Nuclear localization of Schizosaccharomyces pombe Mcm2/Cdc19p requires MCM complex assembly

Abstract

The minichromosome maintenance (MCM) proteins MCM2-MCM7 are conserved eukaryotic replication factors that assemble in a heterohexameric complex. In fission yeast, these proteins are nuclear throughout the cell cycle. In studying the mechanism that regulates assembly of the MCM complex, we analyzed the cis and trans elements required for nuclear localization of a single subunit, Mcm2p. Mutation of any single mcm gene leads to redistribution of wild-type MCM subunits to the cytoplasm, and this redistribution depends on an active nuclear export system. We identified the nuclear localization signal sequences of Mcm2p and showed that these are required for nuclear targeting of other MCM subunits. In turn, Mcm2p must associate with other MCM proteins for its proper localization; nuclear localization of MCM proteins thus requires assembly of MCM proteins in a complex. We suggest that coupling complex assembly to nuclear targeting and retention ensures that only intact heterohexameric MCM complexes remain nuclear.

Figure 1

MCM protein localization is disrupted in mcm mutant strains. (A) The MCM proteins are constitutively nuclear proteins in wild-type cells. Fission yeast cells FY798 (for Mcm2p-HA) and FY1023 (for Mcm3p-HA) were incubated at 32°C and prepared for immunofluorescence as described in MATERIALS AND METHODS. Cells are shown stained with DAPI (i–iv) or antibodies (v–viii) against Mcm2p-HA (v), Mcm3p-HA (vi), Mcm4p (vii), and Mcm6p (viii). (B) Mcm2 protein localization is compromised in mcm mutant strains but not in other S phase mutant strains. mcm mutant strains FY836 (cdc21-M68, i) and FY979 (mis5-268, ii) and other S phase mutants FY835 (orp1-4, iii), FY832 (cdc22-M45, iv), FY831 (pol1-1, v), and FY837 (cdc17-K42, vi) were prepared for immunofluorescence as described in MATERIALS AND METHODS after incubation for 4 h at 36°C. Antibody staining is shown for Mcm2p-HA. (C) MCM protein localization is perturbed in cdc19 mutant strains. Strains FY243 (cdc19-P1) and FY1100 (cdc19-P1 mcm3-HA), for detection of Mcm3p-HA, were prepared for immunofluorescence as described in MATERIALS AND METHODS after incubation for 4 h at 36°C. Antibody staining is shown for Mcm3p-HA (i), Mcm4p (ii), and Mcm6p (iii). Bar, 10 μm.

Figure 2

Relocalization of MCM proteins in mcm mutant is crm1 dependent. (A) mcm3-HA cells fail to form colonies at 17°C. Assay of colony formation of wild-type and mcm3-HA cells at 32°C for 4 d or 17°C for 13 d. (B and C) MCM protein localization in mcm3-HA. Cells were incubated at 32°C or the restrictive temperature (17°C) for 6 h before fixation for immunofluorescence. Cells were visualized with DAPI staining (i and iii) or antibody staining (ii and iv) for Mcm3p-HA (B, ii and iv) or Mcm6p (C, ii and iv). (D) MCM protein localization in crm1-809 mcm3-HA (i) and crm1-11R cdc19-P1 (ii) mutant strains. Double mutant cells were incubated at the restrictive temperature (17°C) for 6 h or at 36°C for 4 h before fixation for immunofluorescence. Antibody staining is shown for Mcm6p. MCM proteins remain in the nucleus in wild-type or crm1 mutants at 17°C (our unpublished results). Bar, 10 μm.

Figure 3

Multiple regions of Mcm2p affect nuclear localization. (A–J) Wild-type fission yeast cells expressing either wild-type or mutant Mcm2p were prepared for immunofluorescence as described in MATERIALS AND METHODS. Note that some cells fail to stain because of variation in the plasmid copy number or the loss of the episome. DAPI (left panel) and Mcm2p-HA (right panel) antibody staining are shown for wild-type Mcm2p (A), overexpressed wild-type Mcm2p (B), zinc finger point mutant Mcm2p-M2 (C), NTP-binding point mutant Mcm2p-M7 (D), NLS1 point mutant Mcm2p-M9 (E), NLS2 point mutant Mcm2p-M10 (F), small amino-terminal deletion Mcm2p-D3 (G), carboxy-terminal truncation Mcm2p-D6 (H), amino-terminal truncation Mcm2p-D7 (I), and MCM core deletion Mcm2p-D10 (J). Bar, 10 μm. (K) Scheme of Mcm2p structural motifs and location of mutations. Mcm2p structural motifs as well as positions of point and deletion mutations are indicated (Forsburg et al., 1997). See Table 4 for description of mutations.

Figure 4

Mcm2p NLS1 is a function of NLS. Upper panel lists the fission yeast strains used for the NLS reporter assay as described in MATERIALS AND METHODS. Localization of GFP-βgal is summarized as C for cytoplasmic or N for nuclear. Lower panels show localization of NLS reporter proteins in fission yeast. Fission yeast cells expressing the GFP-βgal fusion protein were visualized by DAPI staining (i) or GFP fluorescence (ii). (A) No NLS; (B) the SV40 NLS; (C) the mutant SV40 NLS; (D) NLS1 from Mcm2p; (E) the point mutant version of NLS1 from Mcm2p-M9; and (F) NLS2 from Mcm2p. Bar, 10 μm.

Figure 5

Overexpression of NLS mutant Mcm2p-M9 sequesters wild-type MCM proteins in cytoplasm. (A) Mcm2p-M9 with a heterologous NLS is functional. The mutant cdc19-P1 cells harboring a control vector (pREP4X), wild-type Mcm2/Cdc19p (pSLF176), NLS mutant Mcm2/Cdc19p-M9 (pSLF196-M9), mutant SV40NLS-Mcm2/Cdc19p-M9 (pSGP96-M9), or SV40NLS-Mcm2/Cdc19p-M9 (pSGP86-M9) were streaked out on minimal medium plates and incubated for 4 d at 36°C before photography. (B) Overproduction of the NLS mutant protein Mcm2p-M9 is toxic in the cdc19-P1 mcm3-HA strain. The strains expressing wild-type Mcm2/Cdc19p (pSGP56) or mutant Mcm2/Cdc19p-M9 (pSGP56-M9) were streaked out on minimal medium plates with (+ thia) or without (− thia) thiamine and incubated at the permissive temperature (29°C) for 4 d before photography. (C) Localization of MCM proteins in cdc19-P1 mcm3-HA strain when mutant Mcm2p-M9 is overproduced. cdc19-P1 mcm3-HA cells expressing wild-type Mcm2/Cdc19p or mutant Mcm2/Cdc19p-M9 were prepared for immunofluorescence analysis after induction in the absence of thiamine for 20 h at 29°C. Antibody staining is shown for Mcm3p-HA (i and iv), Mcm4p (ii and v), and Mcm6p (iii and vi). Bar, 10 μm.

Figure 6

Mutant Mcm2p-HA localization in crm1 strain. Cytoplasmic mutant Mcm2p-HA proteins do not accumulate in the nucleus in the nuclear export–defective strain crm1-809. Cells harboring a plasmid encoding wild-type or mutant Mcm2 proteins were grown at the restrictive temperature (17°C) for 6 h before fixation for immunofluorescence analysis. Wild-type Mcm2p was overproduced by induction in the absence of thiamine for 16 h followed by incubation at 17°C for 6 h. Cells were visualized with anti-HA antibody. (A) Wild-type Mcm2p; (B) overexpressed wild-type Mcm2p; (C) zinc finger mutant Mcm2p-M2; (D) NLS mutant Mcm2p-M9; (E) NLS mutant Mcm2p-M10; (F) carboxy-terminal truncation Mcm2p-D6; (G) amino-terminal truncation Mcm2p-D7; and (H) MCM core domain deletion Mcm2p-D10. Bar, 10 μm.

Figure 7

MCM complex assembly and nuclear localization. A speculative model coupling MCM complex assembly and nuclear localization. The six MCM proteins depend on NLS sequences supplied by MCM2 and MCM3 for nuclear import. The NLS sequences on MCM2 and MCM3 are inactive until association with the core complex, MCM4/6/7 and MCM5, respectively. Alternatively, the assembly of the full hexameric complex in the cytoplasm may be required for nuclear targeting. Once imported into the nucleus, the MCM subcomplexes have active NESs that are inactivated by the association of the two subcomplexes as an intact hexameric MCM complex. Inactivation of the conditionally mutant MCM proteins results in disassembly of the hexameric complex, exposure of the active NESs, and subsequent export of the MCM subcomplexes.