Identification of the site of interaction of the 14-3-3 protein with phosphorylated tryptophan hydroxylase.

The 14-3-3 protein family plays a role in a wide variety of cell signaling processes including monoamine synthesis, exocytosis, and cell cycle regulation, but the structural requirements for the activity of this protein family are not known. We have previously shown that the 14-3-3 protein binds with and activates phosphorylated tryptophan hydroxylase (TPH, the rate-limiting enzyme in the biosynthesis of neurotransmitter serotonin) and proposed that this activity might be mediated through the COOH-terminal acidic region of the 14-3-3 molecules. In this report we demonstrate, using a series of truncation mutants of the 14-3-3 isoform expressed in Escherichia coli, that the COOH-terminal region, especially restricted in amino acids 171-213, binds indeed with the phosphorylated TPH. This restricted region, which we termed 14-3-3 box I, is one of the structural regions whose sequence is highly conserved beyond species, allowing that the plant 14-3-3 isoform (GF14) could also activate rat brain TPH. The 14-3-3 box I is the first functional region whose activity has directly been defined in the 14-3-3 sequence and may represent a common structural element whereby 14-3-3 interacts with other target proteins such as Raf-1 kinase. The result is consistent with the recently published crystal structure of this protein family, which suggests the importance of the negatively charged groove-like structure in the ligand binding.

The 14-3-3 protein family consists of acidic, dimeric proteins with relative molecular masses of ϳ60 kDa distributed widely among eukaryotic cells (for reviews, see Refs. 1 and 2). Numerous biochemical activities have been attributed to this family of proteins, including calmodulin (CaM) 1 kinase II-dependent activation of enzymes involved in neurotransmitter synthesis, Ca 2ϩ -dependent stimulation of noradrenalin secretion, and the regulation of the activity of Ca 2ϩ -phospholipid-dependent protein kinase C (2). Recent findings, particularly in fission yeast, have shown that the 14-3-3 family is associated with the products of proto-oncogenes and oncogenes such as Raf-1, Bcr, Bcr-Abl, and Polyomavirus middle tumor antigen, suggesting that this family of proteins is also involved in cell transformation and mitogenic signaling pathways (for a review, see Ref. 3). We have previously shown that the 14-3-3 protein activates TPH in concert with phosphorylation of the hydroxylase by CaM kinase II (4,5) and demonstrated that this activation results from the binding of 14-3-3 protein to the phosphorylated hydroxylase (6). From these results, together with the structural features of the 14-3-3 protein (4), we proposed that the acidic COOH-terminal region of the 14-3-3 protein might be involved in the interaction with phosphorylated TPH. However, little evidence has been provided in this and other systems on the structural requirements of the 14-3-3 protein for its biological activities.
In this study, we used the TPH system as a model to assess the functional region of the 14-3-3 protein and showed that the acidic COOH-terminal region, especially restricted in residues 171-213, is a primary site for the interaction of 14-3-3 with phosphorylated TPH.
Expression and Purification of GST Fusion Proteins-Cultures of Escherichia coli strain JM109, transformed by 14-3-3 plasmids, were grown and induced with isopropyl-1-thio-␤-D-galactopyranoside for expression as described (8). The bacteria were collected by centrifugation and resuspended in buffer A (20 mM Tris/HCl, 150 mM NaCl, 0.1 mM dithiothreitol, pH 7.5). Vigorous sonication was performed before centrifugation at 12,000 ϫ g for 20 min. The resulting supernatant was loaded onto a column (1.4 ϫ 3 cm) packed with 3 ml of cross-linked * This work was supported by Grants-in-aid for Scientific Research from the Ministry of Education, Science and Culture of Japan and also by a research grant from Ciba-Geigy Foundation (Japan) for the Promotion of Science. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
Phosphorylation and Binding Assay-Phosphorylation of TPH (ϳ1 g) was carried out at 30°C for 20 min in a reaction mixture (50 mM Hepes, pH 7.6, 5 mM Mg(CH 3 COO) 2 , 0.1 mM CaCl 2 , 0.5 mM ATP, 1 g of CaM, 1 g of CaM kinase II, and various amounts of GST-fused protein (see figure legends)) in a final volume of 200 l. The control experiment was performed under the same conditions described above except that CaM kinase II was removed from the reaction mixture. For phosphorylation of rat brainstem extract (25-55% saturated (NH 4 ) 2 SO 4 fraction), the extract (500 g) was incubated in a buffer (200 l) containing 50 mM Hepes, pH 7.6, 5 mM Mg(CH 3 COO) 2 , 0.1 mM CaCl 2 , 0.5 mM ATP, 4 g of CaM, and 5 g of GST-fused protein at 30°C for 20 min. Control experiments were performed without ATP. For binding assay, glutathione-agarose beads (ϳ50 l) were added to the reaction mixture and incubated for 30 min at 4°C, and the protein complexes bound to the beads were washed three times with buffer A and solubilized in SDS sample buffer. The bound TPH was analyzed by SDS-PAGE followed by Western blotting using a TPH antibody. In some experiments, bound TPH was measured with its enzymatic activity using the washed beads directly.
Others-TPH activity was measured fluorometrically according to the procedure described previously (7). One unit of TPH was defined as the amount that catalyzes the formation of 1 nmol of 5-hydroxytryptophan per min at 30°C. The amounts of recombinant protein were evaluated from the densitometric quantitation of the protein bands obtained after SDS-PAGE (Coomassie Blue staining).

RESULTS AND DISCUSSION
We used the bacterial expression system that produces proteins in E. coli as fusions with an affinity tag, GST, to prepare recombinant proteins, because this system permits affinity purification of active proteins. Bovine brain 14-3-3 isoform was expressed in this system, and the GST-fused protein was purified by affinity chromatography on cross-linked glutathione resin. The purified protein showed a single protein band on SDS gel with an expected molecular mass of ϳ55 kDa (see Fig.  2B, lane 1), and this protein cross-reacted with polyclonal antibodies to bovine brain 14-3-3 protein (data not shown).
Before truncation of the recombinant protein, we examined, using TPH system, whether the GST-fused protein produced in E. coli indeed shares similar properties with bovine brain isoform (Fig. 1). The fused protein added to the TPH assay mixture stimulated the activity of TPH about 2-fold more than the level of TPH measured in the absence of the fused protein (Fig. 1A). This effect was dose-dependent, and the concentration of the protein necessary for half-maximal activation (V max /2) of TPH was about 20 nM. These values were almost equal to the published values with bovine brain 14-3-3 (4, 5). In addition, no stimulation of TPH activity was observed, as analyzed in the absence of CaM kinase II or in the presence of GST alone (Fig. 1A, dotted lines), confirming the previous data (4,5) that this activation of TPH needs both phosphorylation of TPH and the 14-3-3 protein.
We next examined whether the fused protein can interact with TPH in a phosphorylation-dependent manner. TPH was incubated with the fused protein under its phosphorylating or nonphosphorylating conditions (see "Experimental Procedures"), and glutathione-agarose beads that bind with the fused protein were added to the mixtures. Bound TPH with the fused protein immobilized on the beads was then assayed by its enzymatic activity and by Western blot with a TPH antibody. As illustrated in Fig. 1B (lanes 3 and 4), TPH bound with the fused protein only under its phosphorylating condition. Incorporation of phosphate to TPH under this condition was confirmed using [␥-32 P]ATP (data not shown, see Ref. 6), suggesting that, like the bovine brain protein (6), the recombinant protein binds with TPH in a phosphorylation-dependent manner.
We also performed similar experiments using crude brainstem extract supplemented with the recombinant protein (Fig. 1C). As shown in Fig. 1C, lane 4, TPH present in the brainstem extract also bound to the added protein, indicating that this binding occurs in the crude extract. Again, no interaction of TPH to the protein was detected under nonphosphorylating conditions (lane 3). In both experiments, GST alone did not bind to TPH (Fig. 1, B and C, lanes 1 and 2). All these properties of the expressed fusion protein were same as the reported characteristics of brain 14-3-3 protein. Furthermore, like the bovine brain protein (9, 10), the expressed protein  1 and 3) or phosphorylating (lanes 2 and 4) conditions in the presence of GST (2 g, lanes 1 and 2) or GST-fused protein (2 g, lanes 3 and 4), and glutathione-agarose beads were added to the mixture. Bound TPH with the fused protein immobilized on the beads was assayed by its enzymatic activity using the beads or by Western blot with TPH antibody (inset). The cross-reacting band was visualized by horseradish peroxidase-conjugated goat antibodies against rabbit IgG and ECL reagent (Amersham Corp.). The arrowhead indicates the position of TPH. C, formation of a complex between the recombinant 14-3-3 protein and the phosphorylated TPH in the brainstem extract. The rat brainstem extract (500 g) was incubated under nonphosphorylating (lanes 1 and 3) or phosphorylating (lanes 2 and 4) conditions in the presence of GST (5 g, lanes 1 and 2) or GST-fused protein (5 g, lanes 3 and 4) and was then analyzed as in B. ND, not detected.
activated protein kinase C. 2 Thus, we concluded that the expression system used in this study can produce the protein that is suitable for analysis of the functional region of the 14-3-3 protein.
A series of truncation mutants were made ( Fig. 2A), and the expressed proteins were purified with the same procedure described above. As judged by SDS-PAGE (Fig. 2B), the mutants were almost pure with a major band of the expected sizes. The truncated GST-fused proteins were then examined in a similar way to that described in the legend of Fig. 1B for their ability to bind with phosphorylated TPH (see also the legend to Fig. 3). This analysis confirmed the importance of the COOH-terminal acidic region of protein (residues 170 -246) for the interaction with TPH, because the deletion mutants lacking this region (mutants 1-78 and 1-170, Fig. 2A) were found no longer bound with TPH (Fig. 3A, lanes 2 and 3, compare with lane 1). In addition, deletions of up to 33 COOH-terminal residues (1-213) almost retained the activity of full-length protein (1-246), but further deletion of 23 residues  reduced the activity (Fig. 3A, lanes 4 -6). These deletions suggest that residues 171-213 located in the COOH-terminal acidic region of the protein may be necessary for the interaction with the phosphorylated TPH.
To test this assumption, deletions were made to produce three additional mutants, the mutant 167-246 (carrying the complete COOH-terminal region), the mutant 167-213 (carrying the 171-213 region), and the mutant lacking the 171-213 region (⌬171-213, Fig. 2A), and these mutants were analyzed in terms of TPH binding. This analysis revealed that the two mutants, 167-246 and 167-213, bound to TPH to a similar extent (Fig. 3B, lanes 1 and 2), but the mutant ⌬171-213 did not (lane 3). These results proved the above assumption and provided direct evidence that the structural region consisting of residues 171-213, which we termed 14-3-3 box I, is a primary site for the interaction of the protein to the phosphorylated TPH. We also observed that all of the mutants which lacked the box I and failed to bind phosphorylated TPH, such as the mutants 1-170 and ⌬171-213, had no activities toward TPH even with the excess amount over TPH. This suggests that the box I structure is also essential for the activity of the protein.
Our results, however, cannot exclude the possibility that an additional region(s) to the box I participates in the interaction with TPH because the truncation of the box I might induce conformational changes in another part of the molecule and thereby prevent the mutants lacking the box I from the interaction with the hydroxylase.
The amino acid sequence of the 14-3-3 box I is one of the highly conservative sequences extended among the members of the 14-3-3 family. Fig. 4 displays the alignment of corresponding sequences of the 14-3-3 box I from bovine brain 14-3-3 isoforms with a known sequence as well as the plant and yeast counterparts, GF14 and BMH1, respectively. The sequence similarity suggests that the 14-3-3 box I represents a common structural element, which may be involved in the association of these 14-3-3 isoforms to TPH. Consistent with this assumption, the previous data have shown that the rat brain TPH could be activated by all these bovine isoforms as well as by the plant GF14 (5,11), although at present whether the yeast BMH1 2 O. Kunihoro, T. Ichimura, and T. Isobe, unpublished data. protein on TPH binding. TPH (ϳ1 g) was phosphorylated by CaM kinase II for 20 min in the presence of the indicated proteins (25 pmol each) and then analyzed for its binding as in the experiment shown in Fig. 1, B and C (inset). B, effects of mutants 167-246, 167-213, and ⌬171-213 on TPH binding. The experimental conditions were same as in A except that 50 pmol of each mutant protein was used. Arrows indicate the position of TPH.
FIG. 4. Alignment of the amino acid sequences (single-letter notation) of 14-3-3 box is of bovine brain isoforms, Arabidopsis GF14, and yeast BMH1. The sequences of bovine ␤, ␥, , and were from Refs. 5 and 9, and that of ⑀ was from our recent study (T. Isobe and T. Ichimura, unpublished data). The amino acid sequences of Arabidopsis GF14 and yeast BMH1 were from Refs. 11 and 17, respectively. Amino acid sequences are numbered from the N terminus of each isoform. Amino acids identical in all these isoforms are shown in reverse type.
Recently, the crystal structures of the homodimeric proteins of the 14-3-3 (13) and (14) isoforms have been reported. These proteins have a similar tertiary fold consisting of a bundle of nine anti-parallel ␣-helices of each monomer, and the dimers form a large negatively charged channel or groove. Both reports have suggested that this groove may represent the ligand binding surface of the 14-3-3 molecules. In the tertiary structure, the 14-3-3 box I represents helices 7 and 8 and a part of the linker between helices 8 and 9, which are located near the edge of the groove. In this viewpoint, our results are consistent with the proposal from the crystal structure studies and further emphasize the important role of the COOH-terminal structure including helices 7 and 8 in the ligand/14-3-3 protein interaction.
Increasing evidence suggests that the phosphorylation may be a common mechanism that regulates the binding of 14-3-3 to its target molecules. For example, the interaction of 14-3-3 with Raf-1 and Bcr protein kinases was ultimately prevented by the treatment of these kinases with protein phosphatase (15). It has also been shown that the 14-3-3 reduces the dephosphorylation of Raf-1 with protein phosphatases through the interaction with Raf-1 and blocks the dephosphorylation-induced enzymatic inactivation of Raf-1 (16). The facts that the mode of interaction between these protein kinases and 14-3-3 is similar to the TPH/14-3-3 interaction described here and that the 14-3-3 box I is a highly conserved structure from yeast to mammals (Fig. 4) imply that the box I structure may serve as a common binding site for many target proteins including these protein kinases, but this awaits further investigation.