Does the Large Number of Isoforms in Multicellular ... - Springer Link

J Mol Evol (2000) 51:446–458 DOI: 10.1007/s002390010107

© Springer-Verlag New York Inc. 2000

Evolution of the 14-3-3 Protein Family: Does the Large Number of Isoforms in Multicellular Organisms Reflect Functional Specificity? Magnus Rosenquist,1 Paul Sehnke,2 Robert J. Ferl,2 Marianne Sommarin,1 Christer Larsson1 1 2

Department of Plant Biochemistry, Lund University, P.O. Box 117, SE-221 00 Lund, Sweden Program in Plant Molecular and Cellular Biology, Department of Horticultural Sciences, University of Florida, Gainesville, FL 32611, USA

Received: 19 April 2000 / Accepted: 24 July 2000

Abstract. 14-3-3 proteins constitute a family of eukaryotic proteins that are key regulators of a large number of processes ranging from mitosis to apoptosis. 143-3s function as dimers and bind to particular motifs in their target proteins. To date, 14-3-3s have been implicated in regulation or stabilization of more than 35 different proteins. This number is probably only a fraction of the number of proteins that 14-3-3s bind to, as reports of new target proteins have become more frequent. An examination of 14-3-3 entries in the public databases reveals 153 isoforms, including alleloforms, reported in 48 different species. The number of isoforms range from 2, in the unicellular organism Saccharomyces cerevisiae, to 12 in the multicellular organism Arabidopsis thaliana. A phylogenetic analysis reveals that there are four major evolutionary lineages: Viridiplantae (plants), Fungi, Alveolata, and Metazoa (animals). A close examination of the aligned amino acid sequences identifies conserved amino acid residues and regions of importance for monomer stabilization, dimer formation, target protein binding, and the nuclear export function. Given the fact that 53% of the protein is conserved, including all amino acid residues in the target binding groove of the 14-3-3 monomer, one might expect little to no isoform specificity for target protein binding. However, using surface plasmon resonance we show that there are large differences in affinity between nine 14-3-3 isoforms of A. thaliana and a target peptide representing a novel binding motif pre-

Correspondence to: M. Rosenquist; e-mail: magnus.rosenquist@ plantbio.lu.se

sent in the C terminus of the plant plasma membrane H+ATPase. Thus, our data suggest that one reason for the large number of isoforms found in multicellular organisms is isoform-specific functions. Key words: 14-3-3 — Isoforms — Functional specificity — Affinity — H+ATPase — Phylogeny — Surface plasmon resonance

Introduction At the time of their discovery, 14-3-3 proteins were thought to be uniquely associated with neuronal tissue (Moore and Perez 1967). During the past decade it has became clear that members of the 14-3-3 protein family are present in all types of eukaryotic cells. 14-3-3s function as homo- or heterodimers, and each ∼30-kDa monomer is able bind a phosphorylated target protein to its amphipathic binding groove (Yaffe et al. 1997). The discovery that 14-3-3s associate with oncogene products, such as Raf-1, Bcr-Adl, Bcr, and polyoma middle Tantigen, attracted a lot of attention to the 14-3-3 protein family (Ferl 1996; Meili et al. 1998). To date, 14-3-3s have been implicated as key regulators of signal transduction events and a large number of other processes, and there are reports demonstrating inhibition, activation, or structural stabilization of at least 35 different enzymes/proteins by 14-3-3s. Recently, a nuclear export signal (NES) was identified in the Schizosaccharomyces pombe 14-3-3 isoform Rad24 that appeared essential for translocating the mitosis inducing protein Cdc25 to the cytosol following DNA damage, thus postponing mitosis

447

(Lopez-Girona et al. 1999). A C-terminal NES was subsequently shown to be a common feature of 14-3-3s (Rittinger et al. 1999), which adds another function (as an attachable NES) to the rapidly growing list of functions for 14-3-3s. In plants, three important enzyme reactions are known to be affected by 14-3-3s. Binding of 14-3-3 to the cytosolic enzyme nitrate reductase inhibits its activity and thus regulates the first step in nitrogen assimilation (Moorhead et al. 1996; Bachmann et al. 1996). Similarly, a key enzyme in the regulation of sucrose synthesis in the cytosol, sucrose phosphate synthase, is inhibited by 143-3s (Toroser et al. 1998), thus regulating the pool of carbohydrate available for circulation in the plant. Finally, binding of 14-3-3 to the plasma membrane H+ATPase activates H+-pumping across the plasma membrane (Jahn et al. 1997; Oecking et al. 1997) and hence regulates cytosolic pH, the opening and closing of voltage-gated channels, and the transport capacity of all H+ symports and antiports residing in this membrane. Furthermore, a number of external stimuli, such as low temperature (Jarillo et al. 1994), high salt concentrations (Chen et al. 1994), hypoxia (de Vetten and Ferl 1995), pathogen attack (Brandt et al. 1992), and the fungal toxin fusicoccin (Roberts and Bowles 1999) stimulate 14-3-3 expression in plants. In at least two of these cases, powdery mildew (Erysiphe graminis) attack on barley (Hordeum vulgare) and hypoxia in maize (Zea mays) roots the result is increased expression of a specific 14-3-3 isoform (Brandt et al. 1992; de Vetten and Ferl 1995). In multicellular organisms, a relatively large number of 14-3-3 isoforms have evolved (Wang and Shakes 1996; this work), and the question arises as to the reason for this development. There are some apparent possible reasons. First, if a very large amount of the protein is needed, this need may be met by increasing the number of genes coding for that protein. 14-3-3s are abundant proteins (de Vetten et al. 1992; Lu et al. 1992), and the amount may increase sharply during, e.g., a particular developmental stage. For instance, in the flat worm Echinococcus multilocularis, 14-3-3 protein levels are 10 times higher in the metacestode stage than in the adult stage, with a predominant presence in the germinal cell layer (Siles-Lucas et al. 1998). Second, if specific isoforms are localized to specific subcellular compartments, as is the case for the Arabidopsis thaliana 14-3-3 isoforms epsilon, mu, nu, and upsilon, which are found both in the cytosol and the chloroplast stroma (Sehnke et al. 2000). Third, if different isoforms are expressed during different developmental stages, and/or in different cell types or tissues. For instance, in rat embryos, the protein expression levels of 14-3-3 epsilon and the target protein Raf-1 are regulated coordinately during heart development, and the expression levels of both proteins peak at 14.5 to 16.5 days postcoitum (Luk et al. 1998). Fourth, and perhaps most straightforward, if 14-3-3 isoforms are

more or less specific for their target proteins, as suggested by, e.g., the report of preferential association between particular 14-3-3 isoforms and an inhibitor of tumor necrosis factor-induced apoptosis, A20 (Vincenz and Dixit 1996). The possibility to compare 14-3-3 isoforms for obtaining information on the evolution of this intriguing protein family, on structure/function relationships, and on the reason(s) for the many isoforms in multicellular organisms, has increased dramatically. Thus, the number of unique 14-3-3 sequences available in the databases increased several-fold from 1998 to the end of 1999. In the present work, we have used these data to construct a phylogenetic tree for the 14-3-3 family, and to determine conserved amino acid residues and regions of importance for the various functions of 14-3-3s. Previously proposed conserved regions (Wang and Shakes 1996; Rittinger et al. 1999) were closely examined. We also examined conserved residues and their positions in the three dimensional structure of the 14-3-3 dimer determined by Rittinger et al. (1999). Furthermore, differences in affinity between nine 14-3-3 isoforms from A. thaliana and a novel target sequence identified in the C terminus of the plant plasma membrane H+ATPase (Olsson et al. 1998) were recorded using surface plasmon resonance. This novel 14-3-3 binding motif contains a phosphorylated threonine rather than a serine and also differs substantially in other respects from the binding motifs identified by Muslin et al. (1996) and Yaffe et al. (1997). Therefore, as being rather unique, the phosphothreonine motif was chosen as a good candidate for a motif showing specificity toward the binding 14-3-3, as opposed to the optimal motifs, which seem to bind any 14-3-3 isoform with similar affinity (Yaffe et al. 1997; Rittinger et al. 1999). Material and Methods Computational Analysis of 14-3-3 Protein Sequences. Protein sequences corresponding to different 14-3-3 isoforms were obtained from the Swissprot and Genbank databases via the National Center for Biotechnology Information (NCBI). Sequences were compared within each species to eliminate incorrect entries, identical entries, and incorrectly named entries that show absolutely no homology to 14-3-3 proteins. Clustal W multiple alignments of protein sequences were performed using the MacVector 6.5 software (Oxford Molecular Group plc, UK). The alignments were carried out using the Blossum series matrix, with an open gap penalty of 10 and an extend-gap penalty of 0.05. Alignments were examined and adjusted manually. A heuristic search using the maximum parsimony method was done on the alignment of 125 14-3-3 protein sequences using the PAUP 4.0b2 sofware (Sinauer Associates, Sunderland, MA), with gaps treated as missing data. The Tree Bisection-Reconnection (TBR) branch-swapping algorithm was used. Tree building and calculation of bootstrap values were also carried out using the PAUP software. Alleloforms and protein sequences under 100 amino acids in length were excluded from the phylogenetic analysis. Structural analysis and location of conserved residues was carried out using the 3D software Swiss PDB viewer v3.5, available at www.expasy.ch/spdbv/mainpage.htm (Guex and Peitsch 1997).

448 Expression and Purification of Nine A. thaliana 14-3-3 Isoforms (GF14s). Expression and purification of 14-3-3s was carried out according to the PET System Manual (Novagen, Madison, WI) as described previously (Wu et al. 1997a). E. coli BL21(DE3)pLysS were transformed by heat shock with pET15b expression vectors harboring inserts corresponding to the full-length clone of GF14chi, epsilon, phi, lambda, kappa, nu, omega, upsilon, and psi, respectively. Purification was carried out according to the manufacturer’s protocol by Ni2+charged immobilized metal-affinity chromatography (IMAC), using a Ni2+-HiTrap Chelating column (Pharmacia Biotech, Uppsala, Sweden). Protein concentrations were determined according to Bearden (1978). Synthesis of Target Phosphopeptide. A phosphopeptide corresponding to the last 15 C-terminal amino acids of the A. thaliana plasma membrane H+ATPase isoform AHA2 was synthesized (Saveen Biotech AB, Malmo¨, Sweden). A cysteine residue was added to the N terminus of the sequence to enable immobilization. The resulting sequence was CKLKGLDIETPSHYpTV. Surface Plasmon Resonance Analysis. Experiments were performed using a BIA 3000 (Biacore, Uppsala, Sweden). The phosphorylated 16 amino acid long peptide was immobilized on a CM-5 sensor chip via the N-terminal cysteine residue using sulfo-MBS (mmaleimidobenzoyl-N-hydroxysulfosuccinimide ester) as the coupling reagent. The surface of the chip was regenerated using 0.5% (w/v) SDS, pH 6.8, 50 mM NaCl, between each GF14 run. Nine GF14s (733 nM in running buffer) were passed over the coupled sensor chip at a flow rate of 10 ␮l/min for 2 min. As running buffer, 20 mM Tris-HCl, pH 7.5, with 5 mM MgCl2, 25 mM NaCl, and 15 mM imidazole, was used. The association was monitored by the increase in response unit (RU) caused by a change in refractive index on the chip surface. Dissociation of 14-3-3 from the immobilized peptide was monitored for 10 min after the association phase, during addition of running buffer. Obtained data was analyzed using the BIAevaluation software version 3. That equal protein concentrations of all nine GF14s were used was checked by running SDS-PAGE (Laemmli 1970) and comparing band densities after staining with Coomassie Brilliant Blue R 250.

Results Phylogenetic Analysis An analysis of 14-3-3 database entries shows that out of a total of 392 entries (end of 1999) there were 153 specific belonging to 48 different species (Table 1). 239 entries were duplicates, incorrectly identified, synthetic constructs, partial sequences of full-length entries, or incorrectly sequenced (according to the person who submitted the sequence). Isoforms showing more than 97% identity to another isoform, within the same species, were regarded as alleloforms. EST and BAC clones were not used for the phylogenetic analysis. Historically, identified 14-3-3 isoforms were initially named after the letters of the Greek alphabet; having only 24 letters the Greek alphabet did not suffice, and new isoforms have been named arbitrarily. All organisms in which 14-3-3 isoforms have been found are presented in Fig. 1A, arranged according to classical taxonomy. In mammals, there appears to be seven isoforms: beta/ alpha, gamma, epsilon, eta, sigma, tau (also referred to as theta), and zeta/delta (Fig. 1B). The isoforms alpha and

delta are phosphorylated forms of beta and zeta, respectively (Aitken et al. 1995). To date, the sigma isoform has only been found in Homo sapiens, Ovis aries, and Mus musculus. The human gamma isoform differs slightly from the other mammal gamma isoforms. Two 14-3-3 isoforms, corresponding to the mammalian isoforms epsilon and zeta, have been identified in the fruit fly Drosophila melanogaster, along with one alleloform of zeta called exon6⬘ (accession no. CAA73153). Notably, the two isoforms found in Caenorhabditis elegans, phylum Nematoda, do not correspond to any of the mammalian isoforms. One isoform has been found in the slime mold Dictyostelium discoideum, phylum Dictyosteliida; one isoform in the seaweed Fucus vesiculosus, phylum Stramenopiles; and three isoforms in the amitochondrial protozoa Entamoeba histolytica. In baker’s yeast S. cerevisiae, for which the whole genome has been sequenced, there are two 14-3-3 isoforms, BMH1 and BMH2, and two additional alleloforms of BMH2. In fission yeast, S. pombe, two isoforms, Rad24 and Rad25, have been found along with two additional Rad24 alleloforms. In the plant kingdom, Viridiplantae, there are several species belonging to different orders for which several isoforms can be found in the databases. In A. thaliana, in the order Brassicales, 12 isoforms (GF14s) and 3 additional alleloforms are found. In previous studies, only 10 isoforms have been identified (Wu et al. 1997a). One of the two new GF14s is a partially sequenced BAC clone (accession no. AAD28654), and the other is from the characterized BAC clone F21H2 (accession no. AAD46005). In the order Solanales, 11, 7, and 6 isoforms are found in Lycopersicon esculentum (tomato), Nicotiana tabacum (tobacco), and Solanum tuberosum (potato), respectively. In Glycine max (soybean), order Fabales, and Oryza sativa (rice), order Poales, five and four isoforms, respectively, are found. Thus, comparing the number of isoforms found in different organisms, it seems that unicellular organisms have few isoforms, whereas multicellular organisms have several. An unrooted phylogenetic tree with representative topology is shown in Fig. 1B. For the construction of the tree partial sequences of less than 87 amino acids were excluded, leaving 125 sequences in the analysis. Most of the 14-3-3 isoforms fall into four major groups identical to the kingdoms: Viridiplantae (plants), Fungi, Alveolata, and Metazoa (animals). The F. vesiculosus, D. discoideum, E. histolytica, and the Metazoan epsilon isoforms form separate branches. Hence, the Metazoan epsilon isoforms form their own group separated from the large Metazoan group, probably as a result of an early gene duplication during the Metazoan 14-3-3 evolution. As the inner branches of the unrooted tree are unstable, as indicated by the bootstrap values, differences between some groups are most likely smaller than they appear.

449 Table 1. List of protein identification numbers and names of the 153 14-3-3 sequences used in the present work, with the species they are found in arranged according to classical taxonomy and listed in alphabetical order Entamoebidae Entamoeba histolytica

P42648 Protein 1 P42649 Protein 2 P42650 Protein 3

Dictyosteliida Dictyostelium discoideum P54632 Stramenopiles Fucus vesiculosus

Amphibia Xenopus laevis

Q39757 Mammalia Bos taurus

ALVEOLATA Ciliophora Entodinium caudatum

Arthropode Drosophila melanogaster

Tetrahymena pyriformis

AAC35505 14-3-3-like AAD29406 Protein 2 AAD29407 Protein 3 CAB42995 Protein 4 BAA83080

Coccida Eimeria tenella Neospora caninum Toxoplasma gondii

AAD02687 Q25538 BAA25996

Haemosporida Plasmodium falsiparum Plasmodium knowlesi

AAC17516 AAC17515

Mus musculus

O42766 CAA46959 BMH1 P34730 BMH2a S69338 BMH2b CAA59275 BMH2c

Ovis aries

Homo sapiens

FUNGI Candida albicans Saccharomyces cerevisiae

Schizosaccharomyces pombe

Trichoderma harzianum

P42656 RAD24a CAA55795 RAD24b P42657 RAD25 Q99002

METAZOA Porifera Geodia cydonium Nematoda Caenorhabditis elegans Heterodera glycines

Rattus norvegicus

CAA75860

P41932 Protein 1 Q20655 Protein 2 AAC24579

P92177 Epsilon CAA73153 Exon6 P29310 Zeta

P29309 14-3-3-like AAC41251 Epsilon AAC41252 Zeta Q91896 Zeta 2 P29358 Beta AAC61927 Epsilon P11576 Eta AAC02091 Gamma A47389 Zeta S65013 Zeta 2 P31946 Beta/alpha 5803225 Epsilon CAA40620 Eta 4507951 Eta 2 S38532 Eta 3 AAA35483 Eta 4 AAD48408 Gamma 5454052 Sigma (stratifin) 5803227 Tau (⳱ Theta) 4507953 Zeta AAC14343 Beta BAA13424 Epsilon P11576 Eta BAA13422 Eta 2 AAC14345 Gamma AAC14344 Sigma P35216 Tau BAA11751 Zeta BAA13421 Zeta 2 P29358 Beta/alpha P42655 Epsilon S23305 Eta P29359 Gamma AF071008 Sigma S53753 Tau fragment P29361 Zeta/delta S59910 Zeta fragment P35213 Beta AAC52676 Epsilon BAA04259 Eta P35214 Gamma P35216 Theta P35215 Zeta S59915 Zeta 2

VIRIDIPLANTAE Plathyhelmintes Echinococcus multilocularis Schistosoma japoncium Schistosoma mansoni

Monocotelydons Fritillaria agrestis Commelina communis

AAC48315 AAD56715 Q26540 Protein 1 Q26537 Proetin 2

AAC04811 S48656

Chlorophyta Chlamydomonas reinhardtii Coniferopsida Picea glauca Solanum tuberosum

P52908

AAD27827 CAA72384 Protein 1 CAA72383 Protein 2 CAA72382 Protein 3 Q41418 Protein 4

450 Table 1.

Continued

Hordeum vulgare

Oryza sativa

Zea mays

Eudicotelydons Arabidopsis thaliana

Cucurbita pepo Glycine max

Helianthus annus Lycopersicon esculentum

Maackia amurensis Mesembryanthemum crystallinum Nicotiana tabacum

Oenothera hookeri Pisum sativum

Populus alba × tremula

CAA74592 Protein 1 P29305 14-3-3A Q43470 14-3-3B AAB07456 GF14-b AAB07457 GF14-c AAB07458 GF14-d Q06967 14-3-3 S94 P29306 143X “tau chain” P49106 GF14-6 Q01526 GF14-12 AAD28654 GF14-like AAD46005 BAC F21H2 P42643 Chi P48347 Epsilon AAD51783 Kappa P48348 Kappa 2 P48349 Lambda AAD51784 Mu AAD51782 Nu Q01525 Omega P46077 Phi P42644 Psi S47969 Psi 2(RCI14a) AAB62225 Upsilon P42647 RCI S38861 CAA06198 14-3-3-like Q96450 SGF14A Q96451 SGF14B Q96452 SGF14C Q96453 SGF14D O65352 CAA65150 BLT-like P93206 BLT1 P93208 BLT2 P93209 BLT3 P42652 BLT4 P93210 BLT5 P93211 BLT6 P93212 BLT7 P93213 BLT8 P93214 BLT9 P93207 BLT10 AAC15418 P93259 Q41246 14-3-3-like P93342 Protein A O49995 Protein B P93343 Protein C O49996 Protein D O49997 Protein E O49998 Protein F P29307 (subsp. Elata) CAB42546 Protein 1 P46266 Protein 2 CAB42547 Protein 3 AAD27825 Protein 1 AAD27824 Protein 2 AAD27823 Protein 3 AAD27822 Protein 4

Spinacia oleracea Vicia faba

P93784 Protein 16R Q43643 Protein RA215 P29308 Protein 1 P42653 Protein A P42654 Protein B CAA69347 14-3-3-like

451

Fig. 1. A The 48 organisms in which 14-3-3 isoforms have been found, arranged according to classical taxonomy. Numbers in parentheses indicate how many isoforms, plus additional alleloforms, that have been discovered in that particular species. Model organisms for which relatively large parts of the genomes have been sequenced are indicated with dots and bold text. For two of the organisms, Saccharomyces cerevisiae and Plasmodium falsiparum, the whole genome has been sequenced and is available in the databases. The lengths of the lines do not correspond to time or sequence change. B A phylogenetic tree with topology representative for the 14-3-3 family. A heuristic

search using the maximum parsimony method was done on the alignment of 125 14-3-3 protein sequences using the PAUP software, with gaps treated as missing data. Bootstrap values are indicated on selected branches. Four major groups can be observed: Viridiplantae (Plants), Fungi, Alveolata, and Metazoa (Animals). In addition, Fucus vesiculosus, Dictyostelium discoideum, the Entamoeba histolytica isoforms, and the Metazoan epsilon isoforms form separate branches. Mammalian isoforms that are identical or cluster very close together are in bold text. BLTs are L. esculentum (tomato) isoforms, SGF14s are G. max (soybean) isoforms, and GF14s are A. thaliana isoforms.

Analysis of Conserved Regions and Their Position in the 14-3-3 Structure

gions. Wang and Shakes (1996) identified 5 conserved blocks (indicated with black lines in Fig. 2C), based on analysis of the 46 14-3-3 sequences available at that time. Analysis of all 153 14-3-3 sequences available today (Table 1) reveals several conserved residues both inside and outside these blocks, as well as some not conserved residues within these blocks. The positions of the conserved residues throughout the protein suggest if they are involved in stabilizing the monomer, dimer formation, target protein binding, or other functions (Fig. 2). The majority of the conserved residues appear to stabilize the monomer, since they are inside alphahelices and face toward the interior of the protein. Some conserved residues, such as Lys193, Asp197, Ile200, Asp204, and Arg222 have side chains well exposed on the outer surface of the protein and therefore most likely have other functions.

Figure 2A and 2B show the schematic structure of the H. sapiens 14-3-3 zeta/delta isoform based on the pdb file 1QJB at the Brookhaven Protein Data Bank submitted by Rittinger et al. (1999). Looking at the same isoform in different mammalian species, an identity of 96–100% is observed. However, the difference between isoforms within a species can be as high as 46%. Conserved residues are indicated in the monomer to the right in Figs. 2A and 2B. Figure 2C shows where the conserved regions are found in the sequence of H. sapiens 14-3-3 zeta/delta. The three highly divergent Entodinium caudatum isoforms (Fig. 1B) were not included in the analysis, as they are conserved among themselves but different from all the others within the otherwise conserved re-

452

Fig. 1.

Continued.

A NES was recently discovered near the C terminus of the S. pombe isoform Rad24 and shown to function as an attachable NES for the nuclear export of the 14-3-3 target Cdc25 in response to DNA damage (Lopez-Girona et al. 1999). A C-terminal NES was subsequently shown to be a common feature of 14-3-3s and not unique to Rad24 (Rittinger et al. 1999). Our analysis reveals that all 14-3-3 isoforms contain this NES, and that the Leu/Ile repeats that are conserved in the NES of other proteins (Fukuda et al. 1997) are also conserved in all 14-3-3s (Fig. 2C). 14-3-3s exist and function as dimers, and it has been shown that the N-terminal part of the protein is involved in dimer formation (Liu et al. 1995; Luo et al. 1995; Wu et al. 1997b; Abarca et al. 1999). However, the amino acid residues involved in dimerization have not yet been

identified. Putative hydrogen bonds between monomers in the dimer were therefore calculated in Swiss PDB viewer. These calculations suggest that the amide carbonyl group of Ala16 (helix 1) and the side chain of Ser58 (helix 3) form a bond, that Tyr82 (helix 4) forms bonds with Arg18 and Asp21 (helix 2), and that Glu89 (helix 4) and Arg18 (helix 2) form a bond (Fig. 2). The hydrogen bond distances are between 2.63 and 3.09 Å. Ala16 and Ser58 exhibit 90% conserved identity and 95% conserved similarity, respectively, in all 14-3-3 isoforms. Arg18, Asp21, Try82, and Glu89 are totally conserved. Isoform-Specific Binding One way to determine isoform-specific binding is to examine how 14-3-3s from one species interact with a tar-

453

get protein from the same species. In A. thaliana there are 12 14-3-3 isoforms (GF14s) of which the 10 that are known to be expressed have been cloned (Wu et al. 1997a). As a target for these isoforms we chose a newly identified 14-3-3 binding site located in the C terminus of the plasma membrane H+ATPase. This 14-3-3 binding site constitutes a novel motif containing a phosphothreonine rather than a phosphoserine (Olsson et al. 1998; Svennelid et al. 1999; Fuglsang et al. 1999). Thus, we immobilized a phosphorylated peptide representing the last 15 amino acids of the A. thaliana H+ATPase isoform AHA2 on a chip, and used surface plasmon resonance to study interaction with GF14s. Surface plasmon resonance allows real-time binding measurements to be made with extreme sensitivity (review, Schuck 1997). Figure 3A shows superimposed sensorgrams of the interactions between immobilized AHA2 peptide and different GF14s. The data show a strong variation in relative binding affinity between the nine different 14-3-3 isoforms tested and the AHA2 target peptide. GF14phi shows the highest affinity, followed by GF14chi and nu. GF14psi, upsilon, and epsilon have intermediate binding affinity. GF14omega demonstrates below average binding affinity, whereas GF14kappa and lambda show poor binding affinity. The binding affinities correlate well to the positions of the isoforms on the branches in the phylogenetic tree (Fig. 3B), with the exception of the isoforms GF14omega and epsilon. Thus, when plotting the homology to GF14phi as a function of binding affinity compared to GF14phi, a good correlation is observed for the remaining seven isoforms (R2 ⳱ 0.83) (Fig. 3C). Our affinity data hence suggest that 14-3-3 isoforms have functional specificity, in spite of the fact that the amino acid residues exposed in the binding groove are totally conserved among all isoforms. Discussion Phylogenetic Analysis

Fig. 2. A Schematic structure of a 14-3-3 dimer, top view. Numbered boxes indicate alpha helices. Regions with conserved homology are indicated in black on the monomer to the right. B Schematic structure of a 14-3-3 dimer with phosphorylated target peptides, side view. Regions with conserved homology are indicated in dark gray on the monomer to the right. C Positions of conserved residues in 14-3-3s shown on the sequence of H. sapiens isoform zeta. Boxes indicate alpha-helical regions. Amino acids that show conserved identity in 99% of all sequences are indicated in black, and those with conserved similarity are indicated in gray. Black lines mark the five conserved blocks identified by Wang and Shakes (1996). Black dots indicate residues directly involved in peptide binding (Rittinger et al. 1999). Gray dots indicate residues that might form hydrogen bonds between monomers in the dimer, as calculated with Swiss PDB viewer (A and B are based on the pdb file 1QJB at the Brookhaven Protein Data Bank submitted by Rittinger et al. [1999]).

With the limited number of sequences available earlier, analysis of the evolution of the 14-3-3 protein family recognized three large monophyletic divisions: plants, yeast, and animals (Wang and Shakes 1996). Today, three times as many sequences are available, and four major groups of 14-3-3 isoforms can be distinguished in the phylogenetic tree presented in Fig. 1B. These four major groups coincide with the four kingdoms: Viridiplantae (plants), Fungi, Alveolata, and Metazoa (animals), which form different evolutionary lineages. Hence, it is evident that there has been an ancestral type and that, in each kingdom, this ancestral type has evolved independently into different isoforms, in general agreement with the interpretation of previous analyses (Ferl et al. 1994; Wang and Shakes 1996). The Metazoan epsilon isoforms form their own group

454

Fig. 3. A Superimposed sensorgrams obtained by surface plasmon resonance showing the relative binding affinity of nine A. thaliana isoforms (GF14s) to an immobilized peptide corresponding to the phosphorylated C terminus of the plasma membrane H+ATPase isoform AHA2. Surface plasmon resonance allows real-time binding measurements to be made with extreme sensitivity. The first phase of each curve represents the initial equilibration of the chip harboring the peptide with running buffer. The second phase shows an increase in signal as an effect of 14-3-3 binding. In the third phase of the sensorgram, running buffer is passed over the chip, resulting in the dissociation of 14-3-3. The box to the right shows which isoform corresponds to which line and are ordered according to affinity. B A phylogenetic tree with the 10 A. thaliana 14-3-3 isoforms known to be expressed. Bold text indicates which isoforms were used in the affinity study in A. Bootstrap values are indicated on each branch. C Diagram showing protein sequence homology to GF14phi as a function of the binding affinity, expressed as percentage of the binding affinity of GF14phi. When calculating the function and correlation, GF14epsilon and GF14omega were excluded due to their deviance.

separated from the large Metazoan group. This was observed also in previous studies (Aitken et al. 1992; Ferl et al. 1994; Wang and Shakes 1996), and has led to the suggestion that the epsilon isoforms are relatively conserved and similar to the original, ancestral animal 143-3 protein (Wang and Shakes 1996; McEwan et al. 1999). Notably, a very recent database entry shows that an epsilon isoform (accession no. AAF21436) is present also in Schistosoma mansoni, of the phylum Plathyhelmintes (not included in the analysis). By contrast, no epsilon isoform has yet been found in C. elegans, although most of its genome is now available. In Viridiplantae, gene duplication has apparently occurred late in the evolution of the phyla. The grouping of 14-3-3s belonging to the order Poales (maize, rice, and wheat) and Liliales (Fritillaria agrestis) suggests that gene duplication events mainly took place after the Eudicotyledon and Liliopsida class formations (Fig. 1B). This clearly demonstrated when a phylogenetic analysis is carried out solely on plant isoforms (61 entries; data not shown). Notably, however, some eudicotyledon isoforms, such as the A. thaliana GF14psi, nu, and upsilon; one Mesembryanthemum crystallinum isoform; and three Solanales isoforms; fall on the same branch as all the monocotyledon (Liliopsida) isoforms, suggesting that at least two isoforms were present already before the division into Eudicotyledons and Monocotyledons (Fig. 1A), as suggested also by Wang and Shakes (1996). Most of the gene duplication probably occurred very late, and for the Eudicotyledons only four isoforms may have been present before the split into Caryophyllidae, Rosidae, and Asteridae. Thus, most of the A. thaliana isoforms, namely GF14psi, nu and upsilon, phi, chi and omega, as well as kappa and lambda, are clustered on three minor branches and are more similar to themselves than to any other plant isoform (Fig. 1B). At the time of previous phylogenetic studies (Wang and Shakes 1996; Wu et al. 1997a; McEwan et al. 1999) the three isoforms of E. histolytica were selected as outgroup as they were believed to be the most divergent group of 14-3-3 isoforms. However, it is apparent that among the isoforms discovered since then there are more divergent isoforms, e.g., the E. caudatum isoforms. The more rapid evolution of the 14-3-3s in the Alveolata clade, particularly Ciliophora, is not surprising as these organisms have very short life cycles. Furthermore, Apicomplexa pathogens, such as malaria (Plasmodium falsiparum), are constantly faced with evolving defense systems in host organisms, which promotes a high degree of mutation. Since it is not obvious which 14-3-3 isoform should be used to root a tree we chose to make an unrooted tree. Sequence and Structure Analysis 14-3-3s are active as homo- or heterodimers, with each monomer being able to bind a target protein. Conserved

455

residues are therefore expected to be involved in either stabilizing the monomer, dimer formation, target binding, or other specific functions. The conservation pattern of 14-3-3s shows that the protein has not undergone neutral morphological changes. Evolution of proteins that are involved in protein complex formation are subject to a high selection pressure, which will tend to conserve the parts of the proteins that are involved in structural complex formation (e.g., Johnson 1996). For 14-3-3s, which on one hand form complexes with themselves and on the other hand with other proteins the selection pressure will be very high, which may explain why such large portions of the 14-3-3 proteins are conserved. Dimer formation was initially shown to be due to interactions between the N-terminal parts of the monomers (Jones et al. 1995; Liu et al. 1995; Xiao et al. 1995). Later reports (Luo et al. 1995; Wu et al. 1997b; Abarca et al. 1999) and our present analysis (Fig. 2A and B) indicate that interaction more precisely takes place between the conserved block 1 of one monomer and conserved residues in helices 3 and 4 of the other monomer. In a recent study, using the A. thaliana 14-3-3 isoforms GF14psi and lambda, Abarca and co-workers (1999) showed that only helix 4 of a truncated monomer is essential for dimer formation with an untruncated monomer. To identify amino acid residues that may be involved in dimerization we used the Swiss PDB viewer, which identified the amide carbonyl group of Ala16, and the side chains of Arg18, Asp21, Ser58, Tyr82, and Glu89 as involved in hydrogen bond formation between block 1 in one monomer and helices 3 and 4 in the other monomer. Of these residues, all are conserved throughout the whole 14-3-3 protein family, except for Ala16 (block 1) and Ser58 (helix 3), suggesting that the latter two are not essential for dimer formation. This is supported by the work of Abarca et al. (1999) showing that Ala16 and Ser58 interaction is not essential for dimerization of A. thaliana isoforms GF14psi and lambda. However, the lack of these residues may well affect the stability of the dimer. 14-3-3s preferentially bind phosphorylated target motifs, and the structural work by Yaffe et al. (1997) and Rittinger et al. (1999) show in detail which amino acid residues in the 14-3-3 monomer that are involved in binding of a synthetic target peptide. As the target peptide binding residues are totally conserved, possible binding specificity of different 14-3-3s must derive from other parts of the protein, assuming that there are no more interacting residues to be identified. Presumably, the amino acid sequence in other parts of the protein may affect the positions of the helices contributing to the binding surface and in that way affect binding specificity. Thus, the drastic difference in binding affinity between for instance GF14phi and omega (Fig. 3A) may be due to differences in single amino acid residues. Our analysis shows that all 14-3-3 isoforms found in

the databases contain a NES in the C terminus. This NES motif is totally conserved throughout the whole 14-3-3 protein family, and it is therefore not surprising that 143-3 isoforms can replace each other in performing the nuclear export function. However, out of the four NES Leu/Ile residues at least two are involved in target protein binding (Rittinger et al. 1999; Fig. 2C). This raises the question how the 14-3-3 simultaneously can both bind a target protein and act as an attachable NES. This problem was addressed by Rittinger et al. (1999), who demonstrated that the presence of a 14-3-3-binding peptide at least partially inhibited binding of the 14-3-3 to Crm1 in the nuclear export machinery. As a likely solution to the problem they suggested that one 14-3-3 subunit in the dimer would bind the target protein while the other would provide the NES function. Phosphorylation of a number of 14-3-3 isoforms has been reported (Lu et al. 1994; Autieri and Carbone 1999). Lu et al. (1994) identified the phosphorylated amino acid(s) in the A. thaliana isoform GF14 omega as serine. Autieri and Carbone (1999) showed that human 14-3-3 gamma is phosphorylated, but the phosphorylated amino acid was not identified. Out of the six phosphorylation sites proposed by Autieri and Carbone (1999), our analysis shows that Ser214 is present in all 14-3-3 isoforms in all species, Ser57 and Ser58 are present in all mammalian 14-3-3 isoforms, and Tyr126 and Ser145 are present in several isoforms. These five residues are all involved in hydrogen bond stabilization of the monomer according to calculations made with Swiss PDB viewer, and a phosphorylation of any of these residues may therefore affect the conformation of the protein. However, Ser57, Ser58, and Ser145 are relatively hidden in the protein core and thus not particularly accessible to protein kinases. In contrast, Tyr126 and Ser214 are somewhat more exposed. The sixth proposed phosphorylation site, Tyr112, is only found in the mammalian isoforms eta and gamma, which makes it a likely candidate for an isoform-specific phosphorylation. Dubois et al. (1997) have shown that human 14-3-3 zeta is phosphorylated in vivo on Thr232 (Thr233 in Dubois et al. [1997]), which is only found in 14-3-3 zeta and hence constitutes an isoform-specific phosphorylation. Phosphorylation at either Ser214 or Thr232 would affect the top lateral wall of the binding cleft, and could conceivably affect ligand binding as demonstrated by Dubois et al. (1997) for the binding of human 14-3-3 zeta to c-Raf. Together with the phosphorylation of human isoform beta reported by Aitken et al. (1995), this suggests that many 14-3-3s might phosphorylated. How these phosphorylations relate to the functions of the 14-3-3s remains largely to be determined. Functional Specificity of 14-3-3 Isoforms To date, may 14-3-3 target proteins and several different binding motifs, naturally occurring and synthesized,

456

have been identified (reviewed in Palmgren et al. 1998). Considering the large number of 14-3-3 isoforms and the large number of target proteins in many organisms, the question has arisen to what extent 14-3-3 isoforms are specific for their target proteins. A common postulate is that all 14-3-3 isoforms bind with more or less the same specificity to a defined target, and hence show little specificity. This view is supported by some experimental data (reviewed in Palmgren et al. 1998). For instance, all A. thaliana 14-3-3 isoforms that could be expressed in yeast could also replace the endogenous 14-3-3s (van Heusden et al. 1996). By contrast, Bachmann et al. (1996) found strong differences in the ability of five recombinant A. thaliana 14-3-3 isoforms to inhibit the activity of isolated and phosphorylated spinach nitrate reductase; data which indicate functional specificity of isoforms. There are also other reports on isoformspecific interactions with target proteins, but these have been suggested to be due to particular subcellular localization or transcriptional regulation of isoforms rather than to inherent differences in their ability to bind to a specific target (Rittinger et al. 1999; and references therein). A genetic, rather than a functional, model for 14-3-3 isoform diversity is advocated by Palmgren et al. (1998), who suggest that the important differences between 14-3-3 genes are to be found in their promoter regions. To avoid problems such as specific subcellular localizations of isoforms we expressed nine A. thaliana isoforms in E. coli and tested their ability to bind to a phosphorylated target peptide representing the last 15 amino acids of the A. thaliana H+-ATPase isoform AHA2. This recently identified binding motif contains a phosphorylated threonine rather than a serine (Olsson et al. 1998), and binding of 14-3-3 to the C terminus of the H+ATPase activates H+ pumping (Jahn et al. 1997; Oecking et al. 1997). The 14-3-3 protein hence serves as a positive regulator of H+ATPase activity. Comparing the C termini of all plant H+-ATPase isoforms in the databases, the minimal motif for phosphorylation-dependent 14-3-3 binding was found to be YpTV (p denoting a phosphorylated residue), including an enrichment of Q, H, N, and S at the pT −4 to −2 positions (Svennelid et al. 1999). This motif is thus very different from the two optimal motifs investigated by Yaffe et al. (1997) and Rittinger et al. (1999), R[S/Ar][+]pS[L/E/A/M]P and RX[Y/F][+]pS[L/E/A/M]P (Ar denoting an amino acid with an aromatic side chain, and + an amino acid with a positively charged side chain). The H+ATPase motif contains a phosphothreonine, not a phosphoserine; it has no positively charged amino acid, and it does not have a proline in the +2 position. According to the structural data by Yaffe et al. (1997), the phosphoserine in the binding peptide may well be replaced by a phosphothreonine. The +2 proline is important because it produces a sharp change in direction of the binding polypeptide and

allows it to exit the binding cleft of the 14-3-3 (Yaffe et al. 1997). This structural problem is avoided in the H+ATPase since its C terminus ends at pT +1. The basic amino acids in the optimal motifs constitute a phosphorylation signal for animal protein kinases A and C (Yaffe et al. 1997). In the H+ATPase motif, the corresponding function seems to be fulfilled by the Q, H, N, and S at the pT −4 to −2 positions, which probably act as a phosphorylation signal for a plant protein kinase (Svennelid et al. 1999). Notably, the H+ATPase motif is not a typical plant motif. Hence, both sucrose-phosphate synthase and nitrate reductase, which are two plant enzymes that are inhibited by 14-3-3 binding (Toroser et al. 1998; Moorhead et al. 1996), have binding motifs (RQISSP and KSVSTP, respectively, in A. thaliana) similar to the optimal motifs. The profound differences between the H+ ATPase motif and the optimal motifs may explain why we found strong differences in binding affinity between 14-3-3 isoforms and the H+ATPase motif (Fig. 3A), when others have failed to find differences using various isoforms and motifs similar to the optimal motifs (Yaffe et al. 1997; Rittinger et al. 1999). Thus, it could be that isoform specificity is low for the optimal motifs but may be much more pronounced towards deviating and hence more specific motifs.

Concluding Remarks The evolution of 14-3-3 proteins has resulted in a relatively large number of isoforms in multicellular organisms and in a few isoforms in unicellular organisms. Therefore, the need for a large number of isoforms should somehow be related to the needs of a multicellular organism compared to the needs of a unicellular one. This would rule out subcellular compartmentation (Rittinger et al. 1999) as a good reason, since the subcellular compartmentation of yeast with two isoforms is as sophisticated as that of a mammalian cell with seven isoforms. Rather, 14-3-3 isoform specificity with regard to organ (Luk et al. 1998), tissue (Daugherty et al. 1996), or cell type (Namikawa et al. 1998) distribution would fit with the specific needs of a multicellular organism. As proteins involved in complex formation, 14-3-3s are subject to a high selection pressure, which acts to conserve the parts of the protein involved in protein/protein interaction. However, 14-3-3 target proteins also mutate, and new target proteins evolve, making selection pressure specific for the 14-3-3s, at least when the 14-3-3 binding site is directly or indirectly affected by the mutations. In our view, this evolution should promote 14-3-3 isoform specificity for target proteins once gene duplication has taken place. Possibly, evolution has retained a broad specificity among 14-3-3s for some motifs, the so-called optimal motifs, whereas specificity has evolved for other motifs, such as the H+ATPase motif.

457 Acknowledgments. We would like to thank Biacore (Uppsala, Sweden), especially Anette Persson, for allowing us to perform invaluable affinity measurements at their facility. We would also like to thank Axel Janke, Department of Genetics, Lund University, for lending us his expertise in the field of phylogenetics. We are grateful to Fredrik Svennelid, Department of Plant Biochemistry, Lund University, for aiding in the purification of 14-3-3 proteins, and to Katarina Mo¨ller, Reprostugan AB (Arlo¨v, Sweden) for frequent printouts of figures. This work was supported by the Swedish Foundation for Strategic Research (C.L. and M.S.), the Swedish Natural Science Research Council (C.L. and M.S.), the Swedish Council for Forestry and Agricultural Research (M.S.), and the European Union Biotechnology Programme (C.L.).

References Abarca D, Madueno F, Martinez-Zapater JM, Salinas J (1999) Dimerization of Arabidopsis 14-3-3 proteins: structural requirements within the N-terminal domain and effect of calcium. FEBS Lett 462:377–382 Aitken A, Collinge DB, van Heusden BP, Isobe T, Roseboom PH, Rosenfeld G, Soll J (1992) 14-3-3 proteins: a highly conserved, widespread family of eukaryotic proteins. Trends Biochem Sci 17: 498–501 Aitken A, Howell S, Jones D, Madrazo J, Patel Y (1995) 14-3-3 alpha and delta are the phosphorylated forms of raf-activating 14-3-3 beta and zeta. In vivo stoichiometric phosphorylation in brain at a SerPro-Glu-Lys motif. J Biol Chem 270:5706–5709 Autieri MV, Carbone CJ (1999) 14-3-3 gamma interacts with and is phosphorylated by multiple protein kinase C isoforms in PDGFstimulated human vascular smooth muscle cells. DNA Cell Biol 18:555–564 Bachmann M, Huber JL, Athwal GS, Wu K, Ferl RJ, Huber SC (1996) 14-3-3 proteins associate with the regulatory phosphorylation site of spinach leaf nitrate reductase in an isoform-specific manner and reduce dephosphorylation of Ser-543 by endogenous protein phosphatases. FEBS Lett 398:26–30 Bearden JC Jr (1978) Quantitation of submicrogram quantities of protein by an improved protein-dye binding assay. Biochim Biophys Acta 104:549–555 Brandt J, Thordal-Christensen H, Vad K, Gregersen PL, Collinge DB (1992) A pathogen-induced gene of barley encodes a protein showing high similarity to a protein kinase regulator. Plant J 2:815–820 Chen Z, Fu H, Liu D, Chang PF, Narasimhan M, Ferl R, Hasegawa PM, Bressan RA (1994) A NaCl-regulated plant gene encoding a brain protein homolog that activates ADP ribosyltransferase and inhibits protein kinase C. Plant J 6:729–740 Daugherty CJ, Rooney MF, Miller PW, Ferl RJ (1996) Molecular organization and tissue-specific expression of an Arbaidopsis 143-3 gene. Plant Cell 8:1239–1248 de Vetten NC, Ferl RJ (1995) Characterization of a maize G-box binding factor that is induced by hypoxia. Plant J 7:589–601 de Vetten NC, Lu G, Ferl RJ (1992) A maize protein associated with the G-box binding complex has homology to brain regulatory proteins. Plant Cell 4:1295–1307 Dubois T, Rommel C, Howell S, Steinhussen U, Soneji Y, Morrice N, Moelling K, Aitken A (1997) 14-3-3 is phosphorylated by casein kinase I on residue 233. Phosphorylation at this site in vivo regulates Raf/14-3-3 interaction. J Biol Chem 272:28882–28888 Ferl R (1996) 14-3-3 proteins and signal transduction. Annu Rev Plant Physiol Plant Mol Biol 47:49–73 Ferl RJ, Lu G, Bowen BW (1994) Evolutionary implications of the family of 14-3-3 brain protein homologs in Arabidopsis thaliana. Genetica 92:129–138 Fuglsang AT, Visconti S, Drumm K, Jahn T, Stensballe A, Mattei B, Jensen ON, Aducci P, Palmgren MG (1999) Binding of 14-3-3

protein to the plasma membrane H+-ATPase AHA2 involves the three C-terminal residues Tyr(946)-Thr-Val and requires phosphorylation of Thr947. J Biol Chem 274:36774–36780 Fukuda M, Asano S, Nakamura T, Adachi M, Yoshida M, Yanagida M, Nishida E (1997) CRM1 is responsible for intracellular transport mediated by the nuclear export signal. Nature 390:308–311 Guex N, Peitsch MC (1997) SWISS-MODEL and the SwissPdbViewer: an environment for comparative protein modeling. Electrophoresis 18:2714–2723 Jahn T, Fuglsang AT, Olsson A, Bru¨ntrup IM, Collinge DB, Volkmann D, Sommarin M, Palmgren MG, Larsson C (1997) The 14-3-3 protein interacts directly with the C-terminal region of the plant plasma membrane H+-ATPase. Plant Cell 9:1805–1814 Jarillo JA, Capel J, Leyva A, Martinez-Zapater JM, Salinas J (1994) Two related low-temperature-inducible genes of Arabidopsis encode proteins showing high homology to 14-3-3 proteins, a family of putative kinase regulators. Plant Mol Biol 25:693–704 Johnson JE (1996) Functional implications of protein-protein interactions in icosahedral viruses. Proc Natl Acad Sci USA 93:27–33 Jones DH, Martin H, Madrazo J, Robinson KA, Nielsen P, Roseboom PH, Patel Y, Howell SA, Aitken A (1995) Expression and structural analysis of 14-3-3 proteins. J Mol Biol 245:375–384 Laemmli UK (1970) Cleavage of structural proteins during the assembly of the head of bacteriophage T7. Nature 227:680–685 Liu D, Bienkowska J, Petosa C, Collier RJ, Fu H, Liddington R (1995) Crystal structure of the zeta isoform of the 14-3-3 protein. Nature 376:191–194 Lopez-Girona A, Furnari B, Mondesert O, Russell P (1999) Nuclear localization of Cdc25 is regulated by DNA damage and a 14-3-3 protein. Nature 397:172–175 Lu G, Rooney MF, Wu K, Ferl RJ (1994) Five cDNAs encoding Arabidopsis GF14 proteins. Plant Physiol 105:1459–1460 Lu G, DeLisle AJ, de Vetten NC, Ferl RJ (1992) Brain proteins in plants: an Arabidopsis homolog to neurotransmitter pathway activators is part of a DNA binding complex. Proc Natl Acad Sci USA 89:11490–11494 Luk SC, Ngai SM, Tsui SK, Chan KK, Fung KP, Lee CY, Waye MM (1998) Developmental regulation of 14-3-3 epsilon isoform in rat heart. J Cell Biochem 68:195–199 Luo ZJ, Zhang XF, Rapp U, Avruch J (1995) Identification of the 14-3.3 zeta domains important for self-association and Raf binding. J Biol Chem 270:23681–23687 McEwan NR, Eschenlauer SC, Calza RE, Wallace RJ, Newbold CJ (1999) Protozoal sequences may reveal additional isoforms of the 14-3-3 protein family. Protist 150:257–264 Meili R, Cron P, Hemmings BA, Ballmer-Hofer K (1998) Protein kinase B/Akt is activated by polyomavirus middle-T antigen via a phosphatidylinositol 3-kinase-dependent mechanism. Oncogene 16:903–907 Moore B, Perez V (1967) Specific acidic proteins of the nervous system. In: Carlson FD (ed) Physiological and biochemical aspects of nervous integration. Englewood Cliffs, NJ: Prentice-Hall, pp 343– 359 Moorhead G, Douglas P, Morrice N, Scarabel M, Aitken A, MacKintosh C (1996) Phosphorylated nitrate reductase from spinach leaves is inhibited by 14-3-3 proteins and activated by fusicoccin. Curr Biol 6:1104–1113 Muslin AJ, Tanner JW, Allen PM, Shaw AS (1996) Interaction of 14-3-3 with signaling proteins is mediated by the recognition of phosphoserine. Cell 84:889–897 Namikawa K, Su Q, Kiryu-Seo S, Kiyama H (1998) Enhanced expression of 14-3-3 family members in injured motoneurons. Brain Res Mol Brain Res 55:315–320 Oecking C, Piotrowski M, Hagemeier J, Hagemann K (1997) Topology and target interaction of the fusicoccin-binding 14-3-3 homologs of Commelina communis. Plant J 12:441–453 Olsson A, Svennelid F, Ek B, Sommarin M, Larsson C (1998) A phosphothreonine residue at the C-terminal end of the plasma mem-

458 brane H+-ATPase is protected by fusicoccin-induced 14-3-3 binding. Plant Physiol 118:551–555 Palmgren MG, Fuglsang AT, Jahn T (1998) Deciphering the role of 14-3-3 proteins. Exp Biol Online 3:4 Rittinger K, Budman J, Xu J, Volinia S, Cantley LC, Smerdon SJ, Gamblin SJ, Yaffe MB (1999) Structural analysis of 14-3-3 phosphopeptide complexes identifies a dual role for the nuclear export signal of 14-3-3 in ligand binding. Mol Cell 4:153–166 Roberts MR, Bowles DJ (1999) Fusicoccin, 14-3-3 proteins, and defense responses in tomato plants. Plant Physiol 119:1243–1250 Schuck P (1997) Use of surface plasmon resonance to probe the equilibrium and dynamic aspects of interactions between biological macromolecules. Annu Rev Biophys Biomol Struct 26:541–566 Sehnke PC, Henry R, Cline K, Ferl RJ (2000) Interaction of a plant 14-3-3 protein with the signal peptide of a thylakoid-targeted chloroplast precursor protein and the presence of 14-3-3 isoforms in the chloroplast stroma. Plant Physiol 122:235–242 Siles-Lucas M, Felleisen RS, Hemphill A, Wilson W, Gottstein B (1998) Stage-specific expression of the 14-3-3 gene in Echinococcus multilocularis. Mol Biochem Parasitol 91:281–293 Svennelid F, Olsson A, Piotrowski M, Rosenquist M, Ottman C, Larsson C, Oecking C, Sommarin M (1999) Phosphorylation of Thr-948 at the C terminus of the plasma membrane H+-ATPase creates a binding site for the regulatory 14-3-3 protein. Plant Cell 11:2379– 2392

Toroser D, Athwal GS, Huber SC (1998) Site-specific regulatory interaction between spinach leaf sucrose-phosphate synthase and 143-3 proteins. FEBS Lett 435:110–114 van Heusden GP, van der Zanden AL, Ferl RJ, Steensma HY (1996) Four Arabidopsis thaliana 14-3-3 protein isoforms can complement the lethal yeast bmh1 bmh2 double disruption. FEBS Lett 391:252– 256 Vincenz C, Dixit VM (1996) 14-3-3 proteins associate with A20 in an isoform-specific manner and function both as chaperone and adapter molecules. J Biol Chem 271:20029–20034 Wang W, Shakes DC (1996) Molecular evolution of the 14-3-3 protein family. J Mol Evol 43:384–398 Wu K, Rooney MF, Ferl RJ (1997a) The Arabidopsis 14-3-3 multigene family. Plant Physiol 114:1421–1431 Wu K, Lu G, Sehnke P, Ferl RJ (1997b) The heterologous interactions among plant 14-3-3 proteins and identification of regions that are important for dimerization. Arch Biochem Biophys 339:2–8 Xiao B, Smerdon SJ, Jones DH, Dodson GG, Soneji Y, Aitken A, Gamblin SJ (1995) Structure of a 14-3-3 protein and implications for coordination of multiple signalling pathways. Nature 376:188– 191 Yaffe MB, Rittinger K, Volinia S, Caron PR, Aitken A, Leffers H, Gamblin SJ, Smerdon SJ, Cantley LC (1997) The structural basis for 14-3-3:phosphopeptide binding specificity. Cell 91:961–971