLalpha and hMutLbeta with hMutSalpha

Abstract

Mismatch repair is a highly conserved system that ensures replication fidelity by repairing mispairs after DNA synthesis. In humans, the two protein heterodimers hMutSalpha (hMSH2-hMSH6) and hMutLalpha (hMLH1-hPMS2) constitute the centre of the repair reaction. After recognising a DNA replication error, hMutSalpha recruits hMutLalpha, which then is thought to transduce the repair signal to the excision machinery. We have expressed an ATPase mutant of hMutLalpha as well as its individual subunits hMLH1 and hPMS2 and fragments of hMLH1, followed by examination of their interaction properties with hMutSalpha using a novel interaction assay. We show that, although the interaction requires ATP, hMutLalpha does not need to hydrolyse this nucleotide to join hMutSalpha on DNA, suggesting that ATP hydrolysis by hMutLalpha happens downstream of complex formation. The analysis of the individual subunits of hMutLalpha demonstrated that the hMutSalpha-hMutLalpha interaction is predominantly conferred by hMLH1. Further experiments revealed that only the N-terminus of hMLH1 confers this interaction. In contrast, only the C-terminus stabilised and co-immunoprecipitated hPMS2 when both proteins were co-expressed in 293T cells, indicating that dimerisation and stabilisation are mediated by the C-terminal part of hMLH1. We also examined another human homologue of bacterial MutL, hMutLbeta (hMLH1-hPMS1). We show that hMutLbeta interacts as efficiently with hMutSalpha as hMutLalpha, and that it predominantly binds to hMutSalpha via hMLH1 as well.

Figure 1

Complex formation of hMutLα with hMutSα on 200mer and 81mer DNA substrates. HeLa nuclear extract (150 µg) was incubated with 200mer or 81mer homoduplex DNA coupled to magnetic beads according to the hMutSα–hMutL interaction assay described in Materials and Methods. For each DNA substrate, two identical incubations were performed for 20 min. Then, ATP (250 µM final concentration) was added to one of the two mixtures (signified by ‘+’), and incubations were continued for 5 min. The supernatant of the DNA beads was removed and proteins bound to the substrates were eluted in two fractions, first with 700 mM NaCl (top), then with 1000 mM NaCl (bottom). Both elutions were separated on 10% polyacrylamide gels and blotted. hMSH2, hMSH6 and hMLH1 were detected using specific antibodies. The 700 mM NaCl fraction quantitatively elutes hMLH1 and its partner proteins hPMS1 and hPMS2 (hMutLα and hMutLβ) (27), while both fractions together elute hMutSα quantitatively. The ATP-dependent recruitment of hMLH1 to the DNA substrate becomes visible in the 700 mM fraction. The 1000 mM NaCl fraction serves as control that ATP has taken effect, since this nucleotide abolishes the hMutSα signal in this elution fraction (27).

Figure 2

Expression of hMutLα, hMutLβ, hMLH1, hPMS1 and hPMS2 in 293T cells. 293T cells were either cotransfected with hMLH1-hPMS2 (hMutLα) or hMLH1-hPMS1 (hMutLβ) or transfected with hMLH1, hPMS2 or hPMS1 as described in Materials and Methods. Extracts were prepared according to the protocol in Materials and Methods and 50 µg of total protein were separated on a 10% polyacrylamide gel. For comparison, a 50 µg extract of the MMR proficient cell line TK6 was analysed in parallel. hMLH1, hMSH2, hMSH6 (top), hPMS1 (middle) and hPMS2 (bottom, upper signal; the lower signal is hMSH2) were detected using specific antibodies. The signals of hMutSα (hMSH2 and hMSH6) serve as a loading control.

Figure 3

Interaction of hMutLα expressed in 293T cells with hMutSα can be assessed with HCT-116 nuclear extract. The hMutSα–hMutL interaction assay was performed with magnetic beads coupled with 200 bp homoduplex DNA according to the procedure described in Figure 1 and Materials and Methods. Left: the beads were incubated with nuclear extracts of HeLa cells (150 µg). Right: the DNA beads were incubated with a nuclear extract of the hMutL-deficient cell line HCT-116 (150 µg), or with a combination of HCT-116 extract (145 µg) with a 5 µg extract of 293T cells transfected with hMutLα. In all cases, ATP (250 µM final concentration) was added to one of two identical samples prior to elution (signified by ‘+’). Bound proteins were eluted from the DNA beads with 700 and 1000 mM NaCl. The western blot of hMutSα and hMLH1 of the 700 mM elution fraction is shown.

Figure 4

Expression of hMutLα-mpEA and interaction with hMutSα. (A) hMLH1-hPMS2 (hMutLα) or hMLH1 E34A-hPMS2 E41A (hMutLα-mpEA) were cotransfected into 293T cells. Protein extracts of the transfected cells and, for comparison, extract of the MMR proficient cell line TK6 (50 µg each) were separated in a 10% polyacrylamide gel and blotted. hMSH6, hMSH2, hMLH1 and hPMS2 were detected using specific antibodies. The signals of hMutSα (hMSH2 and hMSH6) serve as a loading control. (B) The hMutSα–hMutL interaction assay was performed with magnetic beads coupled with 200 bp homoduplex DNA according to the procedure described in Figure 1 and Materials and Methods. The beads were incubated with 145 µg of nuclear extract of HCT-116 cells supplemented with a 5 µg extract of 293T cells expressing either wild-type hMutLα or hMutLα-mpEA from the transfection displayed in (A). ATP (250 µM final concentration) was added to one of two identical samples prior to elution (signified by ‘+’). The western blots of hMSH6, hMSH2, hMLH1 and hPMS2 of the 700 mM NaCl elution fraction are shown.

Figure 5

Interaction of hMutLα and hMutLβ with hMutSα. The hMutSα–hMutL interaction assay was performed according to Figure 1 and the protocol described in Materials and Methods with magnetic beads coupled with 200 bp homoduplex DNA. Incubations were performed with a nuclear extract of HCT-116 cells (145 µg) supplemented with a 5 µg extract of 293T cells transfected with hMutLα (lanes 1 and 2), hMutLβ (lanes 3 and 4) or with 2.5 µg of each extract (lanes 5 and 6) from the transfection displayed in Figure 2. ATP (250 µM final concentration) was added to one of two identical samples prior to elution (signified by ‘+’). The western blots of hMSH6, hMSH2, hMLH1, hPMS1 and hPMS2 of the 700 mM NaCl elution fraction are shown.

Figure 6

Interaction of hMutL subunits with hMutSα. (A) The hMutSα–hMutL interaction assay was performed with magnetic beads coupled with 200 bp homoduplex DNA according to the procedure described in Figure 1 and Materials and Methods. The beads were incubated with 145 µg of nuclear extract of HCT-116 cells supplemented with a 5 µg extract of 293T cells expressing either hMLH1, hPMS2 or hPMS1 from the transfection displayed in Figure 2. In one case, the HCT-116 extract was supplemented with two extracts of 293T cells, one prepared from cells transfected with hMLH1, the other from cells transfected with hPMS2 (2.5 µg each). ATP (250 µM final concentration) was added to one of two identical samples prior to elution (signified by ‘+’). The western blots of hMSH6, hMSH2, hMLH1, hPMS1 and hPMS2 of the 700 mM NaCl elution fraction are shown. (B) The hMutSα–hMutL interaction assay was performed with a nuclear extract of HCT-116 (145 µg) supplemented with an extract of 293T cells transfected with hPMS2 (5 µg) according to the procedure described in Figure 1 and the protocol in Materials and Methods. DNA beads coupled with 81mer duplexes were used that were either correctly paired (GC) or contained a mismatch (GT). ATP (250 µM final concentration) was added to one of two identical samples (signified by ‘+’) prior to elution. The western blots of hMSH6, hMSH2 and hPMS2 of the 700 mM NaCl elution fraction are shown. Furthermore, the western blot of hMSH6 and hMSH2 of the 1000 mM NaCl elution fraction is shown to demonstrate that the presence of a mismatch actually affected hMutSα binding: the mismatch confers a greater elution resistance to hMutSα in the absence of ATP, resulting in increased signals of hMSH2 and hMSH6 on heteroduplex compared to homoduplex (third lane versus first lane). For detailed characterisation of this effect, see Plotz et al. (27).

Figure 7

Expression of hMLH1 fragments in 293T cells. 293T cells were transfected or cotransfected with either hMLH1-hPMS2 (hMutLα), LN56 or LC49. Extracts were prepared according to the protocol in Materials and Methods. Fifty micrograms of total protein were separated on a 12.5% polyacrylamide gel. For comparison, a 50 µg extract of the MMR proficient cell line TK6 was analysed in parallel. hMLH1, hMSH2 and hMSH6 were detected using specific antibodies. The hMLH1 antibody also recognised the hMLH1 fragments LN56 (56 kDa) and LC49 (49 kDa). The signals of hMutSα (hMSH2 and hMSH6) serve as a loading control. The non-specific bands visible between 37 and 50 kDa in the extracts of cells transfected with hMutLα (and also, to some degree, in the transfections with LC49) most likely represent degradation products of these proteins.

Figure 8

Expression and co-immunoprecipitation of hPMS2 with hMLH1 fragments. Top: Transfection scheme for cotransfection of 293T cells with hMLH1-hPMS2 (hMutLα), LN56-hPMS2 or LC49-hPMS2. Middle: protein extracts of these transfections were prepared and 50 µg were separated on a 12.5% polyacrylamide gel, followed by western blotting and detection of hMLH1 and hPMS2. Bottom: an immunoprecipitation was performed according to the protocol described in Materials and Methods with an antibody against hMLH1 (G168-728; Pharmingen) using 100 µg of the hMLH1-hPMS2 and LC49-hPMS2 extracts, respectively, and 200 µg of the LN56-hPMS2 extract to adjust the experiment to the different expression levels of the proteins. For one experiment, the extract containing hMutLα was used, while the antibody was omitted as a control for successful washing (data not shown). The western blot of the precipitates is shown (hMLH1 and hPMS2 were detected). A, non-specific signals from degradation products and the immunoprecipitating antibody.

Figure 9

Interaction of hMLH1 fragments with hMutSα. The hMutSα–hMutL interaction assay was performed with magnetic beads coupled with 200 bp homoduplex DNA according to the procedure described in Figure 1 and Materials and Methods. The beads were incubated with 145 µg of nuclear extract of HCT-116 cells supplemented with a 5 µg extract of 293T cells transfected with LN56, LC49, hPMS2 or LC49-hPMS2 from the transfection displayed in Figures 7 and 8. ATP (250 µM final concentration) was added to one of two identical samples prior to elution (signified by ‘+’). (A) Western blot of the original incubation mixtures of this experiment showing the presence and adequate concentration of the hMutL constructs in the samples. (B) Western blots of hMSH6, hMSH2, hMLH1 and hPMS2 of the 700 mM NaCl elution fraction are shown.

Figure 10

Expression of hMutLα Δ9/10 and interaction with hMutSα. (A) 293T cells were cotransfected either with hMLH1-hPMS2 (hMutLα) or hMLH1Δ9/10-hPMS1 (hMutLα Δ9/10) as described in Materials and Methods. Extracts prepared according to the protocol in Materials and Methods were separated on a 10% polyacrylamide gel (50 µg of total protein in each lane) and blotted. For comparison, a 50 µg extract of the MMR proficient cell line TK6 was analysed in parallel. hMLH1 and hPMS2 were detected using specific antibodies. (B) The hMutSα–hMutL interaction assay was performed with magnetic beads coupled with 200 bp homoduplex DNA according to the procedure described in Figure 1 and Materials and Methods. The beads were incubated with nuclear extracts of HCT-116 cells (145 µg) supplemented with a 5 µg extract of 293T cells transfected either with wild-type hMutLα or with hMutLαΔ9/10 from the transfection displayed in (A). ATP (250 µM final concentration) was added to one of two identical samples prior to elution (signified by ‘+’), and bound proteins were eluted from the DNA beads. The western blots of hMSH2, hMSH6, hMLH1 and hPMS2 of the 700 mM elution fraction are shown.

Figure 11

Diagram of the hMLH1 constructs. In the top diagram, the hMLH1 protein is shown with its functional domains as known to date. Numbers refer to amino acid positions. hMLH1 encodes a protein of 756 amino acids including a highly conserved N-terminal ATPase domain (‘ATPase domain’) as well as a variant C-terminal domain (‘PMS interaction domain’), which has been suggested to confer dimerisation with hPMS1 and hPMS2 (56). The ATPase domain also incorporates the conserved loops L45, which have been suggested to contain the MutS–MutL interaction interface in bacterial MMR (44). The diagrams below show the localisation of the hMLH1 fragments LN56 and LC49 as well as the hMLH1 Δ9/10 protein in comparison to the full-length protein. Right: +, interaction; –, no interaction.