The 14-3-3 proteins positively regulate rapamycin-sensitive signaling

Research Paper

1259

The 14-3-3 proteins positively regulate rapamycin-sensitive signaling Paula G. Bertram, Chenbo Zeng, John Thorson, Andrey S. Shaw and X.F. Steven Zheng Background: The kinase Tor is the target of the immunosuppressive drug rapamycin and is a member of the phosphatidylinositol kinase (PIK)-related kinase family. It plays an essential role in progression through the G1 phase of the cell cycle. The molecular details of Tor signaling remain obscure, however. Results: We isolated two Saccharomyces cerevisiae genes, BMH1 and BMH2, as multicopy suppressors of the growth-inhibitory phenotype caused by rapamycin in budding yeast. BMH1 and BMH2 encode homologs of the 14-3-3 signal transduction proteins. Deletion of one or both BMH genes caused hypersensitivity to rapamycin in a manner that was dependent on gene dosage. In addition, alterations in the phosphopeptide-binding pocket of the 14-3-3 proteins had dramatically different effects on their ability to relieve the growth-arresting rapamycin phenotype. Mutations that prevented 14-3-3 from binding to a phosphoserine motif abolished its ability to confer rapamycin resistance. In contrast, substitution of two residues in 14-3-3 that surround these phosphoserine-binding sites conferred a dominant rapamycin-resistant phenotype.

Address: Department of Pathology, Washington University School of Medicine, St. Louis, Missouri 63110, USA. Correspondence: X.F. Steven Zheng E-mail: [email protected] Received: 10 July 1998 Revised: 26 October 1998 Accepted: 27 October 1998 Published: 5 November 1998 Current Biology 1998, 8:1259–1267 http://biomednet.com/elecref/0960982200801259 © Current Biology Ltd ISSN 0960-9822

Conclusions: Our studies reveal 14-3-3 as an important component in rapamycin-sensitive signaling and provide significant new insights into the structure and function of 14-3-3 proteins.

Introduction The 14-3-3 proteins are a family of highly conserved and abundant proteins found throughout the eukaryotic kingdom. They exist as multiple isoforms and have been implicated as key regulators of diverse intracellular signal transduction processes [1]. The 14-3-3 proteins bind to two different types of 14-3-3-binding motifs, RSXpSXP and RXY/FXpSP (in single-letter amino-acid code where X is any amino acid and pS is a phosphoserine residue) [2–4]. Direct interaction of 14-3-3 with a number of other signaling proteins via these phosphopeptide motifs has been demonstrated recently and shown to be critical for the signaling functions in which 14-3-3 is implicated. There are two Saccharomyces cerevisiae 14-3-3 homologs, Bmh1p and Bmh2p, which share redundant functions for cell viability in most budding yeast strains. Their essential function can be complemented by several plant 14-3-3 isoforms [5,6]. Bmh1p and Bmh2p appear to be involved in signaling from the GTPase Ras via protein kinase A, in vesicular transport [7] and in the Ras–mitogen-activated protein (MAP) kinase cascade that functions during pseudohyphal development [8]. In the Σ1278b genetic background, yeast lacking both BMH1 and BMH2 are viable, but have many abnormalities such as temperaturesensitive growth, accumulation of glycogen and hypersensitivity to environmental stresses [8]. In Schizosaccharomyces pombe there are two 14-3-3 homologs, Rad24 and Rad25,

which were identified as DNA-damage checkpoint proteins [9] and proposed to be essential in a checkpoint involving Chk1 and the phosphatidylinositol kinase (PIK)related kinase Rad3 [10]. Rapamycin is a potent immunosuppressant antibiotic. When complexed with its immunophilin receptor FK506binding protein 12 (FKBP12), rapamycin inhibits progression through the G1 phase of the cell cycle in mammalian cells and yeast. The yeast ‘target of rapamycin’ proteins Tor1p and Tor2p and their mammalian homolog the FKBP–rapamycin-associated protein FRAP (also called RAFT or mTOR) were identified recently [11,12] and are members of the PIK-related kinase family. Despite their sequence similarities to phosphatidylinositol kinases, PIKrelated kinases are thought to function as protein kinases [13,14]. FRAP was recently shown to have a Ser/Thr protein kinase activity in vitro. It can autophosphorylate and can phosphorylate the translation inhibitor eIF4ebinding protein 1 (4EBP1/Phas-1) [15–18] and p70 S6 kinase [18]. The FKBP12–rapamycin complex binds to the FKBP–rapamycin-binding (FRB) domain of Tor proteins [19,20] and inhibits the protein kinase activity of Tor1p (C.Z. and X.F.S.Z., unpublished observations). Treating yeast with rapamycin or deleting both TOR1 and TOR2 causes the cells to growth arrest as well as causing several physiological changes characteristic of starvation,

1260

Current Biology, Vol 8 No 23

including an increase in intracellular glycogen storage and severely reduced protein synthesis [21]. In this study, we show for the first time that 14-3-3 positively regulates rapamycin-sensitive signal transduction.

Results BMH1 and BMH2 as multicopy suppressors of the growthinhibitory phenotype caused by rapamycin

To identify components of the rapamycin-sensitive Tor signaling pathway, we screened for multicopy suppressors of the growth-inhibitory phenotype caused by rapamycin. A YEp13-based 2µ (multicopy) yeast genomic DNA library was introduced into yeast strain EGY48 and selected on medium lacking leucine (SC Leu–) and containing rapamycin; 26 clones were isolated that showed plasmid-dependent suppression of the growth-inhibitory phenotype conferred by rapamycin. All the genomic inserts fell into two related groups (Figure 1a). A common feature of the first group was that all the clones contained the BMH1 open-reading frame (ORF), whereas the inserts in the second group contained an intact BMH2 ORF. The majority of the genomic DNA inserts had different breakpoints, suggesting that they were independent clones. BMH1 and BMH2 are related to each other and are the two S. cerevisiae homologs of the 14-3-3 family. Partial or

complete deletion of BMH1 or BMH2 resulted in complete loss of their ability to confer rapamycin resistance (Figure 1a), indicating that BMH1 and BMH2 are the responsible suppressors. We further found that overexpression of either the BMH1 or BMH2 ORF alone was sufficient to confer rapamycin resistance in yeast (Figure 1b). Together, these experiments indicate that BMH1 and BMH2, when overexpressed, can suppress the growthinhibitory phenotype caused by rapamycin. Lack of one or both BMH genes causes hypersensitivity to rapamycin in yeast

We next tested the rapamycin sensitivity of yeast when the endogenous Bmh1p and Bmh2p are absent as a result of gene disruption. Yeast strains derived from the Σ1278b genetic background were chosen here because loss of both BMH1 and BMH2 alleles does not result in lethality [8]. In low concentrations of rapamycin, wild-type yeast cells were still viable, but grew at a much slower rate. Haploid strains lacking either the BMH1 or the BMH2 allele grew significantly more slowly than the wild-type strains (Figure 1c). We noted consistently that the bmh1 BMH2 strain grew slightly more slowly than the BMH1 bmh2 strain when treated with rapamycin (Figure 1c). This difference became more apparent in diploid strains (data not

Figure 1

(a)

– Rapamycin

(b) Apa1

HindIII

RR17 YER176W BMH1

SacI PvuI EcoNI

NcoI

NcoI

+ Rapamycin

XbaI RapR

PDA1

Vector

+ + – + –

BMH1

TOR1SI

BMH2

KpnI RapR

RR6 BMH2 YDR102C STE5 YDR101W YDR101C

+ + + – + – –

(c) bmh1 BMH1 BMH2 bmh2 BMH1 BMH2

Changes in the expression levels of Bmh1p and Bmh2p alter yeast sensitivity to rapamycin. (a) Identification of BMH1 and BMH2 as multicopy suppressors of the growth-inhibitory phenotype caused by rapamycin. Rapamycin-resistance clones (RR) identified in this study fell into two distinct groups represented by RR17 and RR6, respectively. The arrows show the arrangement of the ORFs and restriction enzyme sites are indicated. RR6 and RR17 were subjected to deletional analysis; the deletion clones were assayed for their rapamycin-resistant phenotype (RapR) by incubation on SC Leu– medium with rapamycin at 30°C. Rapamycin-resistant clones are labeled +; rapamycin-sensitive clones, –. (b) Overexpression of BMH1

– Rapamycin

+ Rapamycin

+ Rapamycin

bmh1 bmh2 2 days

5 days

10 days Current Biology

or BMH2 alone conferred rapamycin resistance. EGY48 yeast strains carrying a vector control (pBF339), or carrying plasmids expressing the rapamycin-resistant mutant Tor1S1972I (TORSI) or hemagglutinin (HA)-tagged Bmh1p (ADH1 promoter, 2µ) or HA–Bmh2p (ADH1, 2µ), were streaked onto SC Trp– plates with or without rapamycin and incubated at 30°C. (c) Disruption of BMH1 and/or BMH2 causes hypersensitivity to rapamycin. Wild-type and bmh haploid strains were streaked onto YPD plates with or without rapamycin and incubated for 2 days without rapamycin, or for 5 and 10 days with 100 nM rapamycin, at 30°C.

Research Paper 14-3-3 in rapamycin-sensitive signaling Bertram et al.

shown). Disruption of both BMH1 and BMH2 alleles resulted in dramatic hypersensitivity to rapamycin (Figure 1c). In fact, both the haploid bmh1 bmh2 and the diploid bmh1 bmh1/bmh2 bmh2 strains showed no signs of growth even when incubated for 10 days. In addition, expression of BMH1 or BMH2 on a centromere plasmid reduced rapamycin sensitivity of bmh1 bmh2 strains (data not shown). Bmh1p and Bmh2p are not involved in general drug resistance in yeast

One possible explanation for the effects of Bmh1p and Bmh2p is that they are involved in general drug resistance. Eukaryotic cells have developed sophisticated mechanisms of detoxification. For example, overexpression of multi-drug resistant (MDR) genes can lead to a dramatic reduction in the sensitivity of yeast to a variety of drugs [22]. To explore this possibility, we examined the effects of overexpressing BMH1 and BMH2 on the sensitivity of yeast to cycloheximide. Like rapamycin, cycloheximide is a potent protein synthesis inhibitor of Streptomyces origin. In contrast to rapamycin, however, overexpression of BMH1 or BMH2 had no detectable effect on cycloheximide sensitivity (Figure 2a).

1261

a general drug-resistant mechanism or by interfering with the binding of the drugs to FKBP12. Although both FK506 and rapamycin bind to the same active site of FKBP12, FKBP12–rapamycin inhibits the Tor family kinases, whereas FKBP12–FK506 inhibits calcineurin, a protein phosphatase necessary for Ca2+-dependent cell signaling. In yeast, calcineurin is required for resistance to high salt stress conditions; it promotes export of excessive Na+ and Li+ ions. Inhibition of calcineurin by FKBP12–FK506 impairs the ability of yeast to resist high concentrations of LiCl [23,24]. We found that overexpression of BMH1 or BMH2 did not have any appreciable effect on yeast sensitivity to FK506 in 0.16 M LiCl (Figure 2b). Calcineurin is also required in yeast for the expression of the PMC1 gene that is involved in tolerance to high Ca2+ [25]. PMC–lacZ reporter constructs have been established to measure calcineurin activity in vivo [26,27]. We found that overexpression of BMH1 or BMH2 did not cause any detectable reduction in the FK506-sensitive lacZ expression (Figure 2c). These results indicate that Bmh1p and Bmh2p do not confer resistance to the FKBP12-binding antibiotics. – CHX

(a)

+ CHX

Vector CHXR BMH2 Vector CHXS

We also tested whether Bmh1p and Bmh2p might be involved in drug resistance to the family of structurally related immunosuppressive antibiotics that bind to FKBP12, which includes rapamycin and FK506. Bmh1p and Bmh2p might confer resistance to this drug family via

BMH1 TOR1SI

Normal

(b)

+ Li+

Li+ + FK506

Figure 2 Vector BMH2

TOR1SI

BMH1

(c) 120 100 lacZ expression (%)

Bmh1p and Bmh2p are not involved in general drug resistance. (a) Overexpression of BMH1 and BMH2 did not cause cycloheximide (CHX) resistance. A cycloheximide-resistant strain Y190 (CHX R) carrying a vector control (pBF339), and cycloheximide-sensitive Y187 strains (CHXS) carrying a vector control (pBF339) or carrying plasmids expressing the rapamycin-resistant mutant Tor1p S1972I (TOR1SI), HA–Bmh1p (ADH1, 2µ) or HA–Bmh2p (ADH1, 2µ), were streaked onto SC Trp– plates with or without cycloheximide (10 µg/ml) and incubated at 30°C. Overexpression of BMH1 or BMH2 conferred rapamycin resistance in Y190 and Y187 under similar conditions (data not shown). (b) Overexpression of BMH1 and BMH2 did not confer resistance to FK506 and LiCl. EGY48 strains carrying a vector control (pBF339), or plasmids expressing the rapamycin-resistant Tor1p S1972I, HA–Bmh1p (ADH1, 2µ), or HA–Bmh2p (ADH1, 2µ) were incubated on a SC Trp– plate until colonies appeared. Replicas were made onto YPD, YPD with 0.16 M LiCl, or YPD with 0.16 M LiCl and 1 µM FK506, and incubated at 30°C (we found that SC Trp– plates with LiCl were too soft for these experiments). In a similar replica experiment, HA–Bmh1p (ADH1, 2µ) and HA–Bmh2p (ADH1, 2µ) conferred rapamycin resistance when plated on YPD with rapamycin (data not shown). (c) Overexpression of BMH1 and BMH2 did not cause a reduction in FK506-sensitive Ca 2+-inducible transcription. EGY48 strains carrying pKC190 and a vector control (pBF339), or a plasmid expressing HA–Bmh1p (ADH1, 2µ) or HA–Bmh2p (ADH1, 2µ) were cultured in SC Leu– Trp– at 30°C in the presence or absence of 200 nM FK506 and 100 mM CaCl2. The Ca2+-responsive lacZ expression levels shown are the mean results from three independent experiments.

80 60 40 20 0 FK506 CaCl2

+ + – – – + –+ Control

+ + – – –+ –+ BMH1

+ + – – –+ – + BMH2 Current Biology

1262

Current Biology, Vol 8 No 23

Bmh1p and Bmh2p do not directly relieve Tor1p from inhibition by FKBP12–rapamycin

We have recently shown that Tor1 has an intrinsic Ser/Thr protein kinase activity that uses mammalian 4EBP-1 as an artificial substrate (C.Z. and X.F.S.Z., unpublished observations). FKBP12–rapamycin potently inhibits Tor1p kinase activity both in vivo and in vitro. Bmh1p and Bmh2p could antagonize rapamycin inhibition by activating Tor1p kinase activity or by directly preventing binding of FKBP12–rapamycin to Tor1p or Tor2p. We found that overexpression of BMH1 failed to relieve Tor1p from inhibition by FKBP12–rapamycin at any detectable level in vivo. In contrast, Tor1pS1972I, a rapamycin-resistant mutant of Tor1p that does not interact with FKBP12–rapamycin, showed significant rapamycin-resistant kinase activity

Mammalian 14-3-3 genes can also confer a rapamycinresistant phenotype in yeast

Figure 3

Kinase activity (%)

(a) 100 80 60

Vector BMH1 TOR1SI

40 20 0 0

(b)

0 .1 .5 1

1 10 Rapamycin (nM) 5 0 .1 .5

1

100

5

Bmh1p (µM) Rapamycin HA–Tor1p 4EBP1

(c)

(Figure 3a). In addition, incubation of cell-free yeast extracts with bacterially expressed recombinant glutathione-S-transferase (GST)–Bmh1p or GST–Bmh2p before the extracts were treated with rapamycin did not affect the inhibition of Tor1p protein kinase activity by rapamycin (Figure 3b,c). These observations suggest that Bmh1p and Bmh2p do not directly antagonize the inhibition of Tor1p and Tor2p by FKBP12–rapamycin. In addition, we have not been able to obtain any evidence that Tor1p directly interacts with Bmh1p or Bmh2p (data not shown), which is consistent with the lack of consensus 143-3-binding sites in either Tor1p or Tor2p. Thus, Bmh1p and Bmh2p are unlikely to act directly on Tor1p or Tor2p to relieve rapamycin inhibition.

Bmh1p Bmh2p GST Rapamycin HA–Tor1p

The 14-3-3 proteins exist in multiple isoforms and are conserved across species. The yeast and higher eukaryotic 14-3-3 proteins have highly conserved phosphoserineligand-binding pockets. Several plant 14-3-3 isoforms have been shown to complement the essential functions of BMH1 and BMH2 in certain yeast strains [5,6]. Conversely, Bmh1p and Bmh2p can regulate human c-Raf-1 activity in yeast [28]. We asked whether the rapamycinsuppressing activity of 14-3-3 proteins is isoform-dependent or species-dependent. Three human 14-3-3 isoforms (β, τ and η), when overexpressed, were sufficient to confer rapamycin resistance at a comparable level to Bmh1p and Bmh2p (Figure 4). Thus, the rapamycin-sensitive function of 14-3-3 proteins is conserved across species and appears to be isoform-independent. These results also suggest that 14-3-3 proteins have similar roles in Tor signaling in higher eukaryotes. The phosphoserine-ligand-binding pocket is crucial for 14-3-3 function in rapamycin-sensitive signaling

Direct binding to conserved phosphoserine motifs is critical for the functions of 14-3-3 proteins in regulating diverse Figure 4

4EBP1 – Rapamycin

Current Biology

Bmh1p and Bmh2p fail to relieve Tor1p from inhibition by FKBP12–rapamycin. (a) EGY48 strains containing a vector control, overexpressing Bmh1 (2µ), or expressing HA–Tor1pS1972I as a positive control, were incubated in SC Leu– Trp– with or without different concentrations of rapamycin for 1 h. HA–Tor1p was immunoprecipitated and its protein kinase activity assayed. A typical result from several experiments is shown. (b) A cell-free extract prepared from EGY48 expressing HA–Tor1p was incubated first with Bmh1p at different concentrations for 30 min then with 25 nM rapamycin for another 30 min. HA–Tor1p was immunoprecipitated with monoclonal antibody 12CA5 and assayed for its protein kinase activity towards 4EBP1. (c) The same as (a) except that the cell lysates were preincubated with 0.5 µM of each of GST, Bmh1p or Bmh2p.

+ Rapamycin

Vector τ

η β Current Biology

Overexpression of three mammalian 14-3-3 isoforms confers a rapamycin-resistant phenotype in yeast. EGY48 strains carrying a control vector (BF339), or overexpressing 14-3-3β, η or τ were streaked onto SC Trp– plates with or without rapamycin and incubated at 30°C.

Research Paper 14-3-3 in rapamycin-sensitive signaling Bertram et al.

1263

Table 1 Mutations in the ligand-binding pocket affect the ability of 14-3-3h to confer a dominant rapamycin-resistant phenotype. 14-3-3η Binding to phosphopeptide Rapamycin-resistant phenotype

C

WT

R56A/R60A

V51A

R55A

R132K

E185K

Y216F

n/a

++++

+/–

++++

++++

–

+++

+++

–

++++

+/–

++++

++++

–

–

+++

A summary of the abilities of wild-type (WT) and mutant 14-3-3η proteins to bind to a phosphopeptide (RSApSEP) derived from Raf-1 as described by Thorson et al. [29], and to confer rapamycin resistance (Figure 5a). The relative activities of different 14-3-3η

proteins to bind to RSApSEP and to confer rapamycin resistance are indicated by the number of + symbols. C indicates control vector; n/a, not applicable.

signaling proteins [2,3]. The 14-3-3 protein forms a large, negatively charged pocket, the interior of which contains amino acids that are almost invariant throughout the family. The three-dimensional 14-3-3ζ–phosphopeptide structure provides details of the interaction between this conserved pocket and a phosphoserine ligand [3]. The phosphate is held by an array of interactions involving the side chains of amino-acid residues K49 and R56 from helix C and Y128 and R127 from helix E. Mutations in the ligand-binding pocket of 14-3-3η showed dramatically different affinities for RSApSEP, a phosphopeptide ligand derived from Raf-1 [29]. The 14-3-3 mutants were overexpressed in yeast and assayed for their ability to confer a rapamycin-resistant phenotype; their behavior is summarized in Table 1. Mutations R56A/R60A and R132K abolished the ability of 14-3-3 to bind the phosphoserine peptide. These mutant 14-3-3 proteins also failed to confer the rapamycin-resistant phenotype (Figure 5a). In contrast, three other mutations, V51A,

R55A and Y216F, did not affect phosphopeptide binding and these mutant 14-3-3 proteins conferred the rapamycinresistant phenotype. A third type of mutation we examined was E185K. This mutation was originally identified as a suppressor of the activated RAS allele during Drosophila eye development [30]. E185 corresponds to E183 in 14-33ζ. Although the E185K 14-3-3 mutant was only slightly defective in binding to the Raf-1 phosphopeptide, it was completely inactive in conferring rapamycin resistance (Table 1, Figure 5a). Further investigation is needed to understand the nature of the effect of this mutant. The different 14-3-3η mutants were expressed in yeast at a similar level, suggesting that the differences in their activities in vivo were not due to differences in their stability or expression level (Figure 5b). Three BMH1 mutations, L232S (bmh1-1), G55D (bmh1-2) and A59T (bmh1-3), were recently described as partial

Figure 5

– Rapamycin

(a)

+ Rapamycin

C Y216F

WT R56A/ R60A

E185K R132K

V51A R55A

(b)

R5

5A R1 32 K E1 85 K Y2 16 F

R5

1A V5

0A R6

6A /

W

T

Myc–14-3-3η

C

Different mutations affect the ability of 14-3-3η to confer a rapamycin-resistant phenotype. (a) EGY48 strains carrying a vector control (C; pAUD6-2) or carrying plasmids overexpressing wild-type (WT) or mutant 14-3-3η proteins (ADH1, 2µ) as indicated were streaked onto SC Ura– plates with or without rapamycin and incubated at 30°C. (b) Western blotting analysis of the wild-type and mutant Myc-epitope-tagged 14-3-3η proteins from the yeast used in (a).

Current Biology

1264

Current Biology, Vol 8 No 23

Figure 6 – Rapamycin

+ Rapamycin

BMH1 L232S (ARS/CEN) (ARS/CEN) BMH1 (2µ)

G55D (ARS/CEN) Current Biology

loss-of-function mutations that completely abolish the activity of the MAP kinase signaling cascade that functions during pseudohyphal development [8]. The same mutations do not appear to affect other functions of Bmh1p, however [8]. These mutated residues correspond to L230, G53 and A57, respectively, in 14-3-3ζ, based on sequence alignment. As these mutations lie in the ligand-binding pocket, we also examined their effects on rapamycin-sensitive signaling. To our surprise, a single copy of BMH1L232S or BMH1G55D (on ARS/CEN plasmids) conferred strong rapamycin resistance (Figure 6). In contrast, a single copy of wild-type BMH1 (Figure 6) or BMH1A59T (data not shown) did not confer rapamycin resistance. All three BMH1 mutants and the wild-type BMH1 were carried on the same centromere plasmid vector and were expressed at similar levels under the control of the BMH1 promoter [8] (see Supplementary material published with this paper on the internet). Thus, L232S and G55D appear to be gainof-function mutations that enhance the activity of 14-3-3 in rapamycin-sensitive signaling.

Discussion In this study we have isolated BMH1 and BMH2, two yeast 14-3-3 homologs, as multicopy suppressors of the growth-inhibitory phenotype caused by rapamycin. Interestingly, BMH1 and BMH2 were repeatedly isolated and were the only strong suppressors of the rapamycin phenotype, suggesting that they play a critical role in rapamycinsensitive signaling. Loss of BMH1 and BMH2 individually or in combination led to hypersensitivity to rapamycin. Overexpressing BMH1 or BMH2 did not cause decreased sensitivity of yeast to antibiotics in general or to antibiotics related to FK506 that bind to FKBP12. Bmh1p and Bmh2p did not appear to directly interact with Tor family proteins or to relieve rapamycin inhibition in vivo or in vitro. In addition, the L232S and G55D mutations in BMH1 were sufficient to confer dominant rapamycin resistance. These results demonstrate that 14-3-3 is a key component in rapamycin-sensitive signaling. The 14-3-3 proteins constitute a highly conserved family with multiple isoforms [1]. Our results show that Bmh1p and Bmh2p are redundant in rapamycin-sensitive signaling, and that at least three mammalian isoforms (β, τ and ζ) can complement such a function. Thus, different

Dominant gain-of-function mutations in BMH1 confer a strong rapamycin-resistant phenotype. EGY48 strains overexpressing Bmh1p (from a 2µ plasmid), or expressing Bmh1p, Bmh1pL232S or Bmh1pG55D at low levels (from an ARS/CEN plasmid) were streaked onto SC Leu– plates with or without rapamycin and incubated at 30°C.

14-3-3 proteins share a common function in conferring rapamycin resistance. Given that the rapamycin-sensitive pathway is well conserved, Tor from higher eukaryotes may also use 14-3-3 in its signal transduction pathway. Substitution of invariant residues in the conserved ligandbinding pocket of 14-3-3 showed dramatically different effects on the ability of 14-3-3 to bind to a Raf-1 phosphoserine peptide. Residues R56, R60 and R132 directly interact with the phosphate of the Raf-1 peptide [3]. Mutations R56A/R60A or R132K abolished the ability of 14-3-3 to bind to the Raf-1 peptide. When overexpressed, these mutant proteins also failed to confer a rapamycinresistant phenotype. In contrast, residues V51, R55 and Y216 are not directly involved in substrate interaction. They do not individually affect substrate binding or the ability of 14-3-3 to confer a rapamycin-resistant phenotype. These results demonstrate that binding to a phosphoserine motif(s) is critical for 14-3-3 to mediate Tor signaling. One particularly interesting mutation is E185K, which was originally isolated from Drosophila as a dominant-negative mutation in Ras–Raf signaling during eye development [30]. Although the E185K mutant could bind to the Raf-1 peptide at near wild-type affinity, it failed to confer a rapamycin-resistant phenotype. In contrast, Y216F, another mutation isolated in the same screen, did not affect the ability of 14-3-3 to confer rapamycin resistance. One interpretation for these conflicting observations is that 14-3-3 recognizes a slightly different phosphoserine motif in the two distinct pathways. In a mammalian transfection experiment, Y216F behaved as a dominant-negative in Ras–Raf-1 signaling, whereas E185K did not [29]. We have identified two dominant gain-of-function mutations, L232S and G55D, in Bmh1p. Either mutant can confer a rapamycin-resistant phenotype when expressed as a single copy gene. The effects of these mutations are highly selective. In fact, both mutations were originally isolated as loss-of-function mutations in Ras–MAP kinase signal transduction during pseudohyphal development [8]. How is it possible that these mutations cause dominant gain-of-function in one pathway but loss-of-function in another? Both L232 and G55 are localized in the conserved ligand-binding pocket. One mechanistic model is shown in Figure 7. In this model, residue S232 creates a

Research Paper 14-3-3 in rapamycin-sensitive signaling Bertram et al.

1265

Figure 7 A model for the role of 14-3-3 in rapamycinsensitive signaling. TEF is an effector (or ligand) in the rapamycin-sensitive signaling pathway and is shown with or without a phosphate group (P) on the serine (S) of the 14-3-3 binding motif. The dashed line on the lower right-hand TEF protein indicates a potential new hydrogen bond between S232 of Bmh1p (14-3-3) and an asparagine residue (N) adjacent to the phosphoserine of the ligand molecule. PPase indicates a protein phosphatase that removes the phosphate from TEF, and rap indicates rapamycin. The number of TEF molecules (small or normal numbers) in the cell are indicated. See text for further details.

Tor1p ATP ADP S TEF

Inactive

PPase Pi

14-3-3 L232

P S TEF

S TEF

Active

Normal interaction

Tor1p ATP ADP

14-3-3 L232 P S TEF

S TEF

Active

PPase

FKBP12–rap Tor1p ADP ATP S TEF

Inactive

PPase

Active (small numbers)

Weak signaling Weak growth

(small numbers)

14-3-3 S232 P S TEF

14-3-3 L232 P S TEF

Normal interaction

(small numbers)

Pi

Normal signaling Normal growth

(normal numbers)

(normal numbers) FKBP12–rap

Inactive

14-3-3 L232 P

14-3-3 S232 P N S TEF

Normal signaling Normal growth

High-affinity interaction (normal numbers) Current Biology

Pi

new hydrogen bond with a residue (for example N or Q) that is adjacent to the phosphoserine in the ligand and in close proximity to S232 in the three-dimensional structure of the complex. This new hydrogen bond helps to stabilize the 14-3-3–ligand interaction and enhances Tor signaling. In contrast, it might not affect or might even weaken the interaction with another phosphopeptide with a different residue in the same position (for example the Raf-1 ligand). Given that different 14-3-3 ligands have slightly different sequences, pathway-specific mutant alleles can be useful in dissecting the roles of 14-3-3 in diverse signaling pathways. How do 14-3-3 proteins participate in the rapamycin-sensitive signaling pathway? They might directly interact with a substrate of Tor1p or Tor2p, or they might bind to a phosphoprotein that is a substrate of another protein kinase further downstream. Phosphorylation of this effector of Tor1p or Tor2p is important for successful Tor signaling. Binding of 14-3-3 to the phosphorylation site has been shown to be important for protection of the labile phosphate of the ligand, which maintains the ligand in the correct active state [31,32]. As FKBP12–rapamycin only partially inhibits Tor1p protein kinase activity with maximal inhibition at 70%, one could imagine that overexpression of 14-3-3 might further stabilize the 14-33–ligand complex and lead to accumulation of the phosphorylated effector to a normal physiological concentration. Thus, enough signal could be generated even in the presence of rapamycin to confer rapamycin-resistant growth (Figure 7). Altogether, our study has defined a new role for 14-3-3 in the context of Tor signaling and provided new insights into the structure and function of 14-3-3 proteins.

Materials and methods Yeast strains, library screen and plasmids Yeast strains are listed in Table 2. Standard yeast culture medium was prepared. Various drugs were added as indicated to test yeast sensitivity. A YEp13-based yeast genomic library DNA (from K. Nasmyth) was transformed into EGY48 and plated on SC Leu– plates containing 0.1 µg/ml rapamycin, and incubated at 30°C for 4 days. Large colonies were picked for further analysis. To construct plasmids pBF-BMH1 and pBF-BMH2, DNA fragments flanking the BMH1 and BMH2 ORFs were amplified from RR17 and RR6, respectively, using PCR. BMH1 was amplified using oligonucleotides 5′-CGGGATCCTGTCAACCAGTCGTGAAGA-3′ and 5′-CGGGATCCTTACTTTGGTGCTTCACCTT-3′, and BMH2 was amplified using 5′-CGGGATCCTGTCCCAAACTCGTGAAGA-3′ and 5′-CGGGATCCTTATTTGGTTGGTTCACCTT-3′. PCR products were cloned separately into pBF339. In pBF339, BMH1 and BMH2 are under the control of the ADH1 promoter and were fused with an in-frame HA epitope. DNA sequencing was carried out to confirm that no mutations were generated in the PCR-amplified BMH1 and BMH2. To construct plasmids pGEX-BMH1 and pGEX-BMH2, BMH1 and BMH2 were excised with BamH1, and cloned into the BamH1 sites of pGEX-3X (Pharmacia). For the pAUD6-human 14-3-3 expression plasmids, human 14-33β, η and τ were excised from their pGEX-KT vector with BamH1 and EcoRI and cloned into pAUD6 [33]. These genes are under the control of the ADH1 promoter and are tagged with a c-Myc epitope at their amino termini. To construct the 14-3-3η mutants in pAUD6, V51A, R55A, R56A/R60A, R132K, E185K and Y216F were excised from their pGEX-KT vectors with BamH1 and EcoR1 and cloned into pAUD6. The plasmids carrying bmh1-1 (L232S), bmh1-2 (G55D) and bmh1-3 (A59T) mutants were generous gifts from G. Fink. PKC190 was a generous gift from K. Cunningham. BF339 (HA epitope, ADH1, 2µ, Trp1) was a generous gift from Frank Li. Plasmids expressing HA–Tor1p and HA–Tor1p (S1972I) were previously described [20].

GST fusion proteins, western blot analysis, and β-galactosidase and protein kinase assays GST–Bmh1p, GST–Bmh2p, GST–FKBP12 and GST–4EBP1 were produced in Escherichia coli BL21(DE3) and purified according to the manufacturer’s protocol (Pharmacia). Yeast cells were lysed with glass beads in disruption buffer (DB; 50 mM Tris-HCl pH 7.5, 250 mM NaCl, 5 mM EDTA, 1% NP-40 plus protease inhibitors, 1 mM PMSF and 5 µg/ml

1266

Current Biology, Vol 8 No 23

Table 2 Yeast strains used in this study. Strain

Genotype

Source

EGY48

MATα his3, trp1, ura3, 6LexAop-LEU2

[34]

Y190

MATa ura3-52 leu2-3 trp1-901 his3-200 ade2-101 gal4∆ Cyhr2 LYS2::GAL1 URA3::GAL1LacZ

[35]

Y187

MATα ura3-52 leu2-3 trp1-901 his3-200 ade2-101 gal4∆ Met- gal80∆ URA3::GAL1LacZ

[35]

10560-4a

MATa ura3-52 leu2 trp1 his3

[8]

RRY1063

MATa bmh1::HIS3 ura3-52 his3 leu2

[8]

RRY1266

MATa bmh2::HIS3 ura3-52 his3 leu2

[8]

RRY1216

MATa bmh1::HIS3 bmh2::HIS3 ura3-52 his3 leu2

[8]

L5543

MATa/α ura3-52 his3 trp1

[8]

RRY1064

MATa/α bmh1::HIS3 ura3-52 his3

[8]

RRY1276

MATa/α bmh2::HIS3 ura3-52 his3

[8]

leupeptin). A total of 200 µg protein was used for western blot analysis with anti-HA (12CA10) and anti-Myc (9EC10) monoclonal antibodies. A lacZ reporter plasmid (pKC190) was used to measure yeast sensitivity to FK506. The culture conditions and β-galactosidase activity assays were carried out as described [26]. Yeast expressing HA–Tor1p [20] were grown to OD600 = 1 and were lysed as described above. For immunoprecipitation, 5 µg 12CA5 monoclonal antibodies were incubated with cell lysates for 1 h at 4°C. The immunocomplexes were then captured by protein-A–Sepharose beads (Pharmacia), washed three times with DB and once with kinase buffer (KB; 20 mM Hepes, 10 mM MnCl2, pH 7.5), and incubated with KB containing 1 µg 4EBP1 and 20 µM [γ-32P]ATP for 30 min at 30°C. The phosphorylated 4EBP1 was separated on an SDS–polyacrylamide gel and analysed with a BioRad phosphoimager.

Supplementary material A figure showing the expression levels of wild-type and mutant Bmh1p is published with this article on the internet.

Acknowledgements We thank R. Braichmann, D. Dean, A. Muslin and H. Piwnica-Worms for critically reading this manuscript; R. Braichmann, R. Brent, K. Cunningham, M. Cyert and G. Fink for generously providing plasmids and yeast strains; and M. Yaffe for insightful information on 14-3-3ζ:phosphopeptide structure. This work was supported by a Howard Hughes New Investigator Award and a start-up fund from Washington University School of Medicine (X.F.S.Z.). X.F.S.Z. was a Coleman Foundation Scholar.

References 1. Aitken A: 14-3-3 proteins on the MAP. Trends Biochem Sci 1995, 20:95-97. 2. Muslin AJ, Tanner JW, Allen PM, Shaw AS: Interaction of 14-3-3 with signaling proteins is mediated by the recognition of phosphoserine. Cell 1996, 84:889-897. 3. Yaffe MB, Rittinger K, Volinia S, Caron PR, Aitken A, Leffers H, et al.: The structural basis for 14-3-3:phosphopeptide binding specificity. Cell 1997, 91:961-971. 4. Petosa C, Masters SC, Bankston LA, Pohl J, Wang B, Fu H, et al.: 14-3-3zeta binds a phosphorylated Raf peptide and an unphosphorylated peptide via its conserved amphipathic groove. J Biol Chem 1998, 273:16305-16310. 5. van Heusden GP, Griffiths DJ, Ford JC, Chin AWTF, Schrader PA, Carr AM, et al.: The 14-3-3 proteins encoded by the BMH1 and BMH2 genes are essential in the yeast Saccharomyces cerevisiae and can be replaced by a plant homologue. Eur J Biochem 1995, 229:45-53.

6. van Heusden GP, van der Zanden AL, Ferl RJ, Steensma HY: Four Arabidopsis thaliana 14-3-3 protein isoforms can complement the lethal yeast bmh1 bmh2 double disruption. FEBS Lett 1996, 391:252-256. 7. Gelperin D, Weigle J, Nelson K, Roseboom P, Irie K, Matsumoto K, et al.: 14-3-3 proteins: potential roles in vesicular transport and Ras signaling in Saccharomyces cerevisiae. Proc Natl Acad Sci USA 1995, 92:11539-11543. 8. Roberts RL, Mosch HU, Fink GR: 14-3-3 proteins are essential for RAS/MAPK cascade signaling during pseudohyphal development in S. cerevisiae. Cell 1997, 89:1055-1065. 9. Ford JC, al-Khodairy F, Fotou E, Sheldrick KS, Griffiths DJ, Carr AM: 14-3-3 protein homologs required for the DNA damage checkpoint in fission yeast. Science 1994, 265:533-535. 10. Zeng Y, Forbes K, Wu Z, Moreno S, Piwnica-Worms H, Enoch T: Replication checkpoint requires phosphorylation of the phosphatase Cdc25 by Cds1 or Chk1. Nature 1998, 395:507-510. 11. Brown EJ, Schreiber SL: A signaling pathway to translational control. Cell 1996, 86:517-520. 12. Thomas G, Hall MN: TOR signalling and control of cell growth. Curr Opin Cell Biol 1997, 9:782-787. 13. Hunter T: When a lipid kinase is not a lipid kinase? When a lipid kinase is a protein kinase? Cell 1995, 83:1-4. 14. Keith CT, Schreiber SL: PIK-related kinases: DNA repair, recombination, and cell cycle checkpoints. Science 1995, 270:50-51. 15. Brown EJ, Beal PA, Keith CT, Chen J, Shin TB, Schreiber SL: Control of p70 s6 kinase by kinase activity of FRAP in vivo. Nature 1995, 377:441-446. 16. Brunn GJ, Williams J, Sabers C, Wiederrecht G, Lawrence JC Jr, Abraham RT: Direct inhibition of the signaling functions of the mammalian target of rapamycin by the phosphoinositide 3-kinase inhibitors, wortmannin and LY294002. EMBO J 1996, 15:5256-5267. 17. Brunn GJ, Fadden P, Haystead TAJ, Lawrence JC Jr: The mammalian target of rapamycin phosphorylates sites having a (Ser/Thr)-Pro motif and is activated by antibodies to a region near its COOH terminus. J Biol Chem 1997, 272:32547-32550. 18. Burnett PE, Barrow RK, Cohen NA, Snyder SH, Sabatini DM: RAFT1 phosphorylation of the translational regulators p70 S6 kinase and 4E-BP1. Proc Natl Acad Sci USA 1998, 95:1432-1437. 19. Stan R, McLaughlin MM, Cafferkey R, Johnson RK, Rosenberg M, Livi GP: Interaction between FKBP12-rapamycin and TOR involves a conserved serine residue. J Biol Chem 1994, 269:32027-32030. 20. Zheng XF, Florentino D, Chen J, Crabtree GR, Schreiber SL: TOR kinase domains are required for two distinct functions, only one of which is inhibited by rapamycin. Cell 1995, 82:121-130.

Research Paper 14-3-3 in rapamycin-sensitive signaling Bertram et al.

21. Barbet NC, Schneider U, Helliwell SB, Stansfield I, Tuite MF, Hall MN: TOR controls translation initiation and early G1 progression in yeast. Mol Biol Cell 1996, 7:25-42. 22. Balzi E, Goffeau A: Genetics and biochemistry of yeast multidrug resistance. Biochim Biophys Acta 1994, 1187:152-162. 23. Nakamura T, Liu Y, Hirata D, Namba H, Harada S, Hirokawa T, et al.: Protein phosphatase type 2B (calcineurin)-mediated, FK506sensitive regulation of intracellular ions in yeast is an important determinant for adaptation to high salt stress conditions. EMBO J 1993, 12:4063-4071. 24. Breuder T, Hemenway CS, Movva NR, Cardenas ME, Heitman J: Calcineurin is essential in cyclosporin A- and FK506-sensitive yeast strains. Proc Natl Acad Sci USA 1994, 91:5372-5376. 25. Cunningham KW, Fink GR: Ca2+ transport in Saccharomyces cerevisiae. J Exp Biol 1994, 196:157-166. 26. Matheos DP, Kingsbury TJ, Ahsan US, Cunningham KW: Tcn1p/Crz1p, a calcineurin-dependent transcription factor that differentially regulates gene expression in Saccharomyces cerevisiae. Genes Dev 1997, 11:3445-3458. 27. Stathopoulos AM, Cyert MS: Calcineurin acts through the CRZ1/TCN1-encoded transcription factor to regulate gene expression in yeast. Genes Dev 1997, 11:3432-3444. 28. Irie K, Gotoh Y, Yashar BM, Errede B, Nishida E, Matsumoto K: Stimulatory effects of yeast and mammalian 14-3-3 proteins on the Raf protein kinase. Science 1994, 265:1716-1719. 29. Thorson J, Yu L, Hsu A, Shih N, Graves P, Tanner JW, et al.: 14-3-3 proteins are required for maintenance of Raf-1 phosphorylation and kinase activity. Mol Cell Biol 1998, 18:5229-5238. 30. Chang HC, Rubin GM: 14-3-3 epsilon positively regulates Rasmediated signaling in Drosophila. Genes Dev 1997, 11:1132-1139. 31. Peng CY, Graves PR, Thoma RS, Wu Z, Shaw AS, Piwnica-Worms H: Mitotic and G2 checkpoint control: regulation of 14-3-3 protein binding by phosphorylation of Cdc25C on serine-216. Science 1997, 277:1501-1505. 32. Zha J, Harada H, Yang E, Jockel J, Korsmeyer SJ: Serine phosphorylation of death agonist BAD in response to survival factor results in binding to 14-3-3 not BCL-X(L). Cell 1996, 87:619-628. 33. Marcus S, Polverino A, Barr M, Wigler M: Complexes between STE5 and components of the pheromone-responsive mitogen-activated protein kinase module. Proc Natl Acad Sci USA 1994, 91:7762-7766. 34. Gyrus j, Golemis E, Cherkov H, Brent R: Cdi1, a human G1 and S phase protein phosphatase that associates with Cdk2. Cell 1993, 75:791-803. 35. Harper JW, Adami GR, Wei N, Keyomarsi K, Elledge SJ: The p21 Cdk2-interacting protein Cip1 is a potent inhibitor of G1 cyclindependent protein kinases. Cell 1993, 75:805-816.

Because Current Biology operates a ‘Continuous Publication System’ for Research Papers, this paper has been published on the internet before being printed. The paper can be accessed from http://biomednet.com/cbiology/cub — for further information, see the explanation on the contents page.

1267

S1

Supplementary material The 14-3-3 proteins positively regulate rapamycin-sensitive signaling Paula G. Bertram, Chenbo Zeng, John Thorson, Andrey S. Shaw and X.F. Steven Zheng Current Biology 9 November 1998, 8:1259–1267 Figure S1

D 55

G

1

H

r to

S 32 L2

M

h1 bm

h1 bm

B

c Ve

BMH1 bmh1 bmh1 bmh1 bmh1 bmh1 bmh1 BMH2 bmh2 BMH2 bmh2 bmh2 bmh2 bmh2 Bmh2p Bmh1p

1

2

3

4

5

6

7

Current Biology

Expression of wild-type and mutant Bmh1p and Bmh2p. Levels of Bmh1p and Bmh2p proteins were analyzed by western blotting with an affinity-purified anti-human 14-3-3η polyclonal antibody [29]. Lane 1, yeast strain BMH1 BMH2; lane 2, bmh1 bmh2; and lane 3, bmh1 BMH2. Lanes 4–7, bmh1 bmh2 strains transformed with the following centromere plasmids: control vector, BMH1, bmh1L232S and bmh1G55D.