Functional and Physical Interactions between Autonomously ...

Traffic 2004; 5: 925–935 Blackwell Munksgaard

Copyright

#

Blackwell Munksgaard 2004

doi: 10.1111/j.1600-0854.2004.00233.x

Functional and Physical Interactions between Autonomously Replicating Sequence-Binding Factor 1 and the Nuclear Transport Machinery Christian M. Loch1, Nima Mosammaparast2, Tsuyoshi Miyake1, Lucy F. Pemberton2,* and Rong Li1,* 1 Department of Biochemistry and Molecular Genetics, 2Center for Cell Signaling and Department of Microbiology, School of Medicine, PO Box 800733, University of Virginia, Charlottesville, VA 22908–0733, USA *Corresponding authors: Rong Li, [email protected]; Lucy F. Pemberton, [email protected]

Autonomously replicating sequence-binding factor 1 (Abf1p) is a site-specific DNA binding protein in Saccharomyces cerevisiae that functions to regulate multiple nuclear events including DNA replication, transcriptional activation, and gene silencing. Previous work indicates that the multiple functions of Abf1p are conferred by the carboxy-terminus of the protein, which can be further dissected into two important clusters of amino acid residues (CS1 and CS2). Here we present genetic and cell biological evidence for a critical role of CS1 in proper nuclear localization of Abf1p. Mutations in CS1 cause severe defects in cell growth, nuclear translocation, and Abf1p-mediated gene regulation, which can be rescued by a heterologous nuclear localization sequence (NLS). In addition, the CS1-domain can mediate the import of a CS1-GFP fusion protein. Importantly, the CS1-mediated nuclear import depends on the Ran guanine nucleotide exchange factor Prp20p. Interestingly, a single amino acid change in CS1 (K625I) also causes the protein to be exported out of the nucleus via the Crm1p-dependent pathway. The temperature-sensitive growth phenotype of this particular mutant can be overcome by overexpression of Kap121p/Pse1p, a well-established nuclear transport receptor. Biochemical studies indicate that Pse1p binds to a region of Abf1p upstream of CS1 in a RanGTP-sensitive manner, suggesting that Abf1p has a second distinct NLS and can be imported into the nucleus by several overlapping pathways. We propose that the link between Abf1p and the nuclear transport machinery may also be important for partitioning multiple Abf1pmediated nuclear processes. Key words: ABF1, Kap121, nuclear localization sequence, nuclear import, transcription factor Received 22 April 2004, revised and accepted for publication 25 August 2004

Eukaryotic chromosomal DNA serves as the primary site for the actions of multiple nuclear machineries, such as

those of DNA replication, transcription, DNA repair, and recombination. It is now increasingly evident that these nuclear processes are intimately coordinated in a spatial and temporal manner to ensure an efficient and adequate response to cellular needs and environmental cues. Such coordination is often achieved through the actions of common regulatory proteins shared by these machineries. A well-documented example is site-specific transcription factors that bind to various cis-acting regulatory sequences in the genome. Although they are initially identified on the basis of their functions in transcriptional activation, many of these factors are capable of facilitating chromosomal events in addition to transcription. This is at least in part due to their ability to recruit chromatin-modifying complexes that in turn increase chromatin accessibility for multiple nuclear machineries (1). For example, in addition to their established role in stimulating transcription initiation, a number of transcription factors also bind to auxiliary sequences adjacent to origins of DNA replication (2), recombination ‘hot spots’ (3,4), and transcriptional silencers to facilitate specific chromosomal events at these loci (5). More recently, it was demonstrated that promoterbound transcription factors can stimulate nucleotide excision repair (NER) by inducing chromatin reorganization at the promoter-proximal region (6). Thus, a better understanding of the structural and functional relationship of these multifunctional proteins may provide molecular insight into the coordinated regulation of diverse nuclear processes in eukaryotic cells. Autonomously replicating sequence-binding factor 1 (Abf1p) is an essential, site-specific DNA binding protein in Saccharomyces cerevisiae that stimulates multiple nuclear events including DNA replication (7,8), gene activation (9–14), and gene silencing (15–17). Like many transcription factors, Abf1p can be structurally divided into an amino-terminal DNA binding domain (DBD; approximately aa 1–500) (18), and a carboxy-terminal activation domain (AD; aa 604–731) (19). While the DBD of Abf1p is responsible for recognizing specific DNA sequences, the C-terminal AD confers the stimulatory activity of Abf1p in all three chromosomal events (20). Detailed mutagenesis of the AD indicates the presence of two critical regions, named C-terminal sequence1 (CS1) (aa 624–628) and CS2 (aa 639–662) (20). Although both regions are important for the essential function of Abf1p in supporting cell viability, they seem to play distinct roles in various Abf1p-mediated nuclear events. CS2 is capable 925

Loch et al.

of inducing chromatin remodeling (20), which most likely accounts for its positive role in multiple nuclear events such as transcription and DNA replication. In contrast, the exact molecular basis for the function of CS1 in supporting cell viability remains unclear. In the context of the fulllength protein, an alanine-scanning mutation at CS1 abolishes the gene-silencing function of Abf1p. However, when the AD of Abf1p is fused to the GAL4 DBD and the fusion protein is tested in GAL4 binding site-dependent functional assays, the same CS1 mutation does not impair the ability of the AD to remodel chromatin and facilitate gene activation, gene silencing, or DNA replication (20). Curiously, the GAL4-Abf1p fusion protein harboring the CS1 mutation exhibits a stronger activity than its wildtype counterpart in transcriptional activation and chromatin remodeling (20). In the current study, we show that CS1 plays an important role in the nuclear localization of Abf1p. The defects in cell viability, nuclear localization, and gene activation associated with CS1 mutants can be rescued by a heterologous nuclear localization sequence (NLS). We also establish a genetic and biochemical link between Abf1p and the nuclear transport receptor Kap121p/Pse1p. Kap121p/ Pse1p binds to a second NLS upstream of CS1, suggesting that this essential nuclear protein has several overlapping routes to the nucleus. The connection between Abf1p and the nuclear transport system also suggests a potential role of the latter in coordinating and/or partitioning the diverse functions of Abf1p in the nucleus.

Results Identification of a temperature-sensitive (ts) mutant of ABF1 with impaired CS1 function Previous deletion analysis of Abf1p showed that amino acids (aa) 1–662 of the protein are sufficient for supporting normal cell growth (20). Furthermore, alanine-scanning mutagenesis of the C-terminal activation domain of Abf1p revealed two regions (CS1 for aa 624–628 and CS2 for aa 639–662) that are important for viability (see Figure 1A). In particular, cells harboring the Ala(624–628) or Ala(644–648) mutation in the context of Abf1p (1–662) exhibited a lethal phenotype (20). To facilitate functional characterization of these regulatory regions, we utilized error-prone polymerase chain reaction (PCR) to randomly mutagenize the 30 portion of the gene that encompasses these two regions (residues 604–662). Yeast cells transformed with the resulting library of abf1(1–662) mutants were screened for the temperature-sensitive (ts) growth phenotype at 37 °C. Several ts mutants were identified and found to all contain the same amino acid change within CS1 (K625I) (Figure 1B). The severity of the ts phenotype associated with K625I appears to be more pronounced than that of abf1-1 (data not shown), a previously characterized abf1 mutant with impaired DNA 926

binding capability due to a mutation in the 50 portion of the gene (8).

CS1 plays a key role in mediating nuclear localization of Abf1p As part of the investigation of the exact molecular defect(s) associated with the K625I mutant, we compared the subcellular localization of the wild-type and mutant Abf1p. Green fluorescent protein (GFP) was fused to various Abf1p constructs. The GFP fusion genes were then shuffled in yeast cells and shown to behave the same as their nonfusion counterparts, indicating that the GFP tag did not interfere with Abf1p function (data not shown). Fluorescence microscopy revealed that, as expected, the wild-type Abf1p-GFP fusion proteins (1–731 and 1–662) were exclusively localized to the nuclei (panels 2 and 3 in Figure 1C). In contrast, a significant amount of the Abf1pK625I-GFP protein was mislocalized in the cytoplasm (panel 4), suggesting that the mutant protein was not effectively imported into the nucleus. To confirm the mutant phenotype, we added a heterologous NLS from the SV40 large T antigen to the Abfp1-GFP fusion bearing the K625I mutation. As shown in panel 5 of Figure 1C, the added NLS largely complemented the mislocalization defect of the K625I mutant. The shuffled strain harboring this allele also exhibited wild-type growth at both permissive and nonpermissive temperatures (Figure 1B). Thus, the temperature sensitivity of abf1-K625I is most likely due to its cytoplasmic mislocalization. To further characterize the role of CS1 in nuclear localization of Abf1p, a panel of Abf1p-GFP fusions was constructed and their subcellular localization examined by fluorescence microscopy. All constructs were expressed to levels equal to or greater than full length wild-type Abf1p-GFP (data not shown). The percentage of cells that displayed distinct subcellular fluorescent patterns in each case was calculated and is summarized in Figure 2. Similar to the phenotype of the K625I mutant, the alaninesubstitution mutant at CS1(A624-8) in the context of Abf1p(1–662)-GFP showed dispersed whole-cell fluorescence (compare constructs 4 and 5), suggesting that NLS had been disrupted. In contrast, the alanine-substitution mutant that disrupted the CS2 function (A644-8) was still predominantly localized to the nucleus (construct 3). In fact, the Abf1p-GFP that lacked the entire CS2 region (deletion after residue 643) was still capable of efficient nuclear localization (construct 7). This is consistent with our previous finding that CS1 and CS2 have distinct roles in imparting Abf1p function at the molecular level (20). Our work showed that all CS1-containing fragments of Abf1p examined in the current study were capable of mediating nuclear localization (e.g. constructs 7, 9, 10, 12, and 13). In particular, a short peptide that contains the minimal transcriptional activation domain of Abf1p (residues 604–662) was sufficient to serve as a potent NLS (construct 15). Once again, the NLS activity of this C-terminal peptide Traffic 2004; 5: 925–935

ABF1 and Nuclear Localization

A

DNA binding domain 1 40

91

CS1 CS2 624–628 639–662

323

496

731

ABF1 620

625

640

DAHKSKRQHLSDITLEERNED CS1

B

37 °C

RT

1–731

1

1–662

2 3

1–62/K625I

4

1–662/K625I+NLS

C GFP

1–731

1–662

1–662/K625I 1–662/K625I +NLS

GFP

Phase

1

2

3

was largely impaired by the alanine substitution mutation at CS1 (A624-8), but not that at CS2 (A644-8) (compare constructs 16 and 17). These results lead us to the conclusion that the CS1 region of Abf1p is not only important but also sufficient for efficient nuclear localization of the protein. In addition to the GFP fusions described above, our study also revealed a somewhat unexpected phenotype of two GFP fusions that contained Abf1p sequence (1–607) and (323–607). Neither construct contained the CS1 sequence, yet both exhibited discernible nuclear accumulation as compared to GFP alone (constructs 8 and 11), suggesting the presence of a second NLS. However, the finding that fluorescence was equilibrated between the nucleus and cytoplasm in a significant subpopulation of cells harboring these two constructs (44% and 36%, respectively) indicated that this NLS mediated import of the fusion less efficiently than the CS1 NLS. We interpret this as an Traffic 2004: 5: 925–935

4

5

Figure 1: Characterization of a temperature-sensitive mutant of abf1. A) Diagram of the Abf1p protein. The shaded areas represent the bipartite DNA binding domain. The solid squares represent CS1 and CS2 in the activation domain. Also shown are CS1 (boxed) and the surrounding amino acid sequences. B) abf1-K625I displays the ts phenotype at 37 °C. A serial dilution of the yeast cultures was spotted on plates and grown at permissive and restrictive temperatures. Fusion of the SV40 NLS rescues the temperature sensitivity of abf1-K625I. C) Abf1p-K625I is mislocalized to the cytoplasm. Yeast cells that contain various Abf1p-GFP constructs were exami ned by f luorescence microscopy. Fusion of the nuclear localization signal (NLS) from the SV40 large T antigen restores nuclear localization to Abf1p-K625I (panel 5). The scale bar in panel 1 represents 10 m in length.

indication that, while CS1 is required for exclusive and maximal nuclear accumulation, the sequence upstream of CS1 (residues 323–607) could mediate limited nuclear import in the absence of the entire C-terminal domain. Thus, CS1 may serve as the primary sequence for ABF1 nuclear localization, in the absence of which the upstream region may function as a second NLS. The K625I change also facilitates nuclear export of the mutant protein Strictly speaking, the increased cytoplasmic localization of the K625I mutant of Abf1p could be due to decreased nuclear import and/or increased nuclear export. In fact, the K625I mutation results in the sequence I(625)RQHLSDITL, which loosely fits the consensus sequence of the Crm1p-dependent nuclear export signal (NLS) (L-X(2,3)-[LIVFM]-X(2,3)-L-X-[LI]) (21,22). To test this hypothesis, we utilized a yeast strain that is sensitive to the Crm1p-specific inhibitor leptomycin B (LMB) (23). Two 927

Loch et al.

DNA binding domains

CS2 CS1 (core)

N

C

1) 1–731

98%

2) 1–662

96%

4%

3) 1–662/A644–8

79%

21%

4) 1–662/A624–8

2%

100%

5) 1–662/K625I

100% NLS 98%

2%

7) 1–643

99%

1%

8) 1–607

56%

44%

9) 323–662

98%

2%

10) 323–643

96%

4%

11) 323–607

64%

36%

12) 438–643

100%

13) 496–643

100%

6) 1–662/K625I–NLS

14) 438–607

100%

15) 604–662

99%

1%

16) 604–662/A644–8

88%

12%

17) 604–662/A624–8

15%

85%

mutant forms of Abf1p, Abf1p-K625I and Abf1p-(A624-8), as well as wild-type Abf1p, were expressed as GFP fusions in this strain and analyzed by fluorescence micro-

Figure 2: CS1 is the primary NLS of Abf1p. GFP-tagged fragments of Abf1p were constructed and examined by fluorescence microscopy. The phenotypes observed were classified into two types: distinct nuclear localization (N) and no distinct nuclear accumulation, in which GFP was equilibrated between the nucleus and cytoplasm (C). At least 150 cells in each case were examined and the percentages of the N and C populations were calculated.

scopy. As shown in Figure 3, treatment with LMB dramatically increased the nuclear accumulation of the K625I mutant, although the cytoplasmic fluorescence still

Figure 3: The K625I mutation in Abf1p creates a Crm1pdependent nuclear export signal. Residues 1–662 of wildtype Abf1p, Abf1p-K625I, or Abf1p-(A624-8) were expressed as GFP fusion proteins in a leptomycin B (LMB) sensitive yeast strain with or without LMB (200 nM). Cells were stained with Hoechst and analyzed by fluorescence microscopy. The scale bar in the lower right panel for K625I represents 10 m in length.

928

Traffic 2004; 5: 925–935

ABF1 and Nuclear Localization

ABF1(1–662)

ABF1(1–662K625I+NLS)

abf1–1

ABF1(1–662+NLS)

Genetic interactions of ABF1 with the nuclear transport machinery Karyopherins bind to the nuclear localization sequences present on cargo proteins and mediate their transport

ABF1(1–731;WT)

ABF1(1–662K625I+NLS)

ABF1(1–662+NLS)

ABF1(1–662/K625I)

ABF1(1–662)

abf1–1

ABF1(1–731;WT)

Rescue of localization defect rescues most, but not all, of the gene regulation function of K625I Despite the important roles of Abf1p in stimulating multiple nuclear events in yeast, the essential function of Abf1p in supporting cell viability is mostly likely due to its role in transcriptional activation. In the current work, we sought to assess the impact of the nuclear import function of CS1 on the expression of some of the recently identified Abf1pregulated genes (24). Northern analysis in Figure 4 shows the effects of wild-type and various mutant Abf1p on transcription of three Abf1p-targeted genes, i.e. IPP1, PRO3, and MFA, at 30 °C (lanes 1–6) and 37 °C (lanes 7–12). The transcription levels of IPP1 and PRO3 were lower in the abf1-1 than the wild-type strain background, whereas that

of MFA1 was enhanced in the mutant background (compare lane 1 with 2; 7 with 8). Therefore, Abf1p normally activates transcription of the former two and represses the last. As expected, Abf1p (1–662), behaved the same as the full length wild-type Abf1p (1–731) (compare lane 1 with 3; 7 with 9), consistent with the previous observation that the last 69 amino acids of the protein are dispensable for Abf1p function (20). The K625I mutation in the context of Abf1p (1–662) markedly decreased the mRNA levels of IPP1 and PRO3 but increased that of MFA1 (compare lane 3 with 4; 9 with 10). Importantly, the exogenous NLS fused to the K625I mutant to a large extent restored the regulatory function of Abf1p (compare lane 4 with 6; 10 with 12). Thus, the changes in gene expression in the K625I background are most likely due to the cytoplasmic mislocalization of the mutant protein. However, it is also noteworthy that the NLS-fused K625I mutant did not fully restore the transcription of the target genes to the wildtype levels (e.g. compare lane 5 with 6; 11 with 12, see Discussion).

ABF1(1–662/K625I)

remained at a significant level. This result suggests that the K625I mutation may have indeed fortuitously created an NES within the CS1 NLS. Although the NLS may be able to function, the NES appears more active, resulting in accumulation of the mutant protein in the cytoplasm. In contrast, nuclear accumulation of the Abf1p-(A624-8) mutant was not observed in the LMB sensitive strain (Figure 3), suggesting that the five-alanine substitutions have completely destroyed the NLS function of CS1, but in this case, not created an Crm1p-dependent NES.

7

8

9 10 11 12

IPP1 PRO3 MFA1 ACT1 rRNA 1

2

3

4

30 °C Traffic 2004: 5: 925–935

5

6

36 °C

Figure 4: Impact of CS1 on Abf1p-mediated gene regulation. Northern analysis of several Abf1p-regulated genes in the presence of various ABF1 alleles. IPP1 and PRO3 represent genes normally activated, while MFA1 represents genes normally repressed by Abf1p. In the abf1K625I strain (columns 4, 10), all three genes show misregulation of transcript levels relative to wild-type strains (columns 1, 3, 5 and 7, 9, 11). Exogenous NLS fusion to the K625I mutant protein restores some, but not all, of the regulatory activity of Abf1p (columns 6, 12). ACT1 in the Northern blots and ethidium bromide-stained rRNA were used as internal controls.

929

Loch et al.

through the nuclear pore complex (NPC) (21,22,25). This process requires Ran, a small guanosine-50 -triphosphatase (GTPase) that exists in two conformations (22,26). The predominant GTP-bound Ran in the nucleus and GDP-bound Ran in the cytosol establish a gradient that is critical to the nuclear transport process (22). In an attempt to determine the underlying mechanism by which CS1 dictates nuclear import or accumulation of Abf1p, we used the GFP fusion protein containing the CS1 region (496–643) as a reporter and screened through a panel of mutant strains that are defective for different karyopherins known to function in nuclear import. We tested strains deleted for the genes encoding KAP104, KAP123, MTR10/KAP111, KAP114, SXM1/KAP108, NMD5/ KAP119, KAP120, PDR6/KAP122, and MSN5, and bearing temperature-sensitive alleles of KAP60, KAP95, and KAP121/PSE1. The nuclear localization of the GFP-CS1 fusion protein was not significantly affected in any of the single mutant backgrounds (data not shown). As Abf1p has an essential nuclear function, it is likely that CS1 is recognized by more than one karyopherin and that these different transport receptors may act in a functionally redundant manner to mediate Abf1p import. This could explain our observation that no single karyopherin mutants affected CS1 function. The Ran GTPase cycle and maintenance of the RanGTP/RanGDP gradient across the nuclear membrane are required for the function of all karyopherins. We sought to determine whether CS1mediated nuclear import would depend upon the Ran guanine nucleotide exchange factor (GEF) Prp20, which converts the guanosine-50 -diphosphate (GDP)-bound to the GTP-bound form of Ran, in common with other karyopherin-mediated import pathways (27). Abf1p (496–673)-GFP was introduced into a ts mutant prp20-1 and the isogenic wild-type strain. GFP fusion of Sua7, a nuclear protein imported by Kap114p (L.F. Pemberton and J.L. Hodges, personal communication), was used as a positive control in the study. The cells were shifted to 37 °C for 1 h and observed under the microscope. As shown in Figure 5, a large proportion of both Abf1p and Sua7-GFP fusion proteins were clearly relocalized from the nucleus to the cytoplasm in the mutant, but not wild-type, Prp20 background at the elevated temperature, demonstrating that their import is regulated by the activity of the small GTPase Ran. The same result was obtained in a Ran–GAP mutant background (data not shown). These results support the notion that nuclear import of Abf1p is mediated by one or more karyopherin family members. As an alternative approach to elucidate the role of the CS1 region in nuclear localization, a multicopy suppressor screen was undertaken to identify potential genes that, when overexpressed, could suppress the temperature sensitivity of abf1-K625I. Four nearly identical genomic clones that restored growth of abf1-K625I at the nonpermissive temperature were isolated. Subsequent mapping 930

Abf1p-GFP

Hoechst

Sua7-GFP

Hoechst

WT

prp20-1

WT

prp20-1

Figure 5: CS1-mediated nuclear import requires a functional Ran nucleotide exchange factor, Prp20p. Abf1p(496–673)-GFP (A) and Sua7-GFP (B) were introduced into prp20-1 and isogenic wild-type strain. The cells were shifted to 37 °C for 1 h and observed under the microscope for GFP and Hoechst staining. The scale bar in the bottom right panel represents 10 m in length.

of the genomic fragments identified KAP121/PSE1 as the gene responsible for the multicopy suppression (Figure 6). Kap121p/Pse1p represents one of the few essential karyopherins in yeast (22,28,29) and has been implicated in the nuclear import of a variety of nuclear proteins, including Pho4p (30), Ste12p (31), and Yap1p (32). Suppression of the temperature sensitivity by KAP121 was specific to abf1-K625I, as two other temperature-sensitive strains, abf1–1 and abf1(1–607), were not complemented by KAP121 overexpression (Figure 6). The results suggested that Kap121p may be one of the karyopherins that can mediate the nuclear import of Abf1p. Kap121p binds the central region of Abf1p Suppression of the growth phenotype of abf1-K625I by KAP121 overexpression indicated a genetic interaction Traffic 2004; 5: 925–935

ABF1 and Nuclear Localization

+ pRS425 +pRS425/KAP121

RT

import of Abf1p. Our results suggest that while Kap121p/ Pse1p recognizes the second NLS upstream of CS1, other members of the karyopherin family are most likely responsible for direct mediation of the NLS function of CS1.

1–662 abf1–1 Discussion

1–662/ 1–607 K625I

30°C

37°C

Figure 6: KAP121 suppresses the temperature sensitivity of abf1-K625I. Comparison of growth phenotypes of cells harboring different alleles of ABF1 at various temperatures. Strains of each ABF1 background were complemented with either a multicopy KAP121-expressing plasmid (shaded) or the corresponding empty vector (open). Note abf1-K625I is more sensitive to elevated temperature than abf1-1.

between these two proteins. To explore a direct physical interaction between the two proteins, various fragments of Abf1p were fused with the glutathione-S-transferase (GST) (Figure 7A). The purified GST fusion proteins were tested in vitro for their ability to bind to purified Kap121p that was fused to the maltose binding protein (MBP) (33). Surprisingly, results from the GST pulldown assay indicate that the C-terminal AD, which contains the CS1 region (604–662), was neither necessary nor sufficient for binding to Kap121p/Pse1p (lane 7). However, Abf1p(323–604) was sufficient for a specific interaction with Kap121p (Figure 7B; lanes 5 and 6). Interestingly, this is the same region of Abf1p that could mediate moderate nuclear localization in the GFP fusion assay (construct 11 in Figure 3), suggesting that Kap121p/Pse1p interacted with the second Abf1p NLS that was upstream of CS1. In nuclear import, the nuclear RanGTP form binds to karyopherins and facilitates nuclear release of cargo by dissociating the newly imported karyopherin-cargo complex (22). The ability of RanGTP to prevent karyopherin binding to another protein is characteristic of a typical karyopherin–cargo interaction. As shown in Figure 7C, preincubation of Kap121p with physiological levels of RanGTP(10 mM) is sufficient to prevent the subsequent interaction between Kap121p and Abf1p (lanes 3 and 4). This gives further evidence that a direct interaction between the two proteins may be relevant to nuclear Traffic 2004: 5: 925–935

In the current work we demonstrate that a previously identified transcription regulatory region of Abf1p functions as an important signal in mediating effective nuclear import of the protein. The nuclear mislocalization and growth defects caused by an amino acid change in this region of Abf1p can be rescued by an exogenous nuclear localization sequence. Furthermore, the CS1-mediated nuclear import occurs in a Ran-GEF-dependent manner. Finally, we provide genetic and biochemical evidence for a link between Abf1p and the nuclear transport receptor Kap121p. CS1 is apparently not involved in the in vitro interaction between Abf1p and Kap121p, suggesting that another nuclear import factor binds directly to the CS1 region and mediates its nuclear import. This indicates that Abf1p contains at least two functional NLSs. The specific karyopherins that recognize CS1 remain to be elucidated; however, our results suggest that CS1-dependent import may be mediated by several overlapping karyopherin pathways. This is in agreement with previous reports that several other essential proteins in yeast have more than one route into the nucleus (33,34). Nuclear import not only plays a pivotal role in ensuring the function of nuclear proteins at their desired subcellular location, but also provides an important mechanism by which the activities of these proteins are regulated in response to changes in physiological conditions. Given the constitutive presence of Abf1p at ARS1 throughout the cell cycle (35), as well as its relatively high abundance (10,36), CS1 may simply function to sustain a constitutive presence of the protein in the nucleus. However, it remains possible that under certain conditions the nuclear function of Abf1p may be modulated through masking the NLS activity of CS1. In light of our observation that the change of a single residue within CS1 (K625I) is responsible for the gross mislocalization of the protein, it is tempting to speculate that in vivo modification of the same lysine residue (e.g. acetylation) could result in the same signal switch. Our in vitro biochemical data clearly indicate a direct interaction between Kap121p and the central region of Abf1p (aa 323–604). Furthermore, the protein–protein interaction is sensitive to RanGTP, a salient characteristic of cargo– karyopherin binding. Importantly, the same region of Abf1p is capable of mediating modest, but discernible, nuclear localization. These results indicate the presence of a CS1independent NLS that functions through its interaction with Kap121p. However, a ts mutant strain of kap121 does not exhibit obvious defects in the nuclear localization 931

Loch et al.

A

CS1 CS2

DBD 1–731 323–662 323–604

p1 2

32

1

1/10 input

32

+

P( MB

+ + +

–)

Ka

+ +

P–

GST (–) + MBP MBP-Kap121p +

MB

B

3– 6 32 62 3– 6 60 04 4– 66 2

3– 6 32 62 3– 6 60 04 4– 66 2

604–662

MK 198 115 93

49.8

5

4 60

GST

32

32

MBPKap121p

+

+

+ +

+ +

RanGTP

9

3–

66

60

8

32

3–

66 3–

7

32

C

6

2

4

3–

3

4

2

2

1

198

115 93

1

2

3

of wild-type Abf1p at the restrictive temperature (data not shown), suggesting that, like CS1, the central region may also be capable of utilizing multiple karyopherins. It is also interesting to note that the second NLS in the central region of Abf1p did not always appear to be active. For example, whereas Abf1p(1–607) and Abf1p(323–607) both display a moderate degree of nuclear localization (constructs 8 and 11 in Figure 3), Abf1p-(A624-8), which contains an inactive CS1 but a wild-type central region, shows no obvious signs of nuclear localization (construct 4 in Figure 3). One possible explanation is that the C-terminal AD of Abf1p (i.e. aa 604–662) may modulate the weak NLS activity of the central region in an intramolecular manner. In such a scenario, deletion of the entire C-terminal AD may expose the central region and thus liberate the other932

4

Figure 7: Abf1p interacts physically with Kap121p. A) Fragments of Abf1p were fused to the GST tag. B) Anti-MBP immunoblot of the electrophoretically separated, GST pulldown protein samples reveals that MBP-Kap121p binds to GST-Abf1p containing the central region of Abf1p. C) The Kap121p–Abf1p interaction was specifically abrogated by preincubation of MBP-Kap121p with a physiological concentration of RanGTP.

wise masked weak NLS activity. Consistent with this possibility, Abf1p(1–662) with alanine-substitution mutations at CS1 or CS2 caused a lethal phenotype, whereas ABF1(1–607), which lacked the entire C-terminus of the protein, was still capable of supporting cell viability at room temperature (20). The presence of two NLSs would ensure the efficient import of Abf1p, and raises the possibility that each NLS may be differentially regulated. The transcription study clearly indicates that an ectopic NLS only partially rescues the defects of K625I in gene regulation. It is possible that the NLS could not completely reverse the Crm1-mediated nuclear export of this mutant. Another explanation is that CS1 may confer an additional Traffic 2004; 5: 925–935

ABF1 and Nuclear Localization

role in the nucleus, which might be distinct from its function in nuclear localization. This would be consistent with our previous observation that a GAL4–Abf1p fusion bearing the CS1 mutation behaved differently from its wildtype counterpart in transcriptional activation and chromatin remodeling (20), even though the nuclear localization of the fusion protein was fully conferred by the NLS from GAL4. If a second function of CS1 indeed exists in the context of the native Abf1p protein, it may or may not be mechanistically linked to the role of CS1 in nuclear localization. For example, it is possible that the same CS1 region may interact with distinct molecules in the cytoplasm and nucleus to achieve the functions in the separate locations. Alternatively, the putative association of CS1 with the nuclear import machinery and/or nuclear pore complex (NPC) could be responsible for the dual activities of CS1. In this regard, it is worth noting that several nuclear transport proteins and components of NPC are implicated in the organization of special chromatin structure that in turn impacts on gene expression at particular loci in the yeast genome (37,38). Given the functions of Abf1p in mediating diverse nuclear processes and the association of heterochromatin with the nuclear periphery near NPC, a link between CS1 and the nuclear transport system may impart to Abf1p the ability to organize transcriptionally inactive chromatin structure and/or to tether silenced heterochromatin there. A separate pool of nuclear Abf1p that no longer associates with the nuclear transport system may find its way to the euchromatic region of the chromosomes and thus facilitate gene activation and/or firing of early origins of DNA replication.

Materials and Methods Yeast media, transformation, and manipulation Standard techniques were used for experimental manipulation of Saccharomyces cerevisiae (39). Strains and plasmids Strains TMy86 and TMy87 are derivatives of strain W303, and have been described previously (20). TMy86 has the following genotype: MAT a ade2–1 his3–11,15 leu2–3112 trp1–1 ura3–1 can1–100 Dabf1::HIS3MX6 pRS416-ABF1. TMy87 is the same as TMy86 except for MAT a. The mutant prp20–1 strain and its isogenic wild-type strain ACY109 (MAT a, prp20–1, trp1, leu2, ura3) were generous gifts from Dr. Anita Corbett (Emory University, GA). The leptomycin-B sensitive strain, KWy175 (xpo1::LEU2 XPO1(T539C) HIS3 trp1-1 ura3-1 ade2-1) was a kind gift of Dr. Karsten Weis (University of California, Berkeley, CA). pRS415-ABF1 has been described previously (20). Its TRP1-marked derivative was constructed by ligating the BglI-digested, TRP1-containing fragment from pRS414 to the BglI-digested pRS415-ABF1 vector. KAP121 was subcloned into pRS425 by PCR, using the genomic DNA fragment isolated from the multicopy screening as the Traffic 2004: 5: 925–935

template. The ABF1-GFP fusion genes were made by a three-step PCR strategy. First, ABF1 and GFP fragments were amplified separately, each containing common overlapping sequences at the junctions. These two PCR products were then used as the templates in the final PCR to generate the ABF1-GFP fusion gene. The resulting insert was digested with BamHI and XbaI, and inserted into either the BamHI/XbaI-digested pRS415-ABF1 construct harboring the native ABF1 promoter or a pRS414-derived plasmid containing the ADH1 promoter. The GFP constructs bearing exogenous nuclear localization sequence (NLS) were constructed in the same manner as above, except that the 30 primers used for GFP PCR contained the sequence that encodes the NLS from SV40 large T antigen (40) between the GFP coding sequence and the stop codon. GST-ABF1 fusion constructs were made by PCR amplification of ABF1 from pRS415-ABF1, and ligation of the restriction-digested fragments into XhoI/EcoRIdigested pGEX-5X-1 (Amersham, Piscataway, NJ). MBPtagged KAP121/PSE1 was constructed as previously described (33). Construction of a randomly mutagenized plasmid library and temperature-sensitive mutant screening To generate the randomly mutagenized library of the Abf1p C-terminal region, we amplified the DNA fragment that encodes the C-terminal region of Abf1p (aa 604–662) in the presence of MnCl2 (41). The amplified fragment was digested with XbaI and XhoI, and used to replace the corresponding region in pRS415-ABF1(1–662). The resulting library was transformed into TMy86. The shuffled colonies were then screened for temperature sensitivity. One thousand independent colonies were examined and the most severe temperature-sensitive mutants were further characterized. Northern blot analysis Cells that carry different alleles of ABF1 were grown in YPAD medium until optical density (OD)600 reached 1.0– 1.2 at 30 °C. The culture was mixed with an equal volume of 30 °C or 42 °C preheated YPAD to reach the final temperature of 30 °C or 36 °C, respectively. After 45 min of incubation at the indicated temperature, cells were harvested and RNA was prepared as described previously (42). Multicopy suppressor screening The TRP-marked abf1-K625I plasmid was shuffled into TMy86 and the temperature sensitivity of the resulting strain was confirmed. The strain was then transformed with a LEU-marked 2 m plasmid-based yeast genomic DNA library purchased from ATCC (ATCC37323). Plates were incubated at room temperature for 24 h and then moved to the 35 °C restrictive temperature. Plates were monitored closely for the appearance of colonies. Approximately 5000 transformants were screened in the current study. Plasmids were isolated from candidate yeast 933

Loch et al.

colonies and the multicopy plasmids were identified by restriction digestion. The identity of the inserts was revealed by DNA sequencing. Fluorescence microscopy GFP-tagged ABF1 plasmids were transformed into TMy86 and shuffled when they could support viability. Mid-log phase cells were stained with Hoechst 33342 (5 mg/mL) and examined by fluorescence microscopy using a Nikon E800 epifluorescence microscope. The images were captured with a Hamamatsu CCD camera and processed using Openlab 3.0.9 software. Cells were viewed through the appropriate fluorescent filter under a 100
objective. Saved images of at least 150 cells were analyzed for to quantitate the percentage of cells with distinct nuclear GFP localization versus no distinct nuclear localization. For the experiment shown in Figure 6, the GFP constructs were introduced into prp20-1 and isogenic wild type strain (ACY109, a kind gift from Dr. Anita Corbett, Emory University, GA). Protein purification MBP-Kap121p was produced and purified as described (34). GST-Abf1p fusion proteins were induced and purified according to the manufacturer’s instructions. Briefly, IPTG was added to log-phase cells harboring the GST plasmids at a final concentration of 0.2 mM, and the induced cells were grown overnight at  18 °C. The cell pellet was resuspended in 8 mL Buffer A (50 mM Tris pH 8.0, 15 mM EDTA, 2% Triton X-100 w/v, 150 mM NaCl) with protease inhibitors (5 mg/mL leupeptin, 5 mg/mL aprotinin, 1 mg/mL pepstatin A, 1 mg/mL benzamidine, 2 mM PMSF) and 0.1 mg/mL lysozyme. This mixture was then sonicated and pelleted at 14 000 g for 30 min at 4 °C. To the supernatant, 150 mL 50% glutathione-Sepharose beads (prewashed in Buffer A) was added, and the mixture rotated for 1–2 h at room temperature. Following incubation, beads were washed extensively with Buffer A containing 350 mM NaCl (final concentration) and Buffer A with 150 mM NaCl. To elute protein from beads, elution buffer (EB) (100 mM Tris pH 8.0, 150 mM NaCl, 10 mM glutathione, 1 mM dithiothreitol (DTT) and protease inhibitors as in Buffer A) was added to the bead pellet. This slurry was rotated 10 min at 4 °C, and centrifuged at 4000 g for 3 min at 4 °C. The supernatant was then dialyzed overnight against two changes of 500 mL storage buffer, TGED buffer (10 mM Tris pH 8.0, 50% glycerol v/v, 1 mM EDTA, 1 mM DTT). GST pulldown assay Glutathione-Sepharose beads were washed twice in TBT20 (20 mM HEPES, 0.11 M KOAc, 2 mM MgCl2, 1% Tween-20 by vol.) and then blocked in TB-T20 plus 10% BSA by rotation at 4 °C for at least 2 h. Each binding reaction contained 30 mL blocked beads (10 mL beads þ 20 mL buffer), 5 mg of GST fusion proteins, 2 mg MBP fusion proteins, and the total volume was brought to 100 mL with TB-T20. Beads were rotated at 4 °C for 1 h and then washed three times in 1 mL TB-T20 by rotating for 10–15 min per 934

wash. Beads were then washed with TB-T20 containing 10 mM maltose, and finally with 1
phosphate-buffered saline. Proteins from the beads were eluted with 2
SDS-PAGE loading buffer and analyzed by SDS-PAGE. In the assays including RanGTP, 10 mM RanQ69L (a nonhydrolyzable mutant version of human Ran) was incubated with the recombinant Kap for 30 min at 4 °C, before addition of other components of the binding reaction as listed above.

Acknowledgments We thank Drs. K. Weis and A. Corbett for the yeast strains. We also thank Amy Shaw, Katie Dorfler, Ping-Chiao Tsai, and Hongjun Zhong for technical assistance. The work was supported by NIH grants (GM57893 for R.L and GM65385 for L.F.P). N.M. was supported by the NIH Cell and Molecular Biology Training Grant and the Medical Scientist Training Program.

References 1. Fyodorov DV, Kadonaga JT. The many faces of chromatin remodeling: SWItching beyond transcription. Cell 2001;106:523–525. 2. Melendy T, Li R. Chromatin remodeling and initiation of DNA replication. Front Biosci 2001;6:D1048–D1053. 3. Nicolas A. Relationship between transcription and initiation of meiotic recombination: toward chromatin accessibility. Proc Natl Acad Sci USA 1998;95:87–89. 4. Koren A, Ben-Aroya S, Kupiec M. Control of meiotic recombination initiation. a role for the environment? Curr Genet 2002;42:129–139. 5. Morse RH. RAP, RAP, open up! New wrinkles for Rap1 yeast. Trends Genet 2000;16:51–53. 6. Frit P, Kwon K, Coin F, Auriol J, Dubaele S, Salles B, Egly JM. Transcriptional activators stimulate DNA repair. Mol Cell 2002;10:1391–1401. 7. Diffley JF, Stillman B. Similarity between the transcriptional silencer binding proteins Abf1 Rap1. Science 1989;246:1034–1038. 8. Rhode PR, Sweder KS, Oegema KF, Campbell JL. The gene encoding ARS-binding factor I is essential for the viability of yeast. Genes Dev 1989;3:1926–1939. 9. Buchman AR, Kornberg RD. A yeast ARS-binding protein activates transcription synergistically in combination with other weak activating factors. Mol Cell Biol 1990;10:887–897. 10. Della Seta F, Ciafre SA, Marck C, Santoro B, Presutti C, Sentenac A, Bozzoni I. The ABF1 factor is the transcriptional activator of the L2 ribosomal protein genes in Saccharomyces cerevisiae. Mol Cell Biol 1990;10:2437–2441. 11. Galius-Durner V, Xie J, Chintamaneni C, Vershon AK. Participation of the yeast activator ABF1 in meiosis-specific expression of the HOP1 gene. Mol Cell Biol 1996;16:2777–2786. 12. Morrow BE, Johnson SP, Warner JR. The rRNA enhancer regulates rRNA transcription in Saccharomyces cerevisiae Mol Cell Biol 1993;13:1283–1289. 13. Ozsarac N, Bhattacharyya M, Dawes IW, Clancy MJ. The SPR3 gene encodes a sporulation-specific homologue of the yeast CDC3/10/11/12 family of bud neck microfilaments and is regulated by ABF1. Gene 1995;164:157–162. 14. Silve S, Rhode PR, Coll B, Campbell J, Poyton RO. ABF1 is a phosphoprotein and plays a role in carbon source control of COX6 transcription in Saccharomyces cerevisiae. Mol Cell Biol 1992;12:4197–4208. 15. Diffley JF, Stillman B. Purification of a yeast protein that binds to origins of DNA replication and a transcriptional silencer. Proc Natl Acad Sci USA 1988;85:2120–2124. 16. Laurenson P, Rine J. Silencers, silencing, and heritable transcriptional states. Microbiol Rev 1992;56:543–560.

Traffic 2004; 5: 925–935

ABF1 and Nuclear Localization 17. Pryde FE, Louis EJ. Limitations of silencing at native yeast telomeres. EMBO J 1999;18:2538–2550. 18. Cho G, Kim J, Rho HM, Jung G. Structure-function analysis of the DNA binding domain of Saccharomyces cerevisiae ABF1. Nucl Acids Res 1995;23:2980–2987. 19. Li R, Yu DS, Tanaka M, Zheng L, Berger SL, Stillman B. Activation of chromosomal DNA replication in Saccharomyces cerevisiae by acidic transcriptional activation domains. Mol Cell Biol 1998;18:1296–1302. 20. Miyake T, Loch CM, Li R. Identification of a multifunctional domain in autonomously replicating sequence-binding factor 1 required for transcriptional activation, DNA replication, and gene silencing Mol. Cell Biol 2002;22:505–516. 21. Kaffman A, O’Shea EK. Regulation of nuclear localization: a key to a door. Annu Rev Cell Dev Biol 1999;15:291–339. 22. Weis K. Regulating access to the genome: nucleocytoplasmic transport throughout the cell cycle. Cell 2003;112:441–451. 23. Neville M, Rosbash M. The NES-Crm1p export pathway is not a major mRNA export route in Saccharomyces cerevisiae. EMBO J 1999;18:3746–3756. 24. Miyake T, Reese J, Loch CM, Auble DT, Li R. Genome-wide analysis of ARS binding factor 1 (Abf1p)-mediated transcriptional regulation in Saccharomyces cerevisiae. J Biol Chem 2004: in press. 25. Marelli M, Dilworth DJ, Wozniak RW, Aitchison JD. The dynamics of karyopherin-mediated nuclear transport. Biochem Cell Biol 2001;79:603–612. 26. Dasso M. The Ran GTPase: theme and variations. Curr Biol 2002;12:R502–R508. 27. Aebi M, Clark MW, Vijayraghavan U, Abelson J. A yeast mutant, PRP20, altered in mRNA metabolism and maintenance of the nuclear structure, is defective in a gene homologous to the human gene RCC1 which is involved in the control of chromosome condensation. Mol Gen Genet 1990;224:72–80. 28. Pemberton LF, Blobel G, Rosenblum JS. Transport routes through the nuclear pore complex. Curr Opin Cell Biol 1998;10:392–399. 29. Chook YM, Blobel G. Karyopherins and nuclear import. Curr Opin Struct Biol 2001;11:703–715. 30. Kaffman A, Rank NM, O’Shea EK. Phosphorylation regulates association of the transcription factor Pho4 with its import receptor Pse1/ Kap121. Genes Dev 1998;12:2673–2683.

Traffic 2004: 5: 925–935

31. Leslie DM, Grill B, Rout MP, Wozniak RW, Aitchison JD. Kap121pmediated nuclear import is required for mating and cellular differentiation in yeast. Mol Cell Biol 2002;22:2544–2555. 32. Isoyama T, Murayama A, Nomoto A, Kuge S. Nuclear import of the yeast AP-1-like transcription factor Yap1p is mediated by transport receptor Pse1p, and this import step is not affected by oxidative stress. J Biol Chem 2001;276:1863–1869. 33. Mosammaparast N, Ewart CS, Pemberton LF. A role for Nucleosome Assembly Protein-1 in the nuclear transport of histones H2a H2b. EMBO J 2002;21:6527–6538. 34. Mosammaparast N, Guo Y, Shabanowitz J, Hunt DF, Pemberton LF. Pathways mediating the nuclear import of histones H3 and H4 in yeast. J Biol Chem 2002;277:862–868. 35. Dahmann C, Diffley JF, Nasmyth KA. S-phase-promoting cyclindependent kinases prevent re-replication by inhibiting the transition of replication origins to a pre-replicative state. Curr Biol 1995;5:1257–1269. 36. Lee TI, Rinaldi NJ, Robert F, Odom DT, Bar-Joseph Z, Gerber GK, Hannett NM, Harbison CT, Thompson CM, Simon I, Zeitlinger J, Jennings EG, Murray HL, Gordon DB, Ren B, et al. Transcriptional regulatory networks in Saccharomyces cerevisiae. Science 2002;298:799–804. 37. Galy V, Olivo-Marin J, Scherthan H, Doye V, Rascalou N, Nehrbass U. Nuclear pore complexes in the organization of silent telomeric chromatin. Nature 2000;403:108–112. 38. Ishii K, Arib G, Lin C, Van Houwe G, Laemmli UK. Chromatin boundaries in budding yeast: the nuclear pore connection. Cell 2002;109:551–562. 39. Adams A, Gottschling DE, Kaiser CA, Stearns T. Methods in Yeast Genetics. Cold Spring Harbor: Cold Spring Harbor Laboratory Press, 1997: 123–127. 40. Benditt JO, Meyer C, Fasold H, Barnard FC, Riedel N. Interaction of a nuclear localization signal with isolated nuclear envelopes and identification of signal binding proteins by photoaffinity labeling. Proc Natl Acad Sci USA 1989;86:9327–9331. 41. Cadwell RC, Joyce GF. Randomization of genes by PCR mutagenesis. PCR Methods 1992;2:28–33. 42. Iyer V, Struhl K. Absolute mRNA levels and transcriptional initiation rates in Saccharomyces cerevisiae. Proc Natl Acad Sci USA 1996;93:5208–5212.

935