14-3-3 Proteins: A Number of Functions for a Numbered Protein

See allHide authors and affiliations

Science's STKE  09 Aug 2005:
Vol. 2005, Issue 296, pp. re10
DOI: 10.1126/stke.2962005re10


Many signal transduction events are orchestrated by specific interactions of proteins mediated through discrete phosphopeptide-binding motifs. Although several phosphospecific-binding domains are now known, 14-3-3s were the first proteins recognized to specifically bind a discrete phosphoserine or phosphothreonine motif. The 14-3-3 proteins are a family of ubiquitously expressed, exclusively eukaryotic proteins with an astonishingly large number of binding partners. Consequently, 14-3-3s modulate an enormous and diverse group of cellular processes. The effects of 14-3-3 proteins on their targets can be broadly defined using three categories: (i) conformational change; (ii) physical occlusion of sequence-specific or structural protein features; and (iii) scaffolding. This review will describe the current state of knowledge on 14-3-3 proteins, highlighting several important advances, and will attempt to provide a framework by which 14-3-3 functions can be understood.


Cell signaling results from a tightly regulated interplay of protein-protein interactions and posttranslational modifications of signaling pathway components. The most prevalent means of posttranslational modification is probably reversible protein phosphorylation (1). Phosphorylation can modulate the catalytic activity, substrate specificity, subcellular localization, and stability of various proteins, as well as the assembly and disassembly of macromolecular complexes (14). It is therefore not surprising that direct links between phosphorylation state and protein-protein interactions have been uncovered in many biological systems. One way in which protein phosphorylation can regulate protein-protein interactions is through modules that specifically bind to a phosphorylated motif in another protein. Such modules include 14-3-3 proteins and FHA (forkhead-associated), WW, WD40, LRR (leucine-rich repeat), and BRCT (BRCA-1 C-terminal) domains for serine and threonine phosphorylation, as well as SH2 (Src-homology 2) and PTB (phosphotyrosine binding) domains for tyrosine phosphorylation (58). This simple, yet elegant, mechanism enables particular proteins to interact with each other only when an upstream phosphorylation event has previously occurred.

The 14-3-3 proteins were discovered and named in 1967 as part of an examination of brain tissue proteins (9); however, their importance was not recognized until they were identified as activators of the tyrosine and tryptophan hydroxylases (10, 11) and as inhibitors of protein kinase C and activators of Raf (1215). Around the same time that Raf1 was discovered to bind 14-3-3 proteins, it was also reported that the middle T antigen (MTAg) of polyoma virus bound 14-3-3 proteins (16). This was followed by the discovery in 1996 that 14-3-3 proteins bind to specific phosphorylated motifs in protein targets (17).

This review summarizes our current knowledge of 14-3-3 proteins and describes some exciting new developments in this field. Readers are directed to several excellent reviews on 14-3-3 function for further information (1824).

Structure and Function

The 14-3-3 proteins are small (~30 kD), acidic proteins that form both homo- and heterodimers. The number of isoforms found in various organisms ranges from 2 in yeast, Caenorhabditis elegans, and Drosophila melanogaster to 7 in mammals and 12 in Arabidopsis (25). 14-3-3 proteins have been found in all eukaryotic organisms examined, which indicates an early eukaryotic origin. Even a 14-3-3 of the primitive eukaryote Giardia lamblia, an organism believed to be more than twice as distant from humans as fungi (on the basis of molecular phylogeny) is highly conserved compared with human 14-3-3s (26). At the time of initial writing, we searched 282 prokaryote genome sequences at the National Center for Biotechnology Information (of which 155 were complete) and found no 14-3-3 or 14-3-3–like proteins. All 14-3-3s are highly conserved, with about 50% identity among amino acids both within and across species (Fig. 1A). The importance of 14-3-3 function is illustrated by the lethal phenotype derived from simultaneous knockout of both isoforms in the yeast S. cerevisiae (27). In several cases, it has been shown that knockout of 14-3-3 genes can be complemented by 14-3-3 genes from another organism (28, 29).

Fig. 1.

Conservation of peptide binding by 14-3-3 protein. (A) Sequence alignment of evolutionarily diverse 14-3-3 proteins. Residues that are 100% identical are shaded in black, and residues that are >50% identical are shaded in gray. Included in the alignment are representative proteins from Saccharomyces cerevisiae (Bmh1), Arabidopsis thaliana (GP14κ), D. melanogaster (14-3-3ε), Caenorhabditis elegans (14-3-3 isoform 1), and human (14-3-3ζ). Residues that directly contact the phosphate group of target proteins are indicated with red arrows. (B) Structure of 14-3-3ζ shaded according to residue conservation. Structure analyzed using Consurf (124) with 14-3-3 isoforms from humans (7 isoforms), A. thaliana (12 isoforms), S. cerevisiae (2 isoforms), D. melanogaster (2 isoforms), and C. elegans (2 isoforms). Sequence alignment was performed using ClustalX (116) with residue conservation determined by a maximum likelihood method within Consurf. Structures (B), (C), and (D) were based on 14-3-3ζ [PDB 1QJB (38)]. Legend denotes relative conservation levels in the protein, with teal representing the least-conserved residues and burgundy representing the most highly conserved residues. A CHIME version in which Fig. 1B can be manipulated and viewed from different perspectives is available at;2005/296/re10/DC1. In (C) and (D), the peptide (ARSHpSYPA; shown in green with the phosphate group highlighted in yellow) binds in an extended conformation to 14-3-3 (shown as a ribbon diagram in red or blue) within the conserved groove of each monomer. (D) shows one half dimer from the perspective of the other monomer. Structure figures were generated using Molscript and Raster3D (126, 127).

Initial crystallographic studies of the ligand-free proteins showed that 14-3-3 proteins form highly helical, cup-shaped dimers (30, 31) (Fig. 1, B and C). The current model suggests that this highly rigid structure leads to deformation of the target protein with little or no change in the structure of the 14-3-3 dimer (23, 32). The only exception to this model appears to be the flexible carboxyl-terminal region of the protein, which was recently shown to change in conformation when phosphorylated (33). This relatively unconserved region has been hypothesized to play an autoinhibitory role in ligand binding (33, 34), possibly in an isoform-specific manner. Each subunit of the dimer is able to bind one discrete phosphoserine- or phosphothreonine-containing ligand independently with the peptide ligands in extended antiparallel orientations at a distance of 34 Å between phosphate groups (30, 32, 35) (Fig. 1, C and D). The phosphopeptide-binding pocket is the most highly conserved region both within and across species (Fig. 1, A and B). The residues equivalent to Lys49, Arg56, Arg127, and Tyr128 of human 14-3-3ζ, which bind the phosphate moiety of the phosphorylated residues of the bound protein, are completely conserved in all 14-3-3 proteins known (32, 3538). In addition to the peptide binding groove, other regions of the 14-3-3 surface have more recently been implicated in determining substrate binding specificity (39).

Recent crystallography of SMG7 (suppressor with morphogenetic effects on genitalia 7) revealed unexpected structural homology to 14-3-3 proteins (40). SMG5, SMG6, and SMG7 are proteins that regulate nonsense-mediated decay (the degradation of mRNA transcripts containing nonsense mutations) in metazoans. In spite of having less than 10% sequence identity, SMG7 folded nearly identically to a 14-3-3 monomer. Even more surprisingly, SMG7 was also shown to bind to phosphopeptides in a manner analogous to 14-3-3 proteins. This finding, along with the previously reported similarity in structure and function of alpha synuclein (41), suggests that a number of 14-3-3 –like proteins may have evolved.

Like the C-terminal region, the amino terminal region of 14-3-3 proteins is not highly conserved (Fig. 1A). This region is responsible for the dimerization of 14-3-3 proteins (42, 43). 14-3-3 isoforms heterodimerize in specific combinations in cells (39, 43), and specific allotment of heterodimer pairs could have substantial implications for the biological effects exerted by 14-3-3 proteins.

Examination of 14-3-3–binding targets and application of an oriented peptide library approach uncovered two phosphopeptide motifs that bind 14-3-3 proteins (17, 35, 38). These two motifs are RSXpSXP and RXXXpSXP (where R is arginine, S is serine, X is any amino acid, pS is phosphoserine, and P is proline). Phosphothreonine (pT) can replace pS. In addition to the peptide binding groove, other regions of the 14-3-3 surface have now also been implicated in determining substrate binding specificity (39). Many target proteins do not contain sequences that conform precisely to these motifs. This may relate to the dimeric structure of 14-3-3 proteins. Several binding partners have more than one 14-3-3–binding site (4446), and the affinity of a peptide with two 14-3-3–binding motifs for the 14-3-3 dimer is more than 30 times that of a peptide containing a single phosphorylated motif (38). This most likely occurs because once one phosphopeptide binds, the effective local concentration of the second phosphopeptide increases dramatically. Thus, two "imperfect" sites may be sufficient to bind to the 14-3-3 dimer, if they are in antiparallel orientation and able to achieve the correct spacing. In fact, a binding mode involving two imperfect sites may be the more common mode of interaction between 14-3-3 proteins and their targets (23, 35, 47). Furthermore, the fact that some targets have suboptimal 14-3-3–binding motifs could be physiologically important [see discussion of the plant proton adenosine triphosphatase (ATPase) below].

Another potential benefit of "suboptimal" binding motifs is that they are more likely to dissociate from the 14-3-3 than are motifs comprised of the "perfect" consensus sequence. This is highlighted by considering the interaction among 14-3-3, fusicoccin, and a "suboptimal" binding peptide derived from the 14-3-3 target plant plasma membrane H+-ATPase (37). The proton pumping activity of the H+-ATPase can be blocked by its autoinhibitory C terminus. This inhibition is abolished by phosphorylation of the second-to-last residue of the enzyme (QSYpTV-COOH) and the subsequent docking of 14-3-3 (37, 48, 49). A peptide based on this motif binds 14-3-3 protein with relatively weak affinity [dissociation constant (Kd) of 2.5 μM versus 37.4 nM for the optimal mode 2 peptide] (37, 38). When the fungal toxin fusicoccin is present, the affinity for the peptide is increased dramatically (Kd = 27 nM) (37). A crystal structure of 14-3-3, toxin, and QSYpTV peptide demonstrated that fusicoccin fills in a cavity at the 14-3-3-peptide interface, which is empty because of the shortened 14-3-3–binding motif (37). Fusicoccin causes a constitutive activation of the proton pump (37, 48, 49), which in planta causes constant opening of the stomatal aperture and subsequent wilting. Thus, the ability of 14-3-3 to readily dissociate from an "imperfect" binding partner may in this case be critical to its normal physiological function.

In addition to the phosphopeptide-binding ability of 14-3-3 proteins, it has been well documented that 14-3-3 proteins can also bind unphosphorylated targets (36, 5053). These targets bind to 14-3-3 proteins in the same location as phosphorylated targets and can compete with phosphopeptides for binding, which adds another possible level of complexity to 14-3-3 target recognition (53). To date, nonphosphorylated targets constitute a very small population of the total number of 14-3-3 interactors. This is consistent with a recent article indicating that the vast majority of 14-3-3 targets bind in a phosphorylation-dependent manner (54, 55).


Recent structural and biochemical advances have helped generalize the roles of 14-3-3 proteins in cells. Whereas Aitken classified 14-3-3 proteins in the context of their target proteins (18), we propose a classification system for 14-3-3 proteins on the basis of the following modes of action: (i) directed conformational changes of the target protein, (ii) physical occlusion of sequence-specific or structural features, and (iii) scaffolding. This scheme is an update to the classification systems described by Fu et al. (56) and Tzivion and Avruch (22). It should also be noted that more than one of these modes of action could be used simultaneously.

14-3-3 Directed Conformational Changes

The highly α-helical nature of the 14-3-3 proteins provides them with a very rigid structure. Cocrystallization studies of 14-3-3 proteins with peptides or whole proteins have shown little change in the structure of the 14-3-3 protein compared with the ligand-free form (23, 30, 32, 35, 38). This structural rigidity has been described in terms of a molecular anvil hypothesis: 14-3-3 proteins act as a stable support on which target proteins can be reshaped (23), thereby altering some property of the target (Fig. 2A).

Fig. 2.

Structural effects of 14-3-3 binding. 14-3-3 binding can elicit three effects: (A) A conformational change in the target protein; (B) masking of a specific region on the target (this could include an active site, a ligand-binding region, or a region that interacts with another protein); or (C) colocalization of two proteins. Each monomer of 14-3-3 is shown in a different shade of blue.

The anvil effect is illustrated by several examples. The serotonin–N-acetyltransferase protein is normally catalytically inefficient. After phosphorylation, 14-3-3 binding results in stabilization of the enzyme’s active conformation. This 14-3-3–mediated stabilization elicits an increase in the catalytic efficiency of the enzyme because of increases in both its catalytic rate and substrate affinity (32). Although serotonin–N-acetyltransferase is the only 14-3-3 target for which the anvil hypothesis can be explained structurally, this concept could explain the effects of 14-3-3 proteins in various other systems. Serotonin–N-acetyltransferase also provides a beautiful example of the binding mechanism described above in which phosphorylated sites at two separate regions on one protein each bind to a separate subunit of the 14-3-3 dimer (57).

Several proteins are catalytically inactivated by 14-3-3 binding. One of the best-characterized examples of this is plant nitrate reductase (5860). More recent examples include the plant mitochondrial and chloroplast adenosine 5′-triphosphate (ATP) synthases. The plant ATP-synthase β subunit was originally identified as a 14-3-3 interactor through affinity chromatography on 14-3-3–Sepharose (61). ATP synthases produce ATP in mitochondria and chloroplasts through the generation of a membrane proton gradient (62). Activity of these enzymes must be shut down when the proton gradient across the membrane is abolished because of darkness (chloroplasts) or lack of oxygen (mitochondria), because, under these conditions, the enzyme can function in reverse to hydrolyze ATP. 14-3-3 binds directly to both the mitochondrial and chloroplast ATP synthase β subunits, which results in inactivation of the enzyme and thus prevents reversibility of the ATP synthesis reaction (61, 63). This could conceivably involve the reorientation of the active site to a catalytically inactive conformation as a result of 14-3-3 binding or occlusion of a key catalytic region of the β subunit.

Another means by which 14-3-3–mediated conformational changes can modify a protein is through directed alteration of distant structural features. If changes caused by 14-3-3 binding can result in substantial rearrangement of an active site, then it follows that 14-3-3 binding could conceivably also alter a target protein surface involved in protein-protein interactions.

One example of this involves the class II histone deacetylases (HDACs). After phosphorylation, these proteins bind to 14-3-3 proteins, and this binding results in their relocation to the cytoplasm from the nucleus (64). Initially this shuttling was believed to result solely from masking of a nearby nuclear localization sequence (NLS) adjacent to the 14-3-3–binding site. However, more recent work has shown that removal of a nuclear export sequence (NES) from HDAC5 can abolish the nuclear export of the target protein, even though its 14-3-3 binding–and therefore the NLS exposure–is not affected (65). This suggests that nuclear phosphorylation of HDACs and 14-3-3 binding to an N-terminal region results in the exposure of the C-terminal NES and thereby causes nuclear export. A direct interaction between HDACs and the nuclear export protein CRM1 in a 14-3-3–regulated manner has yet to be formally shown. According to the current model, 14-3-3 blockade of the NLS on class II HDACs is still considered important for regulating nuclear import (65, 66).

14-3-3 Blockage of Sequence-Specific or Structural Features

In contrast to 14-3-3–induced changes of relatively distant structural features, 14-3-3 proteins also act as masks that directly occlude important features of target proteins (Fig. 2B). In the only crystallized 14-3-3-target protein complex, the serotonin–N-acetyltransferase–14-3-3ζ interaction buries 2527 Å2 of surface area of the target protein (32). When a 14-3-3 protein binds a target, it could block access of other possible interactors. In several cases, an NLS is located within close proximity of a 14-3-3–binding site, which leads to the proposal that 14-3-3 binding blocks access of importin proteins to the NLS region of the target protein. In addition to the HDAC example discussed above, this mode of regulation has been well studied with respect to the Cdc25 cell cycle dual-specificity phosphatases. These proteins have an NES that is thought to be constitutively exposed and an NLS that is in close proximity to a 14-3-3–binding site. When phosphorylated (Ser323 in Cdc25B) by checkpoint kinases, 14-3-3 proteins bind to Cdc25 and likely occlude the NLS site(335 to 354 in Cdc25B). As initially described for Cdc25C, this allows the rate of nuclear export of the phosphatase to exceed the rate of nuclear import. Once the phosphatase has been removed from the nucleus, phosphorylation and activation of the Cdc2-cyclin B protein kinase occurs, which leads to mitotic entry (6674). For a more detailed examination of the Cdc25 regulatory mechanism, we refer the reader to a recent review (75).

Another example of 14-3-3 function by masking has emerged in the field of endoplasmic reticulum (ER) protein retention and release. Some proteins synthesized in the ER are retained there by a diarginine motif that is "recognized" by the coat protein complex I (COPI) retention machinery (7678). 14-3-3ζ and ε were identified as interacting proteins for the dibasic motif of the C terminus of the potassium channel α subunit, Kir6.2, through an affinity-binding approach (76). 14-3-3 proteins only recognized the multimeric, not the monomeric, form of the Kir6.2 tail and competed for binding to the dibasic region with the COPI machinery, which suggests that bound 14-3-3 masks the COPI interaction site and thus causes release of Kir6.2 from the ER. The authors proposed that 14-3-3 proteins "probe" the assembly state of the channel and, when they "recognize" a fully assembled target, override the retention signal and allow its release. In a similar work from another group (78), a yeast two-hybrid screen revealed that 14-3-3β bound the C-terminal tail of the dimeric KCNK3 potassium channel. This interaction was phosphorylation-dependent, and the binding site again was a nonclassic 14-3-3–binding motif (RRSpSV) located in the last few residues of the protein. In the KCNK3 protein, the dibasic retention motif is on the N terminus of the protein, but, similar to Kir6.2, KCNK3 binding to 14-3-3 or COPI was mutually exclusive. Two other examples of this type of regulation are the Iip35 major histocompatibility antigen class II–associated invariant chain and the α4 acetylcholine receptor, both of which have overlapping dibasic retention signals and 14-3-3–binding motifs (78, 79). The exact method by which the 14-3-3 protein probes the target for correct folding or assembly is unknown, but one attractive possibility is that only correctly folded or assembled oligomers have their 14-3-3–binding sites at the correct distance and orientation to bind to a 14-3-3 dimer, whereas misfolded or nonassembled oligomers do not.

Members of the FOXO class of transcription factors are active in the nucleus in their dephosphorylated state and have been shown to be phosphorylated by various protein kinases. Phosphorylation leads to their expulsion from the nucleus and termination of their target gene transcription (8082). Two of these sites are phosphorylated by protein kinase B (PKB), and this phosphorylation results in association with 14-3-3 proteins (47, 80, 83, 84). These 14-3-3–binding sites bracket the DNA binding domain and inhibit the DNA binding activity of the protein (47, 81), as well as its nuclear localization (80, 8385), in a 14-3-3–dependent manner. Because FOXO proteins are monomers in solution and they bind a 14-3-3 dimer with a 1:1 ratio, it is thought that each phosphorylation site interacts with one of the 14-3-3 subunits (47). 14-3-3 dimerization is necessary for the inhibition of FOXO DNA binding activity in vitro (83). On the basis of this requirement, it would be interesting to see what effects a dimerization-deficient 14-3-3 protein would have on FOXO-mediated gene transcription. Several other proteins, including p53 (86, 87), topoisomerase II (88), and the viral protein Rep68 (89), also have DNA binding activity that is decreased by interaction with 14-3-3 proteins, although whether masking or deformation is the cause is still unclear.

Another important feature of 14-3-3 masking is the protection of phosphorylation sites from dephosphorylation. Studies have shown that, both in vitro and in vivo, sites that bind 14-3-3 proteins are protected from protein phosphatases (68, 90-92). This is likely because of the inability of the phosphatase to access the phosphate group. This implies that, in order to counteract the effects of 14-3-3 binding, two steps must occur. First, the 14-3-3 protein must dissociate from the target, to expose the phosphate group (68). Second, a protein phosphatase must remove the phosphate. It is also worth noting that suboptimal interactions, such as those discussed earlier for the H+-ATPase, could be more easily reversible because of the faster off-rate of the 14-3-3–target protein interaction. If the interaction fluctuates between the bound and unbound states rapidly, the phosphate group would be exposed to protein phosphatases more often. Therefore, in cases such as the H+-ATPase, phosphatase protection may be less of an issue than for high-affinity, low off-rate interactions.

14-3-3 Colocalization of Proteins

Another function of 14-3-3 proteins is to act as a phosphorylation-dependent scaffold to anchor proteins within close proximity of one another (Fig. 2C). The first study to suggest this mechanism of regulation was published in 1995 (93). For example, one-half of the dimer could bind to a protein kinase, while the other site binds to the kinase substrate. Therefore, when a kinase activation signal is detected, the substrate is rapidly and specifically phosphorylated. This allows for tight regulation of phosphorylation and specific targeting of a site on the substrate because of the restricted orientation of the kinase active site and the substrate. The advantage of this type of regulation is that a priming phosphorylation of both the kinase and the substrate is required first, before full substrate phosphorylation. This allows several tiers of regulation to occur in a system.

One example of 14-3-3 scaffolding involves glycogen synthase kinase 3β (GSK-3β) and tau. These proteins copurify along with a 14-3-3 protein, and 14-3-3 proteins lead to more efficient phosphorylation of tau by GSK-3β (94). There are several other examples of kinases that have been shown to be part of 14-3-3 complexes (55, 93, 95106), which indicates that this could be a relatively widespread means of regulating kinase function. It should be noted, however, that, despite several reports of this type of 14-3-3–based regulation of kinase activity, no experiments as yet have demonstrated this conclusively.

The identification of several enzymes involved in primary metabolism as part of the mammalian 14-3-3 interactome could indicate another 14-3-3 scaffolding role in the spatial colocalization of a metabolic pathway. For example, in the de novo generation of purines from sugars, enzymes catalyzing the first, second, fourth, and seventh through 10th steps were all affinity-purified by 14-3-3–Sepharose (55). Conceivably, this could lead to a 14-3-3–dependent colocalization, coregulation, or both, of several enzymes in this metabolic pathway.

Regulation of 14-3-3 Proteins

Although much is known about how 14-3-3 proteins regulate their targets, comparatively little is known about how the 14-3-3 proteins themselves are regulated. Transcriptionally, most 14-3-3 isoforms are ubiquitously expressed in all tissues, with the exception of mammalian 14-3-3σ, which has a more restricted expression (107). Several plant isoforms have been shown to be tissue-specific (108). Furthermore, although most 14-3-3 isoforms bind their targets with similar affinity, some isoform-specific effects have been described [reviewed in (18, 109, 110)].

14-3-3σ in particular displays a number of isoform-specific effects. 14-3-3σ appears to be the least conserved of all mammalian isoforms and may have derived from a retrotransposon event (39). In addition to its more restricted expression, this isoform differs from other 14-3-3 proteins with respect to the specificity of substrate binding (39, 111). Structural studies on 14-3-3σ have suggested that a small number of amino acid substitutions have resulted in a protein that selectively forms homo- but not heterodimers, with target affinity that differs from that of the 14-3-3s that heterodimerize (39). This difference in target specificity appears to be the result of differences outside of the phosphopeptide binding groove (39).

14-3-3 proteins have long been known to be phosphorylated in vivo on several sites, but the physiological relevance of this was unclear (18, 112, 113). Two groups have demonstrated that the 14-3-3ζ Ser58 site is phosphorylated in vivo and–as predicted by its location in the structure–this phosphorylation promotes the formation of 14-3-3 monomers as opposed to dimers (105, 113). This could have profound effects on 14-3-3 targets that interact only with dimers, although more work must be done to clarify the physiological conditions under which phosphorylation at this site may occur (20). Two recent studies have shown that Ser184 phosphorylation results in decreases the interaction of 14-3-3 isoforms with their targets (114, 115). Another phosphorylation site, Ser233 of 14-3-3ζ, negatively regulates target protein binding (18, 33, 112, 116). This C-terminal region is thought to occupy the phosphopeptide binding groove in the absence of a bound target (117); thus, the phosphorylation of a residue in this region may be predicted to enhance the affinity of the C terminus for the phosphopeptide binding groove. This Ser233 site is part of a suboptimal 14-3-3 binding motif that may competitively inhibit target binding once phosphorylated; however, high-affinity targets may overcome this inhibition. None of these three phosphorylation sites are conserved in every human 14-3-3 isoform. Thus, isoform specificity and heterodimerization may play an important role in the phosphorylation of 14-3-3 proteins.

A further possibly relevant level of 14-3-3 regulation is through cofactors such as adenosine 5′-monophosphate (AMP) and magnesium (118). In the presence of AMP, reduced binding of 14-3-3 proteins to some plant targets has been observed (58, 119). In agreement with this, the global proteolysis of plant 14-3-3–binding proteins after sucrose starvation can be blocked by the addition of 5-amino-4-imidazolecarboxamide riboside, a drug that results in an in vivo increase in the AMP mimic 5′-ZMP(5-aminoimidazole-4-carboxamide riboside monophosphate) (120). AMP can modulate the effects of 14-3-3 proteins on the plant plasma membrane H+-ATPase both in vivo and in vitro (119). Furthermore, AMP can directly bind to the plant 14-3-3 protein GF14-6 (119). Preliminary work in our laboratory has extended this AMP regulation of 14-3-3 target binding to mammals. It is possible, therefore, that stress-dependent increases in AMP cause the release of 14-3-3 proteins from their targets (or selected targets), leading to cellular effects. The recent identification of many primary metabolic enzymes as 14-3-3 targets could conceptually advance this hypothesis (55).

The 14-3-3 Interactome

The majority of 14-3-3–binding partners have been identified from the perspective of the interacting protein. Often, 14-3-3 proteins were found as copurifying cofactors of proteins or in yeast two-hybrid screens, with the target protein used as bait. From the viewpoint of someone who is interested in finding novel 14-3-3 targets, there has, until recently, been far less research. Large-scale studies of the S. cerevisiae 14-3-3 interactome have uncovered 16 proteins that interact with the yeast 14-3-3 isoforms, BMH1 or BMH2, but no two groups have found the same interactions (121, 122). Using a number of affinity approaches, several groups have now performed proteomic studies to identify proteins that bind 14-3-3 directly or that reside in complexes containing 14-3-3 proteins (54, 55, 123, 124). This revealed that many enzymes involved in primary metabolism, signaling, trafficking, and transcription are 14-3-3 interactors. These studies are interesting not only for the collection of proteins identified, but also because the proteins identified by each group differ greatly. We estimate that there is only ~25% identified protein overlap between these studies. We believe that these differences may have to do with differences in starting material and specific isoform, as well biases of the different methodologies. The starting material differed in several of these studies, as did the 14-3-3 isoform used, making it difficult to compare these works. One would predict that the immobilized affinity matrix used by one group (55) might be biased toward lower-affinity (rapidly exchanging) targets. However, the single isoform approach employed by other groups (54, 123, 124) could skew the results toward isoform-specific interactors. Only 17% of the proteins identified in the most recent study (124), which used 14-3-3σ as the target, were those identified by 14-3-3ζ (123), in spite of similar methodology and starting material. As more of these studies are done, more isoform- and tissue-specific interactors will be found. This highlights the importance of determining which method of analyzing the 14-3-3 interactome is most effective from a biochemical standpoint. The recent work implicating isoform-specific effects, combined with heterodimerization complexity, may complicate the results of proteomic studies further.

In the span of only 10 years, 14-3-3 proteins have gone from an almost unknown entity in biochemistry to the forefront of some of the most important processes in biology. Recent work has uncovered many 14-3-3–binding partners, as well as the mechanisms of target identification and modulation. Furthermore, 14-3-3 proteins are being implicated in the regulation of a huge spectrum of both general and specialized signaling pathways, which indicates the key importance of their role in health and disease.


  1. 1.
  2. 2.
  3. 3.
  4. 4.
  5. 5.
  6. 6.
  7. 7.
  8. 8.
  9. 9.
  10. 10.
  11. 11.
  12. 12.
  13. 13.
  14. 14.
  15. 15.
  16. 16.
  17. 17.
  18. 18.
  19. 19.
  20. 20.
  21. 21.
  22. 22.
  23. 23.
  24. 24.
  25. 25.
  26. 26.
  27. 27.
  28. 28.
  29. 29.
  30. 30.
  31. 31.
  32. 32.
  33. 33.
  34. 34.
  35. 35.
  36. 36.
  37. 37.
  38. 38.
  39. 39.
  40. 40.
  41. 41.
  42. 42.
  43. 43.
  44. 44.
  45. 45.
  46. 46.
  47. 47.
  48. 48.
  49. 49.
  50. 50.
  51. 51.
  52. 52.
  53. 53.
  54. 54.
  55. 55.
  56. 56.
  57. 57.
  58. 58.
  59. 59.
  60. 60.
  61. 61.
  62. 62.
  63. 63.
  64. 64.
  65. 65.
  66. 66.
  67. 67.
  68. 68.
  69. 69.
  70. 70.
  71. 71.
  72. 72.
  73. 73.
  74. 74.
  75. 75.
  76. 76.
  77. 77.
  78. 78.
  79. 79.
  80. 80.
  81. 81.
  82. 82.
  83. 83.
  84. 84.
  85. 85.
  86. 86.
  87. 87.
  88. 88.
  89. 89.
  90. 90.
  91. 91.
  92. 92.
  93. 93.
  94. 94.
  95. 95.
  96. 96.
  97. 97.
  98. 98.
  99. 99.
  100. 100.
  101. 101.
  102. 102.
  103. 103.
  104. 104.
  105. 105.
  106. 106.
  107. 107.
  108. 108.
  109. 109.
  110. 110.
  111. 111.
  112. 112.
  113. 113.
  114. 114.
  115. 115.
  116. 116.
  117. 117.
  118. 118.
  119. 119.
  120. 120.
  121. 121.
  122. 122.
  123. 123.
  124. 124.
  125. 125.
  126. 126.
  127. 127.
View Abstract

Navigate This Article