Expression and Characterization of the Catalytic Core of Tryptophan Hydroxylase*

Wild type rabbit tryptophan hydroxylase (TRH) and two truncated mutant proteins have been expressed inEscherichia coli. The wild type protein was only expressed at low levels, whereas the mutant protein lacking the 101 amino-terminal regulatory domain was predominantly found in inclusion bodies. The protein that also lacked the carboxyl-terminal 28 amino acids, TRH102–416, was expressed as 30% of total cell protein. Analytical ultracentrifugation showed that TRH102–416 was predominantly a monomer in solution. The enzyme exhibited an absolute requirement for iron (ferrous or ferric) for activity and did not turn over in the presence of cobalt or copper. With either phenylalanine or tryptophan as substrate, stoichiometric formation of the 4a-hydroxypterin was found. Steady state kinetic parameters were determined with both of these amino acids using both tetrahydrobiopterin and 6-methyltetrahydropterin.

Tryptophan hydroxylase (TRH 1 , EC 1.14.16.4) carries out the 5-hydroxylation of tryptophan via the oxidation of tetrahydropterin and the reductive incorporation of molecular oxygen (Scheme 1). In mammalian metabolism the reaction catalyzed by TRH precedes ␣-decarboxylation and is believed to be the initial and rate-limiting process in the production of the neurotransmitter serotonin (5-hydroxytryptamine). Although TRH has been studied since the early 70s, enzymological characterization has been impeded by the limited quantity of active enzyme available from native or heterologous sources, the exceedingly low specific activity of the isolated enzyme, and the quite rapid decrease in activity observed during purification or storage (1)(2)(3)(4)(5)(6)(7)(8)(9).
TRH is a member of the small family of pterin-dependent aromatic amino acid hydroxylases that includes tyrosine hydroxylase (TYH) and phenylalanine hydroxylase (PAH). Each of these enzymes catalyzes the addition of an oxygen atom to the ring of an aromatic amino acid substrate. The bulk of what is currently known of the reaction mechanism of these enzymes has come from studies of the latter two (10). Both PAH and TYH require ferrous iron for activity (11,12); however, the exact role for the iron in catalysis is undefined. The primary structures of these enzymes are known from a variety of organisms. Sequence comparisons and deletion mutageneses have identified three functional regions: an amino-terminal regulatory domain, a catalytic domain, and a carboxyl-terminal interface (13)(14)(15)(16)(17). The regulatory domains of the three hydroxylases show no similarities, whereas the catalytic domains are homologous, with sequence identities of 32-75%. Enzymes lacking the regulatory domain are catalytically active (13,14,17). The carboxyl-terminal 24 amino acids of TYH form a long helix demonstrated to be responsible for the tetrameric structure of the enzyme (13,18); this helix is presumed to have a corresponding function in TRH and PAH.
We report here the purification and preliminary characterization of a mutant protein containing only the catalytic core of TRH. The rationale for the truncations was to increase the heterologous expression and/or stability of the enzyme by removing both the regulatory and interface domains. This doubly truncated form of the enzyme serves as the first viable model enzyme for detailed mechanistic studies of the catalytic reaction mechanism of TRH. The high specific activity of this enzyme has allowed the analysis of several fundamental properties of this important enzyme, including catalytic specificity and metal dependence.

EXPERIMENTAL PROCEDURES
Materials-Tryptophan, 5-hydroxytryptophan, and ␤-mercaptoethanol were obtained from Sigma. 6-Methyltetrahydropterin was synthesized according to Fitzpatrick (19). Tetrahydrobiopterin was purchased from Calbiochem. MES and dithiothreitol were purchased from Research Organics, Inc. Catalase was purchased from Boehringer Mannheim. Ceramic hydroxyapatite was from Bio-Rad, and Q-Sepharose was obtained from Amersham Pharmacia Biotech. Isopropyl-␤-thiogalactoside was from United States Biochemical Corp. Cuprous chloride, cupric sulfate, ferric sulfate, and ferrous ammonium sulfate were from Sigma. Cobalt chloride was purchased from Mallinkrodt. Agarose (SeaKem) was from FMC. [3,5-3 H]Tyrosine was from Amersham Pharmacia Biotech. Ultma polymerase and deoxynucleotides for polymerase chain reaction were obtained from Perkin-Elmer. Plasmid pET3d and Escherichia coli BL21 (DE) were obtained from Novagen. E. coli CJ236 was from Invitrogen. Plasmid pTZ-18R was from Amersham Pharmacia Biotech. Oligonucleotides were custom-synthesized using an Applied Biosystems model 380B synthesizer. Restriction and DNA modification enzymes were purchased from New England Biolabs. Plasmids were purified using the Qiagen midi-prep plasmid preparation kit.
Vectors for TRH Expression-The construct for expression of wild type TRH was made by polymerase chain reaction subcloning from the rabbit cDNA-derived plasmid prbTRH479 (20). NcoI and BamHI restriction sites were incorporated at the 3Ј and 5Ј ends of the gene, respectively, via non-complementary oligonucleotide tails. The 1.33kilobase pair product was then digested and subcloned into pET3d and pTZ18R to form the wild type TRH constructs pEWOH2 and pWH1 for expression and mutagenesis, respectively. In pEWOH2 the start codon for the TRH gene is 6 bases from its ribosome binding site and 55 bases from the T7 promoter.
Deletion of the amino-terminal 101 amino acids of TRH was achieved using the Bio-Rad adaptation of the methods of Kunkel et al. (21) with single-stranded uracil-containing DNA derived from pWH1. The oligonucleotide 5Ј ATGAAGGAAGAAGCCATGGAGAGTGTTCCTTGGTT-TCCA 3Ј was used to incorporate an NcoI restriction site (in bold) into the TRH gene adjacent to position 101. The resulting plasmid was digested with NcoI and BamHI and ligated into pET3d to obtain pEWOH⌬101. The start codon/ribosome binding site/promoter relationship of pEWOH⌬101 was unchanged from that in pEWOH2.
Exclusion of the carboxyl-terminal interface helix from translation was achieved using a variation of the Stratagene quick change mutagenesis method. pEWOH⌬101 was used as a template for a polymerase chain reaction reaction in which two complementary oligonucleotides (5Ј GCCAAAAGCTAAACGAATGCCTAAAACGAGCTGC 3Ј and 5Ј GCAGCTCGTTTTAGGCATTCGTTTAGCTTTTGGC 3Ј) were used to mutate codons Ile 417 and Met 421 to stop codons (in bold). The resulting transformed BL21 (DE) cells were screened for expression of the doubly truncated protein using SDS-polyacrylamide gel electrophoresis. The mutation was confirmed by sequencing the entire gene of a plasmid from a cell line that expressed active TRH 102-416 ; this was designated pEWOH⌬101⌬H.
Protein Expression and Purification-Aliquots from frozen cell stocks were plated (240 l/liter of culture) on LB agar (100 g/ml carbenicillin). After 9 h at 37°C, the cells from two plates were resuspended in 10 ml of LB broth and used to inoculate 1 liter of LB broth (100 g/ml ampicillin). The culture was grown with vigorous shaking at 37°C for 2 h or until the cell density had reached an A 600 of 0.5. The culture flask was then transferred to a second shaker at 20°C and permitted to grow for a further 30 min or until their cell density had reached an A 600 of 1.0. At this point isopropyl-␤-thiogalactoside was added to a final concentration of 0.1 mM. After 7.5 h the cells were harvested by centrifugation at 4000 ϫ g for 30 min and used immediately for protein purification.
Unless otherwise stated all subsequent purification procedures were undertaken at 4°C. Cells were resuspended using 20 ml of 50 mM Tris-HCl, 100 mM (NH 4 ) 2 SO 4 , 2 mM dithiothreitol, pH 8.0, per liter of culture and lysed with 6 bursts of sonication for 40 s using a Branson sonicator fitted with a blunt tungsten tip. The temperature of the solution was monitored to ensure that it did not exceed 10°C. The lysed cells were then centrifuged at 11,200 ϫ g for 30 min, and the pellet was discarded. Polyethyleneimine was added to the supernatant to a final concentration of 0.01%, and the mixture was allowed to stir for 10 min and then centrifuged at 11,200 ϫ g for 20 min. The supernatant was loaded directly onto a Q-Sepharose column (22 cm 3 liter Ϫ1 of culture) in the above buffer. The initial fractions eluting from the Q-Sepharose column that contained protein were combined. These were brought to 35% (NH 4 ) 2 SO 4 saturation over a period of 20 min and centrifuged at 7,800 ϫ g for 20 min. The resulting supernatant was then brought to 45% (NH 4 ) 2 SO 4 saturation over a period of 20 min and centrifuged at 7,800 ϫ g for 20 min. The 45% supernatant was brought to 55% (NH 4 ) 2 SO 4 saturation over a period of 20 min and again centrifuged. The pellet obtained from this step was redissolved in 50 mM MES, 200 mM (NH 4 ) 2 SO 4 , 10% glycerol, 2 mM dithiothreitol, 100 M ferrous ammonium sulfate, pH 7.0, using 25 ml per liter of initial culture. The redissolved enzyme was applied to a ceramic hydroxyapatite column (11 cm 3 per liter of initial culture) at a flow rate of 1 ml/min. The column was washed with approximately 2 column volumes of 50 mM MES, 200 mM (NH 4 ) 2 SO 4 , 10% glycerol, 2 mM dithiothreitol, 100 M ferrous ammonium sulfate, pH 7.0, and the protein was eluted with a linear gradient (18 column volumes) from this buffer to 300 mM sodium phosphate, 10% glycerol, 2 mM dithiothreitol, 100 M ferrous ammonium sulfate, pH 6.5. TRH 102-416 typically eluted between 150 and 200 mM phosphate. Fractions containing TRH 102-416 were pooled. Contaminating nucleic acids were removed by the further addition of polyethyleneimine to 0.01% and stirring for 20 min. After centrifugation at 11,200 ϫ g for 20 min, the supernatant was concentrated by the addition of (NH 4 ) 2 SO 4 to 65% saturation over a period of 20 min. This sample was then centrifuged at 7,800 ϫ g for 15 min, and the supernatant was discarded. The white protein pellet was redissolved in a minimum of 50 mM MES, 200 mM (NH 4 ) 2 SO 4 , 10% glycerol, 2 mM dithiothreitol, pH 7.0, and stored as small aliquots (typically 200 l of 200 M enzyme) at Ϫ70°C. When the enzyme was to be used for kinetic studies, the storage buffer also contained 100 M ferrous ammonium sulfate. A typical yield from 1 liter of cells was 10 mg with a specific activity of 0.6 mol of hydroxytryptophan produced per min/mg.
Ultracentrifugation-Sedimentation equilibrium analyses were carried out at 10°C in a Beckman model XL-A ultracentrifuge. Concentrated TRH 102-416 was diluted to 5, 10, or 15 M in 200 mM (NH 4 ) 2 SO 4 , 100 mM MES, pH 7.2. The system was assessed as having attained equilibrium when scans at 280 nm taken at 2-h intervals were identical. The data used for analysis were averages of 20 successive scans. The absorbance values as a function of radial position were fit using Kaleidagraph software to either Equation 1, which describes the equilibrium sedimentation of a monomer, or Equation 2, which describes the sedimentation of a self-associating species. The terms used in Equations 1-3 are as follows: N, the stoichiometry of association; M, the molecular mass of an enzyme monomer; A o , the absorbance at the reference radius r o ; K a , the association constant; C, the base-line offset; v, the partial specific volume (calculated to be 0.723 from the amino acid content of TRH 102-416 ); , the buffer density; and , the angular velocity in radians/s. The concentration of protein was determined using an ⑀ 280 value of 35.2 mM Ϫ1 cm Ϫ1 calculated by the method of Pace et al. (22).
Enzyme Assays-An HPLC-based assay for enzyme activity was used for samples during purification. The 500-l reaction mixture contained enzyme, 0.2 mg/ml catalase, 200 M tryptophan, 200 M 6-MePH 4 , 100 M ferrous ammonium sulfate, 15 mM ␤-mercaptoethanol, 100 mM MES, pH 7.0. The reaction was initiated by the addition of tetrahydropterin after equilibrating all other substituents at 37°C for 1 min. Aliquots (100 l) were withdrawn at times between 0 and 30 s and quenched into 10 l of 40% trichloroacetic acid. These samples were then centrifuged at 12,000 ϫ g for 10 min and loaded onto a Rainin microsorb MV reverse phase C18 HPLC column (50 ϫ 4.6 mm inner diameter) using a 10-l injection loop. The mobile phase was 40 mM sodium acetate, 5% acetonitrile, pH 3.5, at flow rate of 1 ml/min. 5-Hydroxytryptophan and tryptophan were detected using a Waters 470 fluorescence detector with an excitation of 290 nm and an emission of 340 nm. Under these conditions 5-hydroxytryptophan and tryptophan had retention times of 1.7 and 3.9 min, respectively. When phenylalanine was used as a substrate, tyrosine was detected with an excitation of 275 nm and an emission of 310 nm, using 40 mM sodium acetate, pH 3.5, as the mobile phase. Similarly, the product of tyrosine hydroxylation, 3,4-dihydroxyphenylalanine, could be detected with an excitation of 280 nm and an emission of 310 nm in the same mobile phase.
Steady state kinetic measurements with tryptophan as substrate were made using an Applied Photophysics stopped-flow apparatus operating in the fluorescence mode. TRH 102-416 (typically 2-10 M) in air-saturated 200 mM (NH 4 ) 2 SO 4 , 100 mM MES, 100 M ferrous ammonium sulfate, 25 g/ml catalase, pH 7.0, was rapidly mixed with aerobic solutions containing varied concentrations of tetrahydropterin and tryptophan in 10 mM HCl and 12 mM dithiothreitol at 15°C. The excitation wavelength was 300 nm; all emitted light which passed a 320-nm wavelength cut-off filter perpendicular to the light source was collected. Under such conditions, the formation of 5-hydroxytryptophan could be monitored independently of the substrate or the formation of the fluorescent 7,8-dihydropterin. The amount of 5-hydroxytryptophan generated was quantified by comparing the fluorescence yield to that of 5-hydroxytryptophan standards measured on the same instrument in the presence of comparable quantities of other substrates. The concentration dependence data were fit to Equations 4 and 5. Equation 5 was used when substrate inhibition was observed; K ai is the inhibition constant for the substrate.
The metal requirement of TRH 102-416 was determined using sequential stopped-flow spectrophotometry. The apoenzyme ( 4 to 100 M, and spectra were recorded every 5 s for 400 s. The amount of hydroxylated amino acid produced was then determined by HPLC. The spectra generated during the reaction were analyzed globally using the program Specfit (Spectrum Software Associates) to determine the spectra of the pterin products.
Iron Determination-The iron content of the enzyme was determined by atomic absorption spectroscopy, using a slight variation of the method of Ramsey et al. (24). Samples were dialyzed into 100 mM MES, 200 mM (NH 4 ) 2 SO 4 , pH 7.0, and then diluted 5-fold in 2.5 M nitric acid. After 30 min on ice the samples were diluted 10-fold with water and centrifuged at 12,000 ϫ g for 10 min. Aliquots of the supernatant were analyzed for iron using a Perkin-Elmer model 2380 atomic absorption spectrophotometer equipped with a graphite furnace.

RESULTS
Expression-The T7 polymerase-based pET expression system of Studier (25) was used to express wild type TRH and two truncated mutant proteins. The wild type TRH was expressed as less than 1% of the total cell protein. This level of expression was unaffected by temperature over the range 20 -37°C (data not shown). At 20°C, the TRH activity reached a maximum level of approximately 4 nmol/min/ml culture 7 h after induction (Fig. 1A). All of the wild type TRH was soluble. In contrast, the protein lacking the amino-terminal regulatory domain, TRH 102-444 , was expressed at much higher levels. Based upon SDS-polyacrylamide gel electrophoresis of cell lysates after induction, TRH 102-444 was expressed as 30 -35% of the total cell protein under all conditions tested. However, when cells were grown at 37°C, 95% of the TRH 102-416 was found in inclusion bodies. The fraction of soluble TRH 102-444 could be increased by decreasing the growth temperature to 17°C, but the greatest activity found under any condition with this protein was only 40% that observed with the wild type protein (see Fig. 1A). Moreover, during attempts to purify TRH 102-444 , the enzyme showed a pronounced tendency to lose activity and precipitate.
Although the absolute levels of expression seen with TRH 102-444 were sufficient, the insolubility of the protein and the low activity suggested that it was not folding properly. TRH is believed to be held together as tetramer by hydrophobic helices at the carboxyl termini of each monomer (6,8,16,26,27), similar to TYH (13,18). Given the tetrameric nature of TRH, it was considered possible that the presence of even a single unfolded subunit within the tetramer would be sufficient to render the entire tetramer unstable. Since the interface helix is not required for activity in these hydroxylases (13,15,27), a mutant protein lacking both the regulatory domain and the carboxyl-terminal helix was examined. This form of TRH, TRH 102-416 , was expressed at a level similar to that observed with TRH 102-444 . However, based on SDS-polyacrylamide gel electrophoresis and enzyme activity, 80% of the protein was soluble, so that the active, soluble TRH 102-416 was 25-30% of the total cell protein. The amount of enzyme activity produced as TRH 102-416 was approximately 10 times that observed with the wild type enzyme and 25 times that obtained with TRH 102-444 (Fig. 1A). Consequently, TRH 102-416 was selected for further characterization.
Purification-TRH 102-416 could be purified in three steps, a Q-Sepharose column, ammonium sulfate fractionation, and a hydroxyapatite column (Fig. 1B). TRH 102-416 showed marked instability at low ionic strength. Maintaining the enzyme in a minimum of 100 mM (NH 4 ) 2 SO 4 during purification greatly enhanced recovery from each chromatographic step. The enzyme stability was further enhanced by ferrous iron and dithiothreitol. The enzyme could be stored at Ϫ70°C indefinitely without loss of activity. Because of the improved stability which resulted, the enzyme was typically stored in the presence of 100 M ferrous ammonium sulfate. If ferrous ammonium sulfate was omitted from purification buffers, the resulting enzyme contained no detectable iron (Ͻ0.05 atom/monomer).
Ultracentrifugation-Since TRH 102-416 no longer contains the helix that is proposed to be necessary for oligomerization, the mutant protein should be a monomer. Equilibrium ultracentrifugation was used to analyze its quaternary structure. Data were collected using initial protein concentrations of 5-15 M and rotor speeds of 16,000, 19,000, and 22,000 rpm. When the data were analyzed assuming that a single monomeric species was present, the average molecular weight over all conditions was 42,400 Ϯ 4381, compared with a molecular weight of 36,319 calculated from the DNA sequence. These results suggested that species larger than a monomer were present. To examine this possibility, the data were fit to models describing monomer-dimer, monomer-trimer, monomer-tetramer, monomer-hexamer, and monomer-octamer equilibria. In each case the molecular weight was fixed at 36,319. The data were best fit using a monomer-tetramer model (Table I). No improvements in the quality of the fits were seen if species larger than a tetramer were considered. The average K a value obtained by fitting each data set to a model describing a monomer-tetramer equilibrium was 1.33 ϫ 10 13 M Ϫ3 . Fig. 2 shows representative fits to this model of data obtained at three enzyme concentrations using the average K a value and a monomer molecular weight of 36,319. Based on this association constant, TRH 102-416 is 50% monomer at a protein concentration of 42 M.
Hydroxypterin Formation-With tyrosine hydroxylase and phenylalanine hydroxylase, the initial pterin product of catalysis is a 4a-hydroxypterin (28 -30). Although it is assumed that this is also a product with tryptophan hydroxylase, as shown in Scheme 1, formation of a hydroxypterin by tryptophan hydroxylase has not been demonstrated directly. Although hydroxypterins are not stable in solution for extended periods, they can be observed spectrally if the enzyme concentration is high enough to rapidly generate micromolar levels prior to hydrolysis. Consequently, high concentrations of TRH 102-416 (10 M) were used to consume 100 M tetrahydrobiopterin in the presence of excess tryptophan or phenylalanine. Near ultraviolet absorbance spectra of the reaction were collected every 5 s using a diode array spectrophotometer. The formation of the hydroxypterin was clearly detectable at 246 nm, where its absorbance is maximal, when either tryptophan or phenylalanine was the amino acid substrate (Fig. 3). After the spectral changes were complete, the amount of hydroxylated amino acid produced was determined by HPLC. With tryptophan as substrate 103 nmol of hydroxytryptophan were produced after complete oxidation of 100 nmol of tetrahydrobiopterin. Similarly, oxidation of 100 nmol of tetrahydrobiopterin in the presence of phenylalanine produced 99 nmol of tyrosine. Similar results were obtained with 6-MePH 4 (Table II). However when tyrosine was used as a substrate only 1.2 nmol of 3,4-dihydroxyphenylalanine were produced upon the oxidation of 100 nmol of tetrahydrobiopterin.
The sequential spectra collected during turnover were fit globally to the model in Scheme 2 to determine the spectra of the individual pterin species produced. The data were well fit by such a model (results not shown). The spectrum of the 4a-hydroxybiopterin determined by this method agreed well with previously published spectra (31), with an ⑀ 246 value of 18.1 mM Ϫ1 cm Ϫ1 (Fig. 3). The spectra of the subsequently formed pterin species agreed with previously described spectra of the quinonoid dihydrobiopterin and 7,8-dihydrobiopterin (31,32).  Steady State Kinetic Analyses for TRH 102-416 -Continuous assays were developed as a consequence of the substantial increase in activity of TRH 102-416 compared with TRH from other sources. The assay for hydroxylation of tryptophan followed the increased fluorescence of the product at wavelengths greater than 320 nm. Selectively exciting the reaction at 300 nm minimized the emission contributions from the substrate and fluorescent pterin products. With the alternate substrate phenylalanine, the accumulation of tyrosine could be observed as an absorbance increase at 275 nm.
Tryptophan, phenylalanine, and tyrosine were examined as substrates for TRH 102-416 . Both tryptophan and phenylalanine were hydroxylated efficiently, producing exclusively 5-hydroxytryptophan and 4-hydroxyphenylalanine, respectively. Under the experimental conditions, no production from phenylalanine of 3-hydroxyphenylalanine or 3,4-dihydroxyphenylalanine could be detected by HPLC. Very small quantities of 3,4-dihydroxyphenylalanine formed from tyrosine could be detected. The hydroxylation rate of tyrosine by TRH 102-416 under the conditions of the experiment was at least 5000-fold slower than that observed with tryptophan and phenylalanine. For this reason steady state kinetic parameters were determined only with tryptophan and phenylalanine using 6-MePH 4 and BH 4 . Significant substrate inhibition was seen when either amino acid was varied at a fixed concentration of either BH 4 or 6-MePH 4 (Fig. 4). Substrate inhibition was also seen when either tetrahydropterin was varied with tryptophan, but little or no inhibition was seen with either tetrahydropterin when phenylalanine was the fixed substrate (Fig. 4). Table III summarizes the kinetic parameters.
The Metal Requirement of TRH 102-416 -Both tyrosine hydroxylase and phenylalanine hydroxylase have been shown to be iron-containing enzymes (33)(34)(35), so that tryptophan hydroxylase is also assumed to require iron for activity. If no iron was included in the buffer when purifying TRH 102-416 , the resulting protein contained no significant iron. Apoenzyme prepared in this fashion was used to determine the metal requirement of tryptophan hydroxylase. The assays were done at high (Ͼ2 M) enzyme concentrations to avoid ambiguities due to the presence of contaminating metals. The apoenzyme was mixed with tryptophan and 6-MePH 4 and then 1 s later with individual metals at a final concentration of 50 M. The formation of hydroxytryptophan was followed by fluorescence. Representative kinetic traces are shown in Fig. 5. No activity was observed in the absence of added metal. In the presence of 50 M ferrous iron there was a constant rate of product formation beginning almost immediately after mixing. The same rate was seen with 10 M ferrous iron (results not shown). In contrast, in the presence of an equal concentration of ferric iron, a significant lag was seen before the activity reached the same level as was SCHEME 2.  seen with ferrous iron. The enzyme was not active with copper or cobalt.

DISCUSSION
The three tetrahydropterin-dependent hydroxylases, TYH, PAH, and TRH, constitute a small family of proteins that catalyze the hydroxylation of aromatic amino acids. Because of the availability of PAH from liver, studies of this enzyme have the most extensive history (36). More recently, the availability of recombinant TYH has resulted in a much greater understanding of that enzyme (10). In contrast, understanding the structure and mechanism of tryptophan hydroxylase has made little progress. Natural sources such as brain have proved to contain too little enzyme for useful purification, resulting in very low amounts of protein with low specific activities (1, 2, 4). A number of groups have described preparations of TRH from recombinant sources, but these typically involved enzyme of low or indeterminate specific activity (16,37,38). The limited amounts of material available have severely restricted analyses of the mechanism of this important enzyme.
Sequence comparisons of all three hydroxylases routinely show that the carboxyl-terminal 340 amino acids form a homologous region, whereas the amino-terminal sequences diverge widely (20). Deletion mutageneses of all three enzymes have defined the catalytic core which contains the residues required for catalysis. PAH lacking the amino-terminal 141 amino acid residues and the carboxyl-terminal 43 residues is reported to retain activity (15). Similar results have been reported for TYH (13,14), whereas TRH is reported to retain some activity if as many as 106 residues are deleted from the amino terminus and as many as 19 residues are deleted from the carboxyl terminus (9,39). Based upon such results, the amino-terminal portions of these proteins are generally accepted to be regulatory domains, in that they contain phosphorylation sites and are required for allosteric properties (37, 40 -42). The 42 carboxyl-terminal residues of each monomer are responsible for tetramer formation, primarily due to the presence of a 24-residue helix at the end (13,18). The remaining 300 residues form the catalytic domains of each hydroxylase. Removal of the regulatory domains from either TYH or PAH has only subtle effects on the substrate specificities or catalytic rates (17). In our hands, expression of the wild type rabbit TRH resulted in low levels of expression, consistent with the observations of others (38). Although removal of the regulatory domain resulted in a significant increase in the level of expression of TRH 102-444 , this form of the protein was only slightly soluble. It was only upon removal of the long helix in the tetramer interface that high levels of soluble active enzyme were obtained. A possible reason for this increase in solubility of TRH 102-416 , the protein lacking both the regulatory domain and the tetramerization helix, is that tetramers of TRH 102-444 contain mixtures of correctly folded and incorrectly folded monomers. Any improperly folded subunit may render an entire tetramer unstable. Even if not all of the monomeric TRH 102-416 is correctly folded, any improperly folded subunits would not be expected to affect the stability of other monomers. Irrespective of whether this is the correct reason for its increased solubility, TRH 102-416 is expressed at sufficiently high levels for mechanistic studies.
Wild type TRH is a tetramer (1,4). In contrast, TRH 102-416 is monomeric at and above concentrations typically encountered in kinetic experiments and only forms oligomers at a relatively high concentration. 2 This is the result expected upon removal of the intersubunit helix. There are clearly still some interactions among the monomers in TRH 102-416 , despite the lack of this helix. The structure of TYH shows that the tetramerization domain contains the carboxyl 42 residues, which include the intersubunit helix (18). In addition, there are other interactions across dimer interfaces. Similar interactions in TRH are presumably the reason for the weak formation of tetramers.
TRH 102-416 clearly requires ferrous iron for activity, as do both TYH and PAH (11,12,43). Although TRH 102-416 is active with ferric iron, there is a significant lag in formation of hydroxytryptophan in the presence of this metal. Both TYH and PAH are routinely found to have the active site iron in the ferric form when purified (44,45). The iron must be reduced to the ferrous form for catalysis; tetrahydrobiopterin appears to be the physiological reductant (11,24,43). The lag seen in the formation of hydroxytryptophan by TRH 102-416 in the presence of ferric iron is consistent with a similar phenomenon occurring with this protein. The lag would be due to the relatively slow reduction of the ferric enzyme by the tetrahydropterin.
Based upon the precedents with TYH and PAH (28 -30), it was expected that the initial pterin product with TRH would be the 4-hydroxypterin. The high levels of TRH 102-416 have made it possible to demonstrate this directly for the first time, establishing the reaction shown in Scheme 1 for TRH. Moreover, the stoichiometry of one tetrahydropterin consumed per hydroxylated amino acid produced shown in Scheme 1 has been established for both the physiological substrate BH 4 and the synthetic substrate 6-MePH 4 with both tryptophan and phenylalanine as the amino acid substrate. Thus, based upon the degree of coupling of tetrahydropterin consumption to tyrosine formation, phenylalanine is as good a substrate as tryptophan for TRH 102-416 . With both TYH and PAH it has commonly been observed that use of nonphysiological substrates results in an excess of tetrahydropterin consumed over amino acid hydroxylated (36,46,47). Indeed this is also the case when tyrosine is the substrate for TRH 102-416 where there is an 80-fold greater tetrahydropterin oxidation than amino acid hydroxylation.
The steady state kinetic analyses presented here for TRH 102-416 provide insight into the substrate specificity of TRH and allow comparison with PAH and TYH. Qualitatively, tyrosine is a poor substrate for TRH 102-416 , and both tryptophan and phenylalanine are good substrates. Indeed, given that the kinetic parameters in Table III could not be determined at saturating concentrations of the nonvaried substrates due to substrate inhibition, the kinetic parameters for tryptophan and phenylalanine are probably not significantly different. These results can be compared with the substrate specificities of TYH and PAH. It is best to use results obtained with the comparable catalytic domains of the latter enzymes because of the compli- 2 It is possible that the oligomeric state of the enzyme in kinetic experiments is influenced by the presence of substrates. This possibility has not been investigated.  cations caused by the need for prior activation of wild type PAH. Other than release from cooperative substrate activation of PAH, removal of the regulatory domains of PAH and TYH does not affect the ability of these enzymes to hydroxylate the other aromatic amino acids (17). TYH will hydroxylate phenylalanine (48,49). Whereas the rate of tetrahydropterin oxidation by TYH in the presence of phenylalanine is comparable to that seen in the presence of tyrosine, only a fraction of the reducing equivalents are used to form tyrosine (17,49). This is in contrast to the situation with TRH 102-416 , in which tetrahydropterin oxidation and amino acid hydroxylation are stoichiometric. Tryptophan is also a substrate for TYH, but with a V max value only 20% that seen with tyrosine and with a K m value 20-fold higher (50). Thus, tyrosine is clearly the preferred substrate for TYH. PAH is unable to hydroxylate tyrosine, although tyrosine does stimulate a low rate of tetrahydropterin oxidation (17,49,51). In contrast, PAH is able to hydroxylate tryptophan (42,52) but with a K m value a 1000-fold higher and a V max value one-tenth that of phenylalanine. Thus, PAH strongly prefers phenylalanine over the other two amino acids as a substrate. While TRH 102-416 resembles PAH qualitatively in its ability to hydroxylate only phenylalanine and tryptophan, TRH 102-416 shows no discernible preference for its physiological substrate. It is not clear whether there is physiological relevance to this lack of specificity.
In conclusion, the goal of the work presented here was to obtain a form of TRH that would permit study of catalytic properties. The results indicate that rabbit TRH 102-416 is a stable, highly active form of TRH that can be expressed to very high levels in E. coli. This mutant enzyme clearly appears to be valid for mechanistic studies of wild type TRH.