Proteasome-driven turnover of tryptophan hydroxylase is triggered by phosphorylation in RBL2H3 cells, a serotonin producing mast cell line Yoshiko Iida 1 , Keiko Sawabe 1 , Masayo Kojima 1 , Kazuya Oguro 1,2 , Nobuo Nakanishi 3 and Hiroyuki Hasegawa 1,2 1 Department of Bioscience, and 2 Biotechnology Research Center, Teikyo University of Science and Technology, Yamanashi, Japan; 3 Departments of Biochemistry, Meikai University School of Dentistry, Sakado, Saitama, Japan We previously demonstrated in mast cell lines RBL2H3 and FMA3 that tryptophan hydroxylase (TPH) undergoes very fast turnover driven by 26S-proteasomes [Kojima, M., Oguro, K., Sawabe, K., Iida, Y., Ikeda, R., Yamashita, A., Nakanishi, N. & Hasegawa, H. (2000) J. Biochem (Tokyo) 2000, 127, 121–127]. In the present study, we have examined an involvement of TPH phosphorylation in the rapid turn- over, using non-neural TPH. The proteasome-driven deg- radation of TPH in living cells was accelerated by okadaic acid, a protein phosphatase inhibitor. Incorporation of 32 P into a 53-kDa protein, which was judged to be TPH based on autoradiography and Western blot analysis using anti-TPH serum and purified TPH as the size marker, was observed in FMA3 cells only in the presence of both okadaic acid and MG132, inhibitors of protein phosphatase and proteasome, respectively. In a cell-free proteasome system constituted mainly of RBL2H3 cell extracts, degradation of exogenous TPH isolated from mastocytoma P-815 cells was inhibited by protein kinase inhibitors KN-62 and K252a but not by H89. Consistent with the inhibitor specificity, the same TPH was phosphorylated by exogenous Ca 2+ /calmodulin- dependent protein kinase II in the presence of Ca 2+ and calmodulin but not by protein kinase A (catalytic subunit). TPH protein thus phosphorylated by Ca 2+ /calmodulin- dependent protein kinase II was digested more rapidly in the cell-free proteasome system than was the nonphosphoryl- ated enzyme. These results indicated that the phosphoryla- tion of TPH was a prerequisite for proteasome-driven TPH degradation. Keywords: tetrahydrobiopterin; CaM kinase II; proteasome target; ubiquitin ligase; enzyme turnover. Tryptophan hydroxylase (TPH, EC 1.14.16.4), a member of a family of pterin-dependent aromatic amino acid hydroxy- lases [1], catalyzes the conversion of L -tryptophan to 5-hydroxy- L -tryptophan. This reaction is the initial and rate-limiting step in the biosynthesis of serotonin [2–5]. TPH has been extensively purified from various sources such as bovine pineal gland [6], mouse mastocytoma [7,8], and mammalian brains [9–11]. Physicochemical, enzymic and immunochemical properties differed between TPHs of neural and non-neural tissue origin, and it is accepted that neural TPH might be a different entity from the non-neural enzyme [8,10,12,13]. Complimentary DNAs of TPH have been cloned from various sources but no differences or only trivial variation in amino acid sequences were found among them [14–19]. The molecular basis of differences between the neural and non-neural enzymes has not yet been explained. Both types of cytosolic environment should be studied further to detect differences in the control of gene expres- sion, post-translational modification, and turnover of the enzyme protein in a tissue-specific way. We have demonstrated with RBL2H3, an established cell line that expresses TPH in culture while retaining many of the characteristics of mast cells, that: (a) cellular TPH activity was seriously limited by insufficient supply with the enzyme’s essential cofactor, ferrous iron, and the substrates tryptophan and 6R-tetrahydrobiopterin [20]; (b) immune stimulation lead to a marked increase in TPH level by means of enhanced expression of the TPH gene [21]; and (c) the steady state TPH level of this cell was maintained at extremely low levels by rapid degradation of the enzyme (T 1/2 , 15–60 min) [22,23]. In the latter report, the turnover of TPH protein was shown to be driven by ATP-dependent action of 26S-proteasomes including, at least in part, ubiquitinylation of TPH. Furthermore, it was noted that this rapid turnover was suppressed by a protein kinase inhibitor. Since proteasomes might, in general, be ubiquit- ous in the cell, recognition of the specific target is crucial in terms of the specific protein to be digested. Poly-ubiquiti- nylation represents a major tag for proteasomes. The ubiquitinylation of a specific protein is determined by the ubiquitin ligase complex E3. The molecular basis of the structure–function relationship enabling E3 to specifically recognize a wide variety of substrates is one of the major subjects of investigation in this field. In the ubiquitinylation Correspondence to H. Hasegawa, Department of Bioscience, Teikyo University of Science and Technology, Uenohara, Yamanashi 409–0193, Japan. Fax: + 81 554 63 4431, E-mail: hasegawa@ntu.ac.jp Abbreviations: TPH, tryptophan hydroxylase; CaM kinase II, calcium/calmodulin-dependent protein kinase II; PKA, cyclic AMP-dependent protein kinase; 5HTP, 5-hydroxy- L -tryptophan. Enzyme: Tryptophan hydroxylase (EC 1.14.16.4). (Received 10 March 2002, revised 11 June 2002, accepted 19 August 2002) Eur. J. Biochem. 269, 4780–4788 (2002) Ó FEBS 2002 doi:10.1046/j.1432-1033.2002.03188.x of TPH, a specific tag might be required for targeting by the ligase. In many cases, phosphorylation of the target protein provides the tag for the ubiquitinylation system, especially of such families as the SCF-complex, Skp1/Cullin-1/F-box protein (reviewed in [24,25]). Involvement of phosphoryla- tion in TPH degradation was expected, however, phos- phorylation of non-neural TPH has never been demonstrated, although TPH of brain origin and recom- binant TPH have been known to be phosphorylated by PKA and by CaM kinase II [26–28]. On the other hand, proteasome-driven turnover has only been demonstrated with mast cell lines. The aim of this work was to elucidate whether the phosphorylation of non-neural TPH takes place and, if it does, whether it provides the tag for targeting by the proteasomes involved in the rapid turnover of the enzyme. MATERIALS AND METHODS Materials MG132 (carbobenzoxy-Leu-Leu-Leu-H) and E-64-d [( L - 3-trans-ethoxycarbonyloxirane-2-carbonyl)- L -leucine(3-meth- ylbutyl)amide] were purchased from Peptide Institute (Osaka), lactacystin from Kyowa Medex (Tokyo), okadaic acid and K252a from Alomone Labs (Jerusalem, Israel), and KN-62 from LC Laboratories (La ¨ ufelfingen, Switzer- land). Cycloheximide, creatine kinase, cyclic AMP-depend- ent protein kinase catalytic subunit from beef heart (Cat. No. P2645), rat liver phenylalanine hydroxylase (Cat. No. P6268), phosphocreatine, and sodium orthovanadate were purchased from Sigma. Sodium fluoride was obtained from Nacalai Tesque (Kyoto, Japan). The concentrations of inhibitors used in this study, MG132, E-64-d, lactacystin, okadaic acid, K252a, K252b, KN-62, and cycloheximide, were those that gave the maximum effect on evaluation. TPH was purified from P-815, a mouse mastocytoma, essentially according to Nakata and Fujisawa [8]. Rabbit polyclonal anti-TPH serum was raised against the purified TPH [13]. Bovine liver dihydropteridine reductase was purified up to the second ammonium sulfate fractionation step [29]. (6R)- L -erythro-5,6,7,8-Tetrahydrobiopterin was donated by Suntory (Tokyo, Japan). CaM kinase II and calmodulin, both purified from rat brain, were donated by T. Yamauchi (Department of Biochemistry, Faculty of Pharmaceutical Science, The University of Tokushima, Japan). [ 32 P]H 3 PO 4 (500 mCiÆmL )1 )and[c- 32 P]ATP (tetra- triethylammonium salt; 4500 CiÆmmol )1 ) were purchased from ICN Biochemicals. Cell culture RBL2H3, a mast cell line derived from rat basophilic leukemia cells, was obtained from The Japanese Cancer Research Resources Bank (Tokyo). RBL2H3 cells and FMA3 (Furth’s mastocytoma) cells were maintained as described [23]. One day before experiments, cells were plated to well of a 96-well culture plate (Falcon, Cat. No. 35072) at 1 · 10 5 cells per well. Two hours before the experimental treatment, cells were placed in serum-free medium buffered with 25 m M Hepes/NaOH containing 100 UÆmL )1 of penicillin and 100 lgÆmL )1 of streptomycin, then kept at 37 °C under 10% CO 2 /90% air throughout the experiments except at the time of manipulation. Agents of low solubility in water were dissolved in dimethylsulfoxide at a concentration 100-fold greater than final one used, unless otherwise stated, so that dimethylsulfoxide would be at an equivalent level in each experimental culture with no vehicle effect. Tryptophan hydroxylase assay TPH activity was determined essentially as described previously [13,23]. Cells in monolayer culture in wells of the96-wellplatewereplacedin20lLofNaCl/P i (–), then subjected twice to freezing in liquid nitrogen and thawing in water. Reaction mixtures for the cell-free treatment of purified TPH (phosphorylation and proteolysis as described below) were prepared just prior to measuring the enzyme activity. The disrupted cells or TPH mixture were preincu- bated for 15 min at 30 °Cin0.1 M Tris/HCl ( pH 8.0) containing 30 m M dithiothreitol, 50 l M Fe(NH 4 ) 2 (SO 4 ) 2 , and 4 mgÆmL )1 catalase in a total volume of 100 lL. Subsequently, 50 lL of another cocktail were added to afford a final reaction mixture of 250 l M tryptophan, 400 l M 6R-tetrahydrobiopterin, 500 l M NADH, 1 m M NSD-1015, 2 mgÆmL )1 catalase, and 50 lgÆmL )1 dihydrop- teridine reductase in 0.1 M potassium phosphate buffer ( pH 6.9). The enzyme reaction was allowed to proceed for 10 min at 30 °C, then was terminated by 1 M perchloric acid. The 5HTP formed was measured using an HPLC system equipped with a fluorescence monitor (JASCO model, FP920) set at 302 nm and 350 nm for excitation and emission, respectively. The solid phase was ODS (4.6 · 250 mm, JASCO, Finepak SIL-C18T5), the mobile phase was a 90 : 7 : 5 mixture of 40 m M sodium acetate (adjusted to pH 3.5 with formic acid), acetonitrile and methanol and the flow rate was 1 mLÆmin )1 [30]. Cell-free proteolysis of TPH Extracts from RBL2H3 cells as the source of proteasomes were prepared essentially as described [23]. The cells were homogenized in 5 volumes of 50 m M Tris/HCl (pH 7.5) containing 1 m M dithiothreitol, 2 m M ATP, and 0.25 M sucrose using an Ultra-disperser (model T25; IKA Labor- technik, Staufen, Germany). The homogenate was centri- fuged at 18 000 g for 5 min. In vitro proteolysis was performed in a reaction mixture containing the RBL2H3 cell extracts, 5 m M MgCl 2 ,1m M CaCl 2 ,2m M ATP, 10 lgÆmL )1 creatine kinase, 10 m M phosphocreatine, 0.2 mgÆmL )1 catalase, and 1 m M dithiothreitol in 50 m M Tris/HCl (pH 8.0). Purified TPH from P-815 cells with or without in vitro phosphorylation was used as the sub- strate. Inhibitors of proteasomes and protein kinases were added prior to addition of the substrate TPH. Aliquots were taken after various intervals of incubation (30 °C) for the TPH enzyme activity assay and for Western blot analysis. Phosphorylation of TPH In situ phosphorylation of TPH in FMA3 cells was performed as follows. Cells (2 · 10 6 cells) were adapted to phosphate-free RPMI 1640 (Gibco, Cat. No. 11877–032) supplemented with 5 l M NaH 2 PO 4 for 90 min. Ó FEBS 2002 TPH phosphorylation as proteasome targeting (Eur. J. Biochem. 269) 4781 Subsequently, cells were fed 0.4 mCiÆmL )1 [ 32 P]NaH 2 PO 4 for 30 min. 32 P-Loading was further continued for 120 min in the presence of protein-kinase inhibitors or protein- phosphatase inhibitors. Cells were then rinsed with NaCl/P i and disrupted with 1% NP-40 in 50 m M Tris/HCl ( pH 7.8) containing an inhibitor cocktail (1 m M phenyl- methanesulfonyl fluoride, 2 m M EDTA, 50 m M sodium fluoride, and 1 m M sodium orthovanadate). The cell lysates were mixed with anti-TPH serum (10 lL) and left overnight at 4 °C with agitation. Total IgG was collected by the addition of staphylococcal ghosts (Pansorbin; Calbiochem, La Jolla, CA, USA) as a precipitant, solubilized in 1% SDS, and subjected to SDS/PAGE. Cell-free phosphorylation by PKA was carried out for 30 min at 37 °Cinareaction mixture containing 3 lg of purified TPH as substrate, or rat liver phenylalanine hydroxylase for comparison, 1 lgPKA catalytic subunit and 2 lCi [c- 32 P]ATP in 50 m M Tris/HCl ( pH 7.4) containing 20 l M ATP and 10 m M MgCl 2 in a total volume of 210 lL. For SDS/PAGE, proteins were precipitated by the addition of trichloroacetic acid (5%) in the cold and centrifuged. The pellets were then washed twice with 400 lL of diethylether, dried and dissolved in 50 lLof the lysis buffer for SDS/PAGE. Ca 2+ /calmodulin-depend- ent phosphorylation was carried out with 1 lgTPHas substrate and 0.1 lg CaM kinase II for 30 min at 37 °Cin the presence of 0.1 l M calmodulin, in 210 lLof50m M Tris/HCl ( pH 7.4) containing 10 l M ATP (2 lCi [c- 32 P]ATP), 5 m M MgCl 2 ,120l M CaCl 2 ,and100l M EGTA. Aliquots were taken for the assay of TPH activity or for subjecting to the cell-free proteolysis described above. The remaining reaction mixture was mixed with affinity gel beads DMPH 4 -Affigel-10 [8] for collecting TPH in the presence of the inhibitor cocktail as above and 150 m M NaCl in 50 m M Tris/acetate (pH 8.0), then left overnight at 4 °C with agitation. The proteins obtained were subjected to SDS/PAGE followed by immunoblotting and autoradio- graphy. SDS/PAGE, Western blot analysis, and autoradiography Monolayer cultures washed with NaCl/P i or proteins collected as a pellet as described above were solubilized in 1% SDS and subjected to SDS/PAGE according to Laemmli [31]. Western blot analysis was performed as described previously [23]. The protein signal was visualized using an enhanced chemiluminescence detection system (ECL; Amersham, Buckinghamshire, England). Protein bands were exposed to an X-ray film (Konica, Medical Film 20287). For autoradiography with 32 P, gels following SDS/ PAGE were dried on filter paper, then subjected to exposure either to an X-ray film (Konica) at )80 °Cfor3dayswith an intensifying screen (Kodak, Bio Max MS) or to a fluoro- image analyser (Fujifilm, FLA-3000) using an imaging plate (Fujifilm, BSA-IP MS2040). Graphic images of Western blot analysis or autoradiograms were analyzed using NIH IMAGE ver 1.62 software, Wayne Rasband, National Institute of Health, USA. Other methods Proteins were determined by Bradford’s method [32] using bovine serum albumin as the standard. Data were expressed as means ± SD (n ¼ 4) unless otherwise stated. RESULTS Involvement of protein phosphorylation in TPH degradation in living cells In previous works [22,23], we demonstrated in mast cell lines RBL2H3 and FMA3 that de novo biosynthesis of TPH enzyme protein was accompanied by rapid degra- dation with 26S-proteasomes and that ubiquitinylation of TPH protein was involved in the process, presumably by providing the targeting tag. In search for any connection between protein phosphorylation and TPH turnover, we examined the effect of okadaic acid, a protein phospha- tase inhibitor, on TPH degradation in the living cell system and compared it with those of protease inhibitors (Fig. 1). When protein synthesis was arrested by cycloh- eximide (10 lgÆmL )1 ), a rapid decrease in TPH activity was observed in the absence of the inhibitors (T 1/2 : around 30 min, Fig. 1A). This decrease was much slower or was virtually stopped by proteasome inhibitors MG132 and lactacystin but was not affected by a cystein protease inhibitor, E-64-d; a representative finding showing that the steady state level of TPH was determined by a proteasome-driven degradation process. TPH degradation in the cells was accelerated by okadaic acid (0.25 l M ): the half-life of TPH (T 1/2 ) were estimated to be 29 min and 38 min in the presence and absence of okadaic acid, respectively (Fig. 1A, OA), suggesting an involvement of TPH phosphorylation in recognition of the enzyme by the ubiquitinylation system. Based on this observation, we examined whether TPH was phosphorylated in situ using FMA3 cells in which cytosolic TPH was also rapidly degradated by the proteasome-driven process while the steady state TPH level was roughly 20-fold higher than that of RBL2H3 cells [22,23]. Cellular proteins were labeled by incubating the cells with [ 32 P]orthophosphate, and steady state phosphorylation levels of proteins were performed in the presence and absence of okadaic acid and/or MG132. By Western blot analysis of the whole cell extracts, TPH of molecular mass 53 kDa was locali- zed side-by-side with purified TPH and the anti-TPH serum (Fig. 1B, WB). Addition of both okadaic acid and MG132 caused the immunoreactive band to be twofold thicker than the control band (see plot profiles of WB, right-most patterns), however, no discrete bands of 32 P-incorporation were recognized over the dense back- ground by autoradiography of this blot membrane. In order to concentrate the proteins of interest, immunopre- cipitation of the same cell extracts with the anti-TPH serum was performed before SDS/PAGE as described in Materials and methods. Even after the immunoprecipita- tion, 32 P-labeled TPH-like protein was not detected (lane 1inFig.1B, 32 P), indicating a very low steady state level of phosphorylated TPH or none at all. Addition of either okadaic acid (lane 2 in Fig. 1B, 32 P)orMG132(not shown) made little difference. By simultaneous addition of okadaic acid (0.5 l M ) and MG132 (3 l M ), a protein band of 53 kDa became detectable among several inten- sified proteins (lane 3). We conducted an Ôimage-mathÕ operation to obtain clearer difference by subtracting the image of lane 2 (okadaic acid alone) from the image of lane 3 (okadaic acid plus MG132). A clear band of 32 P- incorporated protein with a molecular mass of 53 kDa 4782 Y. Iida et al. (Eur. J. Biochem. 269) Ó FEBS 2002 was obtained (Fig. 1B, 32 P, lane 4) and was coincident with the TPH visualized by Western blot analysis (Fig. 1B, WB, lane 3 and 4). This operation visualizes 32 P-incorporation into the specific proteins which were protected from proteasome-driven digestion by MG132 among those 32 P-phosphorylated and protected from dephosphorylation by okadaic acid, proteins which oth- erwise would have been digested by proteasomes. Thus the phosphorylated form of TPH was detectable only when the proteasome action and phosphatase were effectively blocked (lane 3 in both Fig. 1B, 32 P and WB). Together with the fact that the blocking of protein phosphatase by okadaic acid resulted in the acceleration of TPH degradation (Fig. 1A, OA), the present result is evidence that phosphorylation takes place on this protein where it functions as the tag for the targeting of TPH by the proteasomes. It was noteworthy that TPH detectable under steady state conditions was unphosphorylated, presumably because the phosphorylated TPH might have been digested away in the absence of proteasome inhibitors (lane 1 in Fig. 1B, 32 P vs. WB). Inhibition of TPH degradation in the cell-free proteasome system by protein kinase inhibitors We examined the involvement of TPH phosphorylation in proteasome-driven degradation of the enzyme in vitro.Our system contained extracts of RBL2H3 cells as the source of proteasomes and ubiquitinylating enzymes [23], and purified TPH from mouse mastocytoma P-815 cells as the substrate for proteolysis. When incubated under complete conditions, the amount of TPH protein of 53 kDa decreased progres- sively concomitantly with the enzyme activity (Fig. 2A,B). Along with the decrease in TPH protein, other TPH-like immunoreactive polypeptide bands of smaller molecular mass were not present; this was in contrast to TPH digestion in cell extracts without ATP, which presumably occurred by the action of lysozomal enzymes [33]. Taken together, our system was representative of a TPH-proteasome system with regard to the sensitivity to specific proteasome inhibitors (Fig. 2A). Degradation of TPH in this cell-free system was inhibited by KN-62 (100 l M ), a potent inhibitor of CaM kinase II (Fig. 2A for TPH activity and Fig. 2B, Fig. 1. Effect of protein phosphatase inhibitor, okadaic acid, and proteasome inhibitors on TPH degradation in RBL2H3 cells and on TPH phos- phorylation in FMA3 cells. (A) RBL2H3 cells (1 · 10 5 cells per well) were cultured in the presence (closed circle) or absence (vehicle, open circle) of respective inhibitors for 60 min to allow the reagents to penetrate the cells. Then, 10 lgÆmL )1 cycloheximide were added to each well and the culture continued. TPH activities at 0, 10, 30, and 60 min after addition of cycloheximide were measured. Inhibitors used were: 10 l M MG132 (A-MG132), 30 l M lactacystin (A-Lact), 10 l M E-64-d (A-E64d), and 0.25 l M okadaic acid (A-OA). Values are means ± SD (n ¼ 4). (B) FMA3 cells (2 · 10 6 cellsÆmL )1 ) were placed in phosphate-free RPMI 1640 medium supplemented with 5 l M NaH 2 PO 4 for 1.5 h, and were exposed to 0.4 mCiÆmL )1 32 P-phosphate. After 30 min exposure, the cells were further treated with okadaic acid and/or MG132 for the next 2 h. The cells were then disrupted with 1% NP-40 in 50 m M Tris/HCl (pH 7.8) containing the protease inhibitor cocktail. ( 32 P): In order to examine 32 P-incorporation into proteins, the cell extracts were subjected to immunoprecipitation followed by SDS/PAGE for autoradiography as described in Materials and methods. The autoradiogram (left panel, 32 P) represents treatment with vehicle (lane 1), 0.5 l M okadaic acid (lane 2), and okadaic acid plus 3 l M MG132 (lane 3). Lane 4 is a graphically created image representing the net increase in the density of lane 3 over lane 2, which was obtained by image math operation (image 3 minus image 2). To create this image, the contrast was pushed twice (pixel densities were doubled). The arrow indicates the expected migration of TPH with a molecular mass of 53 kDa. The adjacent pattern (PP) is the vertical Ôplot profileÕ of lane 4. (WB): The same extracts described in ( 32 P) incorporation were applied (15 lg protein per lane) to SDS/PAGE without the immunoprecipitation, then subjected to Western blot analysis as described in Materials and methods. Lanes 1–3 corresponded to those in the ( 32 P) panel, and lane 4 was the purifed TPH (2.7 ng) as a reference marker for immunostaining. The right-most patterns are the vertical Ôplot profileÕ of lanes 1–4. The Ôimage-mathÕ and Ôplot profileÕ operations were conducted using NIH - IMAGE ver 1.62 software. Ó FEBS 2002 TPH phosphorylation as proteasome targeting (Eur. J. Biochem. 269) 4783 KN62 for protein analysis by Western blotting). KN-62 at 50 l M was also significantly effective, while at concentra- tions higher than 100 l M , the inhibitor showed no further effect (not shown). K252a and K252b (both 50 l M ), protein kinase inhibitors with relatively broad specificity, were also effective as tested (Fig. 2B, K252a), but H-89 was not significantly effective at 50 l M (Fig. 2B, H89). Staurospo- rine and chelerythrine, potent inhibitors of protein kinase C, were not significantly effective (not shown). These results indicated that TPH digestion by proteasomes involved phosphorylation of certain protein(s) as an essential step. Taken together with the outcome of the experiment in Fig. 1, it is likely that the TPH molecule was the protein to be phosphorylated in the selective degradation. Further- more, from the specificity of the protein kinase inhibitors tested above, CaM kinase II seemed to function in the cell extracts, at least in part. Stimulation of TPH degradation in the cell free proteasome system by the phosphorylation of the enzyme by CaM kinase II As described above (Fig. 1), the phosphorylation of non- neuronal TPH was suggested for the first time using RBL2H3 and FMA3 cells. We then examined TPH phosphorylation in vitro in order to determine the type of protein kinase responsible. TPH from brain is known to be phosphorylated by PKA [26,28] and CaM kinase II [27]. Involvement of CaM kinase II was also suggested by results of the experiments shown in Fig. 2. These kinases were tested for the phosphorylation of TPH purified from mastocytoma P-815 cells. Since the available protein specimens contain unidentified proteins, electrophoreto- grams were carefully compared among images visualized by Coomassie Brilliant Blue staining (CBB), by Western blotting using anti-TPH serum (WB), and by autoradio- graphy ( 32 P). PKA was first examined using its catalytic subunit for phosphorylation of TPH. In this experiment, the amount of the PKA-catalytic subunit and that of the other agents added were optimized using rat phenylalanine hydroxylase (PAH, molecular mass of 55 kDa in Fig. 3A). Virtually no phosphorylation of TPH was detected under the conditions used in which phenylalanine hydroxylase (and two other proteins, molecular masses of 95 kDa and 60 kDa that contaminated the preparation) was clearly phosphorylated (Fig. 3A). When an effort was made to enhance the autoradiograph image ( 32 P panel in Fig. 3B), a faint and diffuse signal appeared at a slightly higher position around the TPH region but the signal was too small to determine whether it corresponded to TPH protein (major band in panel WB, and the major band in panel CBB with molecular mass of 53 kDa). The only obviously labeled band (arrowhead, molecular mass of 41 kDa, also seen on the 32 P panel of Fig. 3A) was judged to be from contami- nation of the PKA preparation because this band was not detectable with Western blot analysis (WB in Fig. 3B) and was seen with CBB staining only when PKA was added (CBB panel in Fig. 3B; note that PKA alone had poor recovery through trichloroacetic acid precipitation and diethylether washing to remove trichloroacetic acid for sampling). These results indicated that TPH protein incor- porated virtually no 32 P or far less than the stoichiometric amount of 32 P. On the other hand, TPH was clearly 32 P-labeled by CaM kinase II in vitro in the presence of both Ca 2+ and calmodulin (Fig. 4). Besides TPH (molecular mass of 53 kDa), two diffuse bands appeared on autoradiography (Fig. 4A, lanes 2 and 3). These were thought to be due to autophosphorylation of CaM kinase II proteins of 54 kDa and 63 kDa described by Yamauchi & Fujisawa [34]. This Fig. 2. Inhibitory effect of protein kinase inhibitors on degradation of purified TPH in the cell-free proteasome system. TPH (1 lg) purified from mastocytoma P-815 cells was subjected to a cell-free proteasome system composed of freshly prepared extracts (400 lg protein). The reaction mixture (total volume of 210 lL) was incubated at 30 °C, and aliquots were taken at the indicated times for the TPH activity assay (A) and for Western blot analysis (B). TPH activity was measured as described in Materials and methods after appropriate dilution (200-fold). Preparation of RBL2H3 extracts, composition of the reaction mixtures, and analytical procedures are described in Materials and methods. (A) TPH activity remained after incubation for the indicated times in the absence (open circles) or presence of MG132 (50 l M , closed squares) or KN-62 (100 l M , closed circles). (B) TPH protein remained after the indicated period of incubation. Upper panels: immunoblot images visualized with anti-TPH serum. Lower panels: the digitized densities of corresponding spots were plotted relative to the 0-time density as 100%. The in vitro proteasome system included a vehicle control (open circles in all panels), 50 l M MG132 (closed squares for KN62), 100 l M KN-62 (closed circles for KN62), 50 l M K252a (closed circles for K252a), and 50 l M H-89 (closed circles for H89). 4784 Y. Iida et al. (Eur. J. Biochem. 269) Ó FEBS 2002 phosphorylation was prevented by 50 l M KN-62 (not shown). In addition, PKA run for comparison again gave no indication of phosphorylation of TPH (lane 4, 32 P-panel). These results indicated that TPH purified from mouse mastocytoma P-815 was readily phosphorylated by CaM kinase II. Although the TPH from P-815 cells was easily phosphorylated by this kinase, its enzymic activity was not altered by the phosphorylation reaction. We examined the effect of TPH phosphorylation by CaM kinase II on the susceptibility of the enzyme to degradation in our cell-free proteasome system. As shown in Fig. 4B, degradation of TPH phosphorylated by CaM kinase II prior to the proteasome reaction was much more rapid than that of the nonphosphorylated TPH. This is presumably because the nonphosphorylated TPH had to undergo prior phosphorylation in situ to be targeted by the proteasomes in the reconstituted cell-free system. DISCUSSION In the present study, using RBL2H3 and FMA3 cells as representative non-neural cells, we examined whether TPH is actually phosphorylated and whether phosphorylation is the prerequisite step in the proteasome-driven TPH degra- dation process. We presented evidence that: (a) TPH in FMA3 cells was phosphorylated in vivo; (b) TPH purified from mastocytoma P-815 cells was also phosphorylated in vitro by CaM kinase II but not by PKA; (c) TPH thus phosphorylated was degraded in vitro at a higher rate than was the nonphosphorylated TPH; and (d) living RBL2H3 cells are furnished with a whole proteasome system inclu- ding 26S-proteasomes and a specific ubiquitinylation system that recognizes phosphorylated TPH. As to non-neural TPH, rat pineal enzyme was reported to increase in enzymic activity by treatment with cAMP [35]. The authors, Fig. 3. Insignificant incorporation of 32 P into TPH by protein kinase A. TPH or rat liver phenylalanine hydroxylase (PAH), 3 lg each, were exposed to 1 lg PKA-catalytic subunit in the presence of 2 lCi [c- 32 P] ATP at 37 °C for 30 min. Proteins were precipitated by trichloroacetic acid (5%) on ice, centrifuged, washed with diethylether, and then subjected to SDS/PAGE followed by autoradiography for 32 P incorporation or Western blot analysis using anti-TPH serum. Compositions of the reaction mixtures, and analytical procedures are described in Materials and methods. Size markers are shown in kDa at left. Arrows indicate the estimated position of TPH and the arrowheads indicate contaminating protein in the PKA preparation for comparison between panels. (A) Autoradiogram of phosphorylation products of TPH and PAH (1 lg per lane assuming protein recovery to be consistent in washing procedure). (B) Comparison of protein staining (CBB), Western blot analysis (WB), and autoradiography ( 32 P) with a common gel after SDS/PAGE. Fig. 4. Phosphorylation of TPH and stimulation of its proteasome driven degradation by CaM kinase II. TPH (3 lg) was placed in phosphorylation conditions consisting of CaM kinase II, Ca 2+ andcalmodulinat37°Cfor30mininatotalvolumeof210lL. Exclusion of kinase or the PKA catalytic subunit was also used for comparison. (A) Proteins in the reaction mixture were collected by preferential adsorption of TPH to DMPH 4 - conjugated Affigel-10 gel beads overnight with agitation. Proteins were then extracted from the gel with 1% SDS for SDS/PAGE, followed by (WB) Western blot analysis using anti-TPH serum as the primary antibody and by ( 32 P) autoradiography using a FLA-3000 fluorescence image analyzer (details are described in Materials and methods). Size markers are shown in kDa at left. The arrow on the right indicates the estimated position of TPH. The two right-most patterns (PP) are the Ôplot profileÕ of lanes 2 and 3 of ( 32 P) panels, respectively. (B) TPH in the phosphorylation reaction mixture including CaM kinase II (closed circles in left panel are marked (+) in right panel) or lacking kinase (open circles in left panel are marked (–) in right panel) was subjected to cell-free proteolytic conditions for the indicated period of time, as described in the legend to Fig. 2. TPH remaining undigested was visualized by Western blot immunostaining after SDS/PAGE (WB). The left panel represents the density of the remaining TPH relative to the 0-time density taken as 100. Ó FEBS 2002 TPH phosphorylation as proteasome targeting (Eur. J. Biochem. 269) 4785 however, did not observe phosphorylation of pineal TPH protein, though they described PKA-dependent phosphory- lation of brain TPH. The pineal gland is anatomically classified as being outside the central nervous system and the enzymic properties of pineal TPH were obviously peripheral nature in every aspect we examined [6,36]. Careful investi- gation of TPH from mouse mastocytoma P-815 failed to uncover any activation of enzyme activity by cAMP- dependent protein kinase action [37]. Thus far, no clear evidence has appeared for phosphorylation of non-neural TPH, including that of pineal or neoplastic mastocytoma cells. Evidence of TPH phosphorylation in living cells We first observed that okadaic acid, a protein phosphatase inhibitor, accelerated the degradation of the enzyme, which was already quite rapid (Fig. 1A, OA). This fact suggested that phosphorylation of certain proteins stimulates their degradation. TPH was the protein most likely to be phosphorylated, however, this meant that phosphorylated TPH would be hardly detected unless the proteasome action was blocked. Indeed, in the absence of proteasome inhibitor, phosphorylated TPH was not detectable in the steady state labeling experiment where numerous cellular proteins were 32 P-labeled as shown in Fig. 1B, 32 P.Evenin the presence of either okadaic acid (lane 2) or MG132 (not shown), 32 P-labeled TPH was undetectable, indicating that the rate of either dephosphorylation or proteolytic degra- dation of the putative 32 P-TPH was higher than phosphory- lation of TPH. However, phosphorylated ( 32 P-labeled) TPH-like protein became detectable in FMA3 cells only when both processes were blocked by simultaneous addition of okadaic acid, a protein phosphatase inhibitor, and MG132, a proteasome inhibitor (Fig. 1B, 32 P,lane3).The ‘image math’ operation visualized the net increase in density in lane 3 over lane 2 (Fig. 1B, 32 P, lane 4). This image represented protein bands that fulfilled the following three conditions simultaneously: (a) phosphorylation by cellular protein kinases; (b) protection from dephosphorylation by okadaic acid; and (c) protection by MG132 from protea- some action. A protein band (molecular mass of 53 kDa) had the highest intensity and coincided with that of TPH visualized by Western blot analysis of the same cell extracts run with the authentic TPH protein as the staining reference mark (Fig. 1B, WB). This suggests that the proteasome inhibitor MG132 might accumulate phosphorylated, sub- sequently ubiquitinylated TPH, the presumable substrate of the proteasomes. The experimental result was that MG132 administered to living cells somehow raised TPH activity and increased the amount of TPH-like protein of molecular mass of 53 kDa (Fig. 1A, MG132 and 1B, WB lane 3), suggesting that de-ubiquinylating enzyme is considerably active [23]. Based on this outcome, phosphorylation of non- neural TPH and its role as an essential tag for protein degradation were explored. Phosphorylation of non-neural TPH in vitro Purified TPH from mastocytoma P-815 cells, i.e. TPH of non-neural origin, was demonstrated to be phosphorylated by CaM kinase II in vitro (Fig. 4). Although neural TPH and recombinant enzymes were reportedly phosphorylated by PKA (reviewed in [38]), in the present study, we could not obtain positive evidence for PKA phosphorylation of the TPH from P-815 (Fig. 3). A possibility remains that our TPH preparation was fully phosphorylated at the PKA- specific phosphorylation site before isolation and therefore left no room for further phosphorylation. Indeed, phenyl- alanine hydroxylase purified from rat liver contained endogenously phosphorylated subunits with 1.3 mol of phosphate per tetramer and was fully phosphorylated in vitro to give 1 mol per subunit by the catalytic subunit of PKA [39]. This possibility is difficult to rule out before direct measurement of endogenous phosphate [40]. The present observation that TPH protein incorporated 32 P- phosphate by the action of CaM kinase II but not by PKA does not appear compatible with the idea that the site was shared with PKA and CaM kinase II as is Ser58 of rat TPH [28]. For a solid conclusion, however, further investigation is required, such as site determination of kinase-specific phosphorylation. Phosphorylation of TPH as the tag for proteolysis targeting The role of TPH phosphorylation is that it down regulates the TPH level in the cell by serving a tag for targeting for digestion by proteasomes. This idea is based on the following observations: (a) proteasome-driven TPH degra- dation in vivo (in RBL2H3 cells) was enhanced by okadaic acid, a protein phosphatase inhibitor (Fig. 1); (b) degrada- tion of exogenous TPH in a reconstituted proteasome system composed of RBL2H3 cell extracts was inhibited by both KN-62, a CaM kinase II inhibitor, and K252a, a potent protein kinase inhibitor with broad specificity (Fig. 2); and (c) TPH previously phosphorylated by CaM kinase II was more rapidly degraded than nonphosphoryl- ated TPH in the same in vitro system. The cell-free proteasome system employed in the present study was constructed with fresh extracts from RBL2H3 cells cultured under ordinary conditions and prepared in the presence of 2m M ATP, 1 m M dithiothreitol and 0.25 M sucrose to minimize the dissociation of proteasomes and possible disruption of lysosomes [41]. Based on the sensitivity to protein kinase inhibitors and to proteasome inhibitors, it is obvious that the cell-free proteasome system per se included a relevant protein kinase system for TPH. Although the endogenous protein kinase shares properties with CaM kinase II in terms of sensitivity to inhibitors, the presence of multiple kinase species was also possible since KN-62 alone did not completely prevent the proteolysis, even at concen- trations higher than 100 l M , while K252a of broad specificity did. It was noteworthy that H-89, chelerythrine and staurospoline were not effective in preventing TPH digestion, suggesting little contribution from PKA or PKC as far as the proteasome system in the RBL2H3 cells was concerned. Our observations supported the idea that phosphoryla- tion is a prerequisite for proteasome digestion of non- neural TPH. This phosphorylation enables the cell to severely down-regulate the enzyme level by means of stimulation of proteasome-driven degradation. This is a new role for TPH phosphorylation, which does not necessarily alter its enzyme activity. It will be interesting to learn whether the phosphorylation of neural TPH also has 4786 Y. Iida et al. (Eur. J. Biochem. 269) Ó FEBS 2002 an association with proteasome-dependent degradation, but there is as yet little information about TPH turnover in the central nervous system. In addition, the question remains as to how neural and non-neural TPH can appear so differentiated in view of their common amino acid sequence [19]. Studies on the cellular management of their common TPH gene, including RNA-processing, transla- tion of the message, post-translational processing, and cytosolic machinery such as phosphorylation and proteasome systems are just beginning to elucidate the differences between these two types of TPH. ACKNOWLEDGEMENTS We are grateful to Dr T. Yamauchi for donating purified calmodulin and CaM kinase II and for instructions on its use. This research was supported by the Japan Private School Promotion Foundation and by a Grant-in-Aid for Advanced Scientific Research on Bioscience/ Biotechnology Areas from the Ministry of Education, Science, Sports and Culture of Japan. REFERENCES 1. Kaufman, S. & Fisher, D.B. (1974) Pterin-dependent aromatic amino acid hydroxylases. In Molecular Mechanisms of Oxygen Activation (Hayaishi, O. ed.), pp. 285–369. Academic Press Inc., New York and London. 2. Hosoda, S. & Glick, D. (1965) Biosynthesis of 5-hydro- xytryptophan and 5-hydroxytryptamine from tryptophan by neoplastic mouse mast cells. Biochim. Biophys. Acta. 111, 67–78. 3. Grahame, S.D. (1967) The biosynthesis of 5-hydroxytryptamine in brain. Biochem. J. 105, 351–360. 4. Lovenberg, W., Jequier, E. & Sjoerdsma, A. (1967) Tryptophan hydroxylation: measurement in pineal gland, brainstem, and car- cinoid tumor. Science. 155, 217–219. 5. Ichiyama, A., Nakamura, S., Nishizuka, Y. & Hayaishi, O. (1970) Enzymic studies on the biosynthesis of serotonin in mammalian brain. J. Biol. Chem. 245, 1699–1709. 6. Ichiyama, A., Hasegawa, H., Tohyama, C., Dohmoto, C. & Kataoka, T. (1976) Some properties of bovine pineal tryptophan hydroxylase. Adv. Exp. Med. Biol. 74, 103–117. 7. Hosoda, S. (1975) Further studies on tryptophan hydroxylase from neoplastic murine mast cells. Biochim. Biophys. Acta. 397, 58–68. 8. Nakata, H. & Fujisawa, H. (1982) Tryptophan 5-monooxygenase from mouse mastocytoma P815. A simple purification and general properties. Eur. J. Biochem. 124, 595–601. 9. Tong, J.H. & Kaufman, S. (1975) Tryptophan hydroxylase. Purification and some properties of the enzyme from rabbit hindbrain. J. Biol. Chem. 250, 4152–4158. 10. Nakata, H. & Fujisawa, H. (1982) Purification and properties of tryptophan 5-monooxygenase from rat brain-stem. Eur. J. Bio- chem. 122, 41–47. 11. Cash,C.D.,Vayer,P.,Mandel,P.&Maitre,M.(1985)Trypto- phan 5-hydroxylase. Rapid purification from whole rat brain and production of a specific antiserum. Eur. J. Biochem. 149, 239–245. 12. Kuhn, D.M., Meyer, M.A. & Lovenberg, W. (1980) Comparisons of tryptophan hydroxylase from a malignant murine mast cell tumor and rat mesencephalic tegmentum. Arch. Biochem. Biophys. 199, 355–361. 13. Hasegawa, H., Yanagisawa, M., Inoue, F., Yanaihara, N. & Ichiyama, A. (1987) Demonstration of non-neural tryptophan 5-mono-oxygenase in mouse intestinal mucosa. Biochem. J. 248, 501–509. 14. Grenett, H.E., Ledley, F.D., Reed, L.L. & Woo, S.L.C. (1987) Full-length cDNA for rabbit tryptophan hydroxylase: functional domains and evolution of aromatic amino acid hydroxylase. Proc. Natl Acad. Sci. USA 84, 5530–5534. 15. Darmon, M.C., Guibert, B., Leviel, V., Ehret, M., Maitre, M. & Mallet, J. (1988) Sequence of two mRNAs encoding active rat tryptophan hydroxylase. J. Neurochem. 51, 312–316. 16. Boularand, S., Darmon, M.C., Ganem, Y., Launay, J.M. & Mallet, J. (1990) Complete coding sequence of human tryptophan hydroxylase. Nucleic Acid. Res. 18, 4257. 17. Stoll, J., Kozak, C.A. & Goldman, D. (1990) Characterization and chromosomal mapping of a cDNA encoding tryptophan hydro- xylase from a mouse mastocytoma cell line. Genomics. 7, 88–96. 18. Tipper, J.P., Citron, B.A., Ribero, P. & Kaufman, S. (1994) Cloning and expression of rabbit and human brain tryptophan hydroxylase cDNA in Escherichia coli. Arch. Biochem. Biophys. 315, 445–453. 19. Kim, K.S., Wessel, T.C., Stone, D.M., Carver, C.H., Joh, T.H. & Park, D.H. (1991) Molecular cloning and characterization of cDNA encoding tryptophan hydroxylase from rat central sero- tonergic neurons. Brain Res. Mol. Brain Res. 9, 277–283. 20. Hasegawa, H., Oguro, K., Naito, Y. & Ichiyama, A. (1999) Iron dependence of tryptophan hydroxylase acitivity in RBL2H3 cells and its manipulation by chelators. Eur. J. Biochem. 261, 734–739. 21. Hasegawa, H., Kojima, M., Iida, Y., Oguro, K. & Nakanishi, N. (1996) Stimulation of tryptophan hydroxylase production in a serotonin producing cell line (RBL2H3) by intracellular calcium mobilizing reagents. FEBS Lett. 392, 289–292. 22. Hasegawa, H., Kojima, M., Oguro, K. & Nakanishi, N. (1995) Rapid turnover of tryptophan hydroxylase in serotonin producing cells: demonstration of ATP-dependent proteolytic degradation. FEBS Lett. 368, 151–154. 23. Kojima, M., Oguro, K., Sawabe, K., Iida, Y., Ikeda, R., Yamashita, A., Nakanishi, N. & Hasegawa, H. (2000) Rapid turnover of tryptophan hydroxylase is driven by proteasomes in RBL2H3 cells, a serotonin producing mast cell line. J. Biochem. (Tokyo) 127, 121–127. 24. Nakayama, K. & Nakayama, K. (2001) SCF complex regulating a variety of cellular functions. Exp Med. (Japanese) 19, 132–141. 25. Hershko, A. & Ciechanover, A. (1998) The ubiquitin system. Annu. Rev. Biochem. 67, 425–479. 26. Johansen, P.A., Jennings, I., Cotton, R.G. & Kuhn, D.M. (1996) Phosphorylation and activation of tryptophan hydroxylase by exogenous protein kinase A. J. Neurochem. 66, 817–823. 27. Yamauchi, T. & Fujisawa, H. (1983) Purification and character- ization of the brain calmodulin-dependent protein kinase (kinase II), which is involved in the activation of tryptophan 5-mono- oxygenase. Eur. J. Biochem. 132, 15–21. 28. Kuhn, D.M., Arthur, R. Jr & States, J.C. (1997) Phosphorylation and activation of brain tryptophan hydroxylase: identification of serine-58 as a substrate site for protein kinase A. J. Neurochem. 68, 2220–2223. 29. Hasegawa, H. (1977) Dihydropteridine reductase from bovine liver. Purification, crystallization, and isolation of a binary complex with NADH. J. Biochem. (Tokyo) 81, 169–177. 30. Hasegawa, H. & Ichiyama, A. (1987) Tryptophan 5-mono- oxygenase from mouse mastocytoma: high-performance liquid chromatography assay. Methods Enzymol. 142, 88–92. 31. Laemmli, U.K. (1970) Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature. 227, 680–685. 32. Bradford, M.M. (1976) A rapid and sensitive method for the quantitation of microgram quantities of protein utilizing the principle of protein-dye binding. Anal. Biochem. 72, 248–254. 33. Hasegawa, H., Kojima, M., Oguro, K., Watabe, S. & Nakanishi, N. (1995) Rapid turnover of tryptophan hydroxylase: Demon- stration of proteolytic process in cell-free system. Pteridines. 6, 138–140. Ó FEBS 2002 TPH phosphorylation as proteasome targeting (Eur. J. Biochem. 269) 4787 34. Yamauchi, T. & Fujisawa, H. (1985) Self-regulation of calmodu- lin-dependent protein kinase II and glycogen synthase kinase by autophosphorylation. Biochem. Biophys. Res. Commun. 129,213– 219. 35. Ehret, M., Pevet, P. & Maitre, M. (1991) Tryptophan hydroxylase synthesis is induced by 3¢,5¢-cyclic adenosine monophosphate during circadian rhythm in the rat pineal gland. J. Neurochem. 57, 1516–1521. 36. Ichiyama, A., Hori, S., Mashimo, Y., Nukiwa, T. & Makuuchi, H. (1974) The activation of bovine pineal tryptophan 5-mono- oxygenase. FEBS Lett. 40, 88–91. 37. Yanagisawa, M., Hasegawa, H. & Ichiyama, A. (1982) Trypto- phan hydroxylase from mouse mastocytoma P-815. Reversible activation by ethylenediaminetetraacetate. J. Biochem. (Tokyo) 92, 449–456. 38. Mockus, S.M. & Vrana, K.E. (1998) Advances in the molecular characterization of tryptophan hydroxylase. J. Mol. Neurosci. 10, 163–179. 39. Kaufman, S., Hasegawa, H., Wilgus, H. & Parniak, M. (1981) Regulation of hepatic phenylalanine hydroxylase activity by phosphorylation and dephosphorylation. Cold Spring Harbor Conf. Cell Proliferation. 8, 1391–1406. 40. Hasegawa, H., Parniak, M. & Kaufman, S. (1982) Determination of phosphate content of purified proteins. Anal. Biochem. 120, 360–364. 41. Ugai, S., Tamura, T., Tanahashi, N., Takai, S., Komi, N., Chung, C.H., Tanaka, K. & Ichihara, A. (1993) Purification and char- acterization of the 26S proteasome complex catalyzing ATP- dependent breakdown of ubiquitin-ligated proteins from rat liver. J. Biochem. (Tokyo) 113, 754–768. 4788 Y. Iida et al. (Eur. J. Biochem. 269) Ó FEBS 2002 . Proteasome-driven turnover of tryptophan hydroxylase is triggered by phosphorylation in RBL2H3 cells, a serotonin producing mast cell line Yoshiko Iida 1 ,. phosphorylated by this kinase, its enzymic activity was not altered by the phosphorylation reaction. We examined the effect of TPH phosphorylation by CaM kinase