CDK phosphorylation of a novel NLS-NES module distributed between two subunits of the Mcm2-7 complex prevents chromosomal rereplication

Abstract

Cyclin-dependent kinases (CDKs) use multiple mechanisms to block reassembly of prereplicative complexes (pre-RCs) at replication origins to prevent inappropriate rereplication. In Saccharomyces cerevisiae, one of these mechanisms promotes the net nuclear export of a pre-RC component, the Mcm2-7 complex, during S, G2, and M phases. Here we identify two partial nuclear localization signals (NLSs) on Mcm2 and Mcm3 that are each necessary, but not sufficient, for nuclear localization of the Mcm2-7 complex. When brought together in cis, however, the two partial signals constitute a potent NLS, sufficient for robust nuclear localization when fused to an otherwise cytoplasmic protein. We also identify a Crm1-dependent nuclear export signal (NES) adjacent to the Mcm3 NLS. Remarkably, the Mcm2-Mcm3 NLS and the Mcm3 NES are sufficient to form a transport module that recapitulates the cell cycle-regulated localization of the entire Mcm2-7 complex. Moreover, we show that CDK regulation promotes net export by phosphorylation of the Mcm3 portion of this module and that nuclear export of the Mcm2-7 complex is sufficient to disrupt replication initiation. We speculate that the distribution of partial transport signals among distinct subunits of a complex may enhance the specificity of protein localization and raises the possibility that previously undetected distributed transport signals are used by other multiprotein complexes.

Figure 1.

Wild-type and mutant Mcm2 and Mcm3 nucleocytoplasmic transport signals. (A) Key amino acid sequences of wild-type and mutant Mcm2 NLS and Mcm3 NLS-NES transport module. Putative NLSs and leucines in leucine-rich motif are in bold. Consensus CDK phosphorylation sites are underlined with putative phosphoacceptor residues in gray. Amino acids not spelled out are indicated in parentheses. Amino acid substitutions in the mutant Mcm proteins used in this study are indicated below the wild-type sequences. (B) A schematic of the MCM transport module containing the Mcm2 NLS and the Mcm3 NLS-NES. Below the schematic are the WT and corresponding mutant alleles of this construct that are discussed in the text.

Figure 2.

The Mcm2 and Mcm3 NLSs are each required for nuclear localization of the Mcm2-7 complex. (A) The Mcm2 NLS is required for nuclear localization of Mcm2-GFP. YJL3265 (MCM2::{MCM2-GFP}), YJL1231 (MCM2::{mcm2-nls-GFP}), and YJL1228 (MCM2::{mcm2-nls-GFP-SVNLS₂}) were arrested in G1 phase with α-factor then examined by fluorescence microscopy. (B) The Mcm3 NLS is required for nuclear localization of GFP-Mcm3. YJL2669 (MCM3::{GFP-MCM3}), YJL2675 (MCM3::{GFP-mcm3-nls}), and YJL2665 (MCM3::{SVNLS₂-GFP-mcm3-nls}) were arrested in G1 phase with α-factor then examined by fluorescence microscopy. (C) The Mcm2 NLS is required for nuclear localization of Mcm7-GFP. Cultures of YJL3765 (MCM7-GFP mcm2-td MCM2), YJL3840 (MCM7-GFP mcm2-td mcm2-nls), and YJL3799 (MCM7-GFP mcm2-td mcm2-nls-SVNLS₂) growing exponentially at 23°C were arrested in G2/M phase by addition of nocodazole for 3 h. Galactose was added and cultures were shifted to 37°C for 30 min to degrade Mcm2-td. Cells were then released from G2/M phase into a G1 phase arrest by shifting them to fresh medium containing α-factor for 2 h (still in the presence of galactose and at 37°C). Cells examined by fluorescence microscopy are shown just before the G2/M phase release (NOC arrest) and at the G1-phase block (α-factor arrest). (D) The Mcm3 NLS is required for nuclear localization of Mcm7-GFP. Cultures of YJL3464 (MCM7-GFP mcm3-td MCM3), YJL3469 (MCM7-GFP mcm3-td mcm3-nls), and YJL3474 (MCM7-GFP mcm3-td SVNLS₂-mcm3-nls) were subjected to the same experimental protocol described for Figure 2C.

Figure 3.

Together the Mcm2 and Mcm3 NLSs are sufficient to direct the nuclear localization of a heterologous protein. Overnight cultures of YJL310 containing URA3-marked centromeric plasmids pAR109 (pGAL-SV40NLS-GFP₃), pAR110 (pGAL-NLS2-GFP₃), pAR101 (pGAL-NLS3-GFP₃), pAR113 (pGAL-NLS2-NLS3-GFP₃), pAR126 (pGAL-nls2-NLS3-GFP₃), or pAR127 (pGAL-NLS2-nls3-GFP₃) and growing in SRaf-Ura medium were shifted to YEPRaf medium for 90 min before splitting each culture in two and adding α-factor to one-half and nocodazole to the other. One hour later, as the cultures were approaching a complete arrest, galactose was added to induce synthesis of the GFP₃ fusion proteins. After 2 h of induction, cells were examined by fluorescence microscopy.

Figure 4.

Mcm3 contains an NES, which in combination with the Mcm2 NLS and Mcm3 NLS, directs the cell cycle-regulated localization of GFP₃. (A) The cytoplasmic localization of NLS2-NLS3NES3-GFP₃ in G2/M phase is dependent on the leucine-rich motif of the Mcm3 NES. Cultures of YJL4662 (crm1Δ trp1::{pGAL-NLS2-NLS3NES3-GFP₃),TRP1} [crm1-T539C]) and YJL4860 (crm1Δ trp1::{pGAL-NLS2-NLS3nes3-GFP₃),TRP1} [crm1-T539C]) constitutively expressing their GFP₃ fusion proteins in YEP-Gal medium were arrested for 90 min in either G1 phase with α-factor or G2/M phase with nocodazole before being examined by fluorescence microscopy. (B) NLS2-NLS3NES3-GFP₃ is exported from the nucleus by a Crm1-dependent mechanism. YJL4662 constitutively expressing NLS2-NLS3NES3-GFP₃ in YEPGal medium was arrested in G1 phase with α-factor. At time 0, cells were released from the arrest in the presence or absence of 100 ng/ml leptomycin B (LMB). Samples were collected at the indicated times for flow cytometry and fluorescence microscopy.

Figure 5.

The Mcm3 NES is required for efficient nuclear export of Mcm3 and Mcm7 in cycling cells. (A) Mutation of the Mcm3 NES increases the population of cells containing nuclear GFP-Mcm3 during exponential growth. Exponentially growing cultures of YJL2162 (GFP-MCM3) (top) and YJL2741 (GFP-mcm3-nes) (bottom) were examined by fluorescence and DIC microscopy. Representative fluorescent fields are shown. Cells were categorized as unbudded (UB), small budded (SB), or uninucleate large budded (LB) based on their DIC image and DAPI fluorescence. Binucleate large budded cells, which comprise approximately half of all large budded cells, are unseparated postmitotic cells (many in G1 phase) and were thus not included in the analysis. Each of these categories was further subclassified into cells with or without detectable nuclear GFP fluorescence above cytoplasmic levels as exemplified by the pictures of individual cells. Bar graphs show the percent of cells with nuclear or nonnuclear GFP fluorescence and the total number of cells counted (in parentheses) for each bud stage. (B) Mutation of the Mcm3 NES increases the population of cells containing nuclear Mcm7-GFP during exponential growth. The same experiment and analysis described in Figure 5A was performed for YJL1979 (MCM3 MCM7-GFP) and YJL5439 (mcm3-nes MCM7-GFP).

Figure 6.

The Mcm3 NES is required for the timely nuclear export of Mcm3 and Mcm7. (A) Mutation of the Mcm3 NES delays the nuclear export of GFP-Mcm3. YJL2162 (GFP-MCM3) and YJL2741 (GFP-mcm3-nes) cells were arrested in G1 phase with alpha factor and at time 0 synchronously released into a G2/M phase arrest with nocodazole. At the indicated times, samples were taken for flow cytometry and fluorescence microscopy. (Left) Representative images from 0, 60, and 100 min. (Right) Flow cytometry profiles. (B) Mutation of the Mcm3 NES delays the nuclear export of Mcm7-GFP. YJL1979 (MCM3 MCM7-GFP) and YJL5439 (mcm3-nes MCM7-GFP) were treated and analyzed as described in C, except microscopic images from the 0-, 50-, and 70-min time points are shown.

Figure 7.

In vitro and in vivo phosphorylation of Mcm3 is dependent on the consensus CDK phosphorylation sites in the Mcm3 NLS3NES3 module. (A) GST-Mcm3 is phosphorylated by Cdc28-Clb2 kinase in vitro. In vitro kinase reactions were performed with purified Cdc28-Clb2 kinase mixed with purified GST-Mcm3, GST-Mcm3-cdk5A, or GST-Mcm3-cdk7A. Reaction products electrophoresed on SDS-PAGE were subjected to autoradiography (top) and Coomassie staining (bottom). (B) The CDK consensus sites in the Mcm3 NLS3NES3 are required for in vivo phosphorylation of Mcm3. YJL4110 (MCM3), YJL4313 (Myc₆-MCM3), YJL4324 (Myc₆-mcm3-cdk5A), and YJL4315 (Myc₆-mcm3-cdk7A) were metabolically labeled with ³²P-orthophosphate for 1 h before lysis and immunoprecipitation with 9E10 anti-Myc mAb. Immunoprecipitates were electrophoresed on SDS-PAGE and either subjected to autoradiography (top) or immunoblotted with rabbit anti-Myc polyclonal antibodies (bottom).

Figure 8.

The Mcm3 consensus CDK phosphorylation sites are required for nuclear exclusion of the Mcm2-7 complex. (A) Net nuclear export of NLS2-NLS3NES3-GFP₃ is dependent on the consensus CDK phosphorylation sites in NLS3NES3. YJL4662 (crm1Δ trp1::{pGAL-NLS2-NLS3NES3-GFP₃),TRP1} [crm1-T539C]), YJL5750 (trp1::{pGAL-NLS2-NLS3NES3-cdk5A-GFP₃),TRP1}) and YJL5753 (trp1::{pGAL-NLS2-NLS3NES3-cdk4A-GFP₃), TRP1}) growing exponentially and constitutively expressing their GFP₃ fusion proteins in YEPGal medium were examined by fluorescence microscopy. (B) Mcm3 CDK consensus sites are required for nuclear exclusion of Mcm3. Fluorescent microscopy of nocodazole-arrested YJL2714 (GFP-mcm3-cdk4A), YJL2720 (GFP-mcm3-cdk5A), and YJL2314(GFP-mcm3-cdk7A). (C) Mcm3 Cdk consensus sites are required for nuclear exclusion of other Mcm subunits. Fluorescence microscopy of nocodazole-arrested YJL2033 (mcm2::{MCM2-GFP}), YJL4165 (mcm2::{MCM2-GFP} mcm3-cdk5A), YJL2160 (GFP-MCM3), YJL2720 (GFP-mcm3-cdk5A), YJL2037 (mcm4::{MCM4-GFP}), YJL4169 (mcm4::{MCM4-GFP}, mcm3-cdk5A), YJL2217 (mcm7::{MCM7-GFP}), and YJL4167 (mcm7:: {MCM7-GFP}, mcm3-cdk5A).

Figure 9.

Together the Mcm3 NES and the consensus CDK sites in Mcm3 NLS are required for nuclear export of GFP-Mcm3 and Mcm7-GFP. (A) GFP-Mcm3 is strongly nuclear throughout the cell cycle when both the leucine-rich motif of the Mcm3 NES and the four CDK consensus sites flanking the Mcm3 NLS are mutated. Exponentially growing cultures of YJL2162 (GFP-MCM3), YJL2714 (GFP-mcm3-cdk4A), and YJL5216 (GFP-mcm3-cdk4A-nes) were examined by fluorescence and DIC microscopy. Cells were categorized as unbudded (UB), small budded (SB), or uninucleate large budded (LB) based on their DIC image and DAPI fluorescence. Binucleate large budded cells, which comprise approximately half of all large budded cells, are unseparated postmitotic cells (many in G1 phase) and were thus not included in the analysis. Each bud category was further subclassified into cells with or without detectable nuclear GFP fluorescence above cytoplasmic levels. Bar graphs show the percent of cells with nuclear or nonnuclear GFP fluorescence and the total number of cells counted (in parentheses) for each bud category. (B) Mcm7-GFP is strongly nuclear throughout the cell cycle when both the leucine-rich motif of the Mcm3 NES and the four CDK consensus sites flanking the Mcm3 NLS are mutated. Exponentially growing cultures of YJL1979 (MCM3 MCM7-GFP), YJL5691 (mcm3-cdk4A mcm7::{MCM7-GFP}), and YJL5221 (mcm3-cdk4A-nes MCM7-GFP) were examined by fluorescence and DIC microscopy as described in Figure 9A.

Figure 10.

Phosphomimic mutations of the Mcm3 CDK consensus sites promote net nuclear export of Mcm proteins in G1 phase and this mislocalization impairs replication initiation. (A) (First Row) YJL2160 (GFP-MCM3), YJL1265 (GFP-mcm3-cdk5ED), YJL1260 (SVNLS₂-GFP-mcm3-cdk5ED). (Second Row) YJL2033 (mcm2::{MCM2-GFP}), YJL4162 (mcm2::{MCM2-GFP} mcm3-cdk5ED), YJL4094 (mcm2::{MCM2-GFP} SVNLS₂-mcm3-cdk5ED). (Third Row) YJL2037 (mcm4::{MCM4-GFP}), YJL4103 (mcm4::{MCM4-GFP} mcm3-cdk5ED), YJL4098 (mcm4::{MCM4-GFP} SVNLS₂-mcm3-cdk5ED). (Fourth Row) YJL2217 (mcm7::{MCM7-GFP}), YJL4108 (mcm7::{MCM7-GFP} mcm3-cdk5ED), YJL4096 (mcm7::{MCM7-GFP} SVNLS₂-mcm3-cdk5ED). All strains were arrested in G1 phase with α-factor for 90 min (>95% unbudded) before being examined by fluorescence microscopy. (B) Mcm mislocalization due to the mcm3-cdk5ED mutation impairs replication initiation. Plasmid loss rates per were measured over 12-20 generations for plasmids YCp50 (1 ARS) and pJW1112 (8 ARSs) in YJL310 (MCM3), YJL2160 (GFP-MCM3), YJL1265 (GFP-mcm3-cdk5ED), and YJL1259 (SVNLS₂-GFP-mcm3-cdk5ED). Histogram shows average loss rate per generation and SE for four independent isolates of each plasmid-yeast pair. Complete failure to replicate a plasmid would result in a theoretical loss rate of 50% per generation.