A typical N-terminal Extensions Confer Novel Regulatory Properties on GTP Cyclohydrolase Isoforms in Drosophila melanogaster*

The cofactor tetrahydrobiopterin plays critical roles in the modulation of the signaling molecules dopamine, serotonin, and nitric oxide. Deficits in cofactor synthesis have been associated with several human hereditary diseases. Responsibility for the regulation of cofactor pools resides with the first enzyme in its biosynthetic pathway, GTP cyclohydrolase I. Because organisms must be able to rapidly respond to environmental and developmental cues to adjust output of these signaling molecules, complex regulatory mechanisms are vital for signal modulation. Mammalian GTP cyclohydrolase is subject to end-product inhibition via an associated regulatory protein and to positive regulation via phosphorylation, although target residues are unknown. GTP cyclohydrolase is composed of a highly conserved homodecameric catalytic core and non-conserved N-terminal domains proposed to be regulatory sites. We demonstrate for the first time in any organism that the N-terminal arms of the protein serve regulatory functions. We identify two different modes of regulation of the enzyme mediated through the N-terminal domains. The first is end-product feedback inhibition, catalytically similar to that of the mammalian enzyme, except that feedback inhibition by the cofactor requires sequences in the N-terminal arms rather than a separate regulatory protein. The second is a novel inhibitory interaction between the N-terminal arms and the active sites, which can be alleviated through the phosphorylation of serine residues within the N termini. Both mechanisms allow for acute and highly responsive regulation of cofactor production as required by downstream signaling pathways.

iological conditions and disease states (24,25). The mechanism by which DAHP is able to inhibit GTPCH is similar to that of BH 4 in that GFRP is required for the inhibition and the inhibition is non-competitive (26,27). DAHP, accordingly, has been used to induce BH 4 deficiencies in animal models of BH 4 -deficient neurological disease (28) as well as hyperphenylalaninemia (29).
A second means of negatively regulating GTPCH activity appears to involve the expression of alternative isoforms. In humans, there are at least six alternatively spliced GTPCH mRNAs, with the translated proteins differing only in their C termini, yet only GTPCH type I is enzymatically active (30). Co-expression of GTPCH type I and GTPCH type II, a truncated and non-functional GTPCH protein, in human blood cells depresses the level of GTPCH type I protein (31). It remains unclear whether this mechanism is employed more universally for negative regulation of GTPCH in mammalian cells.
GTPCH is also positively regulated via post-translational modification. It has been reported that stimulation of GTPCH activity in mammalian cell culture occurs through phosphorylation and that GTPCH serves as a substrate for both casein kinase II and protein kinase C (32,33). However, specific sites of phosphorylation have not yet been identified in mammalian GTPCH, nor has the effect of phosphorylation been defined enzymatically.
The sequence of GTPCH is highly conserved; the Drosophila melanogaster homolog is nearly 80% similar to the mammalian form. Moreover, its reaction mechanism appears to be conserved, as does the decameric structure (34). This high degree of conservation in a genetic model organism affords the opportunity to investigate, in greater detail, the mechanisms of regulation and their effects on the organism. Drosophila GTPCH is encoded by the gene Punch (Pu), which has been subjected to extensive genetic and molecular analysis (35)(36)(37). The Pu locus produces at least four transcripts by alternative promoters and alternative splicing. Three of these, transcripts A (1.70 kb), B (1.75 kb), and C (1.80 kb), are well characterized (38). Most of the predicted sequence for each of the resulting polypeptides is virtually identical to the catalytic core protein in mammals. The GTPCH polypeptides from all species analyzed to date have a non-conserved N-terminal domain that extends as an "arm" from the catalytic core. These arms have been variously proposed to be regulatory domains or docking sites for interacting proteins, although evidence is lacking to confirm these hypotheses. The Drosophila isoforms similarly have non-conserved N-terminal domains (supplemental Fig. S1). The isoforms, however, differ from the mammalian forms in three aspects. First, each has a unique N-terminal domain rather than differing at the C terminus as is the case for the mammalian protein.
Second, the N termini are significantly longer than any other characterized GTPCH. Third, unlike the mammalian forms, which seem to be expressed in the same cells, the Drosophila isoforms are expressed in different tissues and at different times during development (36,38,39).
We have undertaken an in vitro analysis of recombinant Drosophila GTPCH isoforms to test for the first time the hypothesis that the N-terminal domains have regulatory functions. The studies reported here demonstrate that each of the Drosophila isoforms is catalytically active, and they provide evidence for a key regulatory region in the N termini of isoforms B and C. We also show that the N-terminal domains of these two isoforms serve as substrates for protein kinase A (PKA) and protein kinase C (PKC). Moreover, we obtained two surprising results. First, the distal residues in these arms strongly affect the kinetic properties of the enzymes, suggesting a heretofore unknown regulatory interaction between the arms and the active sites. Second, the N-terminal extensions serve as functional homologs of the mammalian GFRP despite the non-alignment of sequence, in that they are capable of directing non-competitive inhibition by BH 4 and DAHP in the absence of GFRP.

EXPERIMENTAL PROCEDURES
Cloning of Drosophila GTPCH Isoforms-Cloning of cDNAs corresponding to Drosophila GTPCH isoforms A and B was described previously (38). Isoform C cDNA was cloned from a 0-to 24-h Drosophila embryonic library (Stratagene, La Jolla, CA) containing cDNA clones in the bacteriophage ZAP using methods essentially the same as described previously (38).
Expression and Purification of Recombinant Proteins-PCR products were cloned into the pQE30 vector (Qiagen) for prokaryotic expression of His 6 -tagged GTPCH by standard meth-ods. Plasmid DNA was purified using the QIAprep Spin Miniprep Kit (Qiagen), and sequence confirmation analysis was performed by the Auburn University Genomics and Sequencing Laboratory. The verified DNA was used to transform competent M15 cells harboring the kanamycin-resistant pREP4 plasmid, which provides an inhibitor to expression. Transformed cells were grown on LB plates containing ampicillin (100 mg/liter) and kanamycin (25 mg/liter). A single colony was picked and grown in 1 liter of LB media containing ampicillin (100 mg/liter) and kanamycin (25 mg/liter). When bacterial density reached 0.6 at A 600 , 1 mM isopropyl-1-thio-␤ -D-galactopyranoside (LabScientific, Inc.) was added to induce recombinant protein expression. After 5 h at 37°C, the bacterial culture medium was centrifuged for 20 min at 15,000 ϫ g, and the resulting pellet was stored at Ϫ80°C until use. The cell pellet was thawed for 20 min on ice and resuspended in 20 ml of lysis buffer (50 mM NaH 2 PO 4 , 300 mM NaCl, 10 mM imidazole, pH 8.0). Lysozyme (1 mg/ml) and one tablet of Complete protease inhibitor mixture (Roche Applied Science) were added, and the mixture was stirred on ice for 30 min. Bacteria were lysed by sonication, and the lysate was cleared by centrifugation at 15,000 ϫ g for 20 min. The clarified supernatant was combined with 1 ml of 50% slurry of nickel-nitrilotriacetic acid resin (Qiagen) and mixed gently for 60 min at 4°C. The mixture was applied to a Poly-Prep chromatography column (Bio-Rad). After washing with 20 ml of washing buffer (50 mM Tris, 150 mM NaCl, 0.1% Tween 20, 10 mM imidazole, pH 8.0), the protein was eluted with 500 l of elution buffer (50 mM Tris, 150 mM NaCl, 0.1% Tween 20, 200 mM imidazole, pH 8.0). The eluate was passed through the column two additional times to increase yield. Protein preparations were analyzed for purity by SDS-PAGE and Coomassie Blue protein staining. All kinetic assays were performed on purified recombinant enzymes.
Protein Determination-Protein concentrations were measured by the Bradford protein microassay (40) using the Quick Start Bradford Protein Assay (Bio-Rad) with bovine serum albumin used as the standard.
Protein Sequence Analysis-Sequence alignments were performed using the MUSCLE alignment program with JalView Java Alignment Editor (41,42). Alignment colors pertain to the BLOSUM62 score matrix (43). 32 P-Phosphorylation Labeling-Ten micrograms of protein was suspended in 50 mM Tris, 10 mM MgCl 2 , 200 M ATP, 4 Ci of [␥-32 P]ATP (Amersham Biosciences), with or without 1.25 kilounit of recombinant mouse catalytic subunit PKA (pH 7.5, Calbiochem) or 10 ng of recombinant rat brain catalytic subunit PKC (pH 8.0, Calbiochem) and incubated at 30°C for 30 min. GTPCH phosphorylation reactions were stopped by adding protein loading buffer and boiling for 10 min. Proteins were resolved by 15% SDS-PAGE. Gels were dried in a slab gel dryer (SGD 4050, Savant). X-ray film was placed on the dried gel, exposed, and developed following standard methods.
Kinetic Microplate Assay for GTPCH Activity-Formation of H 2 NTP from GTP (Roche Applied Science) was monitored using a kinetic microplate assay for GTPCH activity (27,44). An increase in A 340 at 15-s intervals over a 1-h period at 37°C was used to measure the accumulation of H 2 NTP using a Thermomax microplate spectrophotometer (Molecular Devices). All reaction mixtures were 200 l (final volume) and contained 100 mM Na ϩ /K ϩ phosphate buffer (93.5 mM Na 2 HPO 4 , 6.5 mM KH 2 PO 4 , pH 7.8) or 50 mM Tris-HCl buffer (50 mM Tris, 300 mM NaCl, 0.1% Tween 20, 10% glycerol, pH 7.5), 0.05 M GTPCH, and various concentrations of GTP and DAHP. A standard curve was generated by complete enzymatic conversion of desired concentrations of GTP to H 2 NTP by 0.2 M GTPCH isoform C as described (27,44).
HPLC-based Assay for GTPCH Activity-GTPCH activity was assayed by a modification of a method described previously (45). All reaction mixtures were 70 l (final volume) and contained 0.2 M purified GTPCH, 0.25 mM GTP (Roche Applied Science), and various concentrations of BH 4 , in either 100 mM Na ϩ /K ϩ phosphate buffer or 50 mM Tris buffer. Reactions were incubated in the dark at 37°C for 60 min. The reaction was stopped, and the product, H 2 NTP, was oxidized to neopterin triphosphate by adding 30 l of 1% I 2 /2% KI solution in 1 M HCl and incubating at 37°C for 60 min in the dark. Samples were decolorized with 15 l of 3% ascorbic acid and centrifuged for 5 min at 10,000 ϫ g. The samples were neutralized with 25 l of 1 M NaOH and dephosphorylated at 37°C for 30 min in a 70-l mixture consisting of 50 l of the oxidized neopterin triphosphate mixture, 2 units of calf intestinal alkaline phosphatase (Roche Applied Science), 7 l of 10ϫ dephosphorylation buffer (Roche Applied Science), 3 l of 1 M NaOH, and 8 l of H 2 O. Neopterin was then quantified by reverse-phase HPLC following centrifugation through micro-spin centrifuge filter tubes (Alltech) with 0.2-m pore size. Chromatographic separations were performed on an ESA CoulArray (Model 5600A) HPLC instrument. The mobile phase contained 75 mM sodium phosphate adjusted to pH 3.0 with phosphoric acid, 0.75 mM octanesulfonic acid, 25 M EDTA, 100 l/liter triethylamine, and 7% acetonitrile. Separations were performed on a Phenomenex Synergi 4-m Hydro-RP (4.6 ϫ 15 cm) column, preconditioned with 500 ml of buffer before use and run with an isocratic flow at 0.5 ml/min. Neopterin was detected by fluorescence at excitation 360 nm/emission 465 nm with a Linear Model LC305 fluorescence detector. Neopterin content was determined by comparison with commercial neopterin (Sigma) as standard using ESA CoulArray software.
PKA Treatment and GTPCH Activity Analysis-Recombinant GTPCH isoforms B and C were expressed and lysed as above. A 45% ammonium sulfate precipitation was performed by dropwise addition of saturated ammonium sulfate solution to the cleared cellular lysate. Following incubation on ice for 1 h, the mixture was centrifuged 10,000 ϫ g for 15 min. The protein pellet, containing GTPCH, was resuspended in 4.5 ml of kinase buffer (50 mM Tris-HCl, 10 mM MgCl 2 , 200 M ATP, pH 7.5). Treatment or control samples of the resulting mixture (2 ml) were treated with either 5 kilounits of recombinant mouse catalytic subunit PKA (Calbiochem) or equal volume kinase buffer, respectively. The samples were incubated at 30°C for 30 min. A 50% slurry of nickel-nitrilotriacetic acid resin (500 l, Qiagen) was added to each sample, and purification of GTPCH was performed as above, except that the protein was eluted with 200 l of buffer. Proteins (0.05 M) were assayed in 100 mM Na ϩ /K ϩ phosphate buffer for enzymatic activity using the kinetic microplate assay.
DAHP Administration-Newly eclosed, male w 1118 flies were collected and aged 72 h on standard media and then maintained on 5% sucrose, 5% Me 2 SO, with or without 10 or 50 mM DAHP for 72 h. Two DAHP-treatment groups were subsequently maintained on standard media for either 72 or 144 h post treatment. For each treatment, 45-85 flies were frozen and homogenized in 70 l of 0.1 M perchloric acid. The resulting homogenate was centrifuged at 12,000 ϫ g for 15 min at 4°C, and the supernatant was filtered through 0.2-m filters. HPLC analysis was conducted as for the GTPCH activity assay above, with the exception that 1 ml/min isocratic flow was used for the separations, and BH 4 was detected electrochemically using an analytical cell ESA Model 5011. Commercial BH 4 (Sigma) was used as a standard.
Kinetic and Statistical Analyses-SOFTmax Pro (Molecular Devices) and GraphPad Prism computer software were used for determination of kinetic properties and statistical analyses.

RESULTS
Kinetic Analysis of Recombinant Drosophila GTPCH Isoforms-To investigate the kinetic properties of Drosophila GTPCH isoforms A-C, recombinant proteins were purified from an Escherichia coli expression system (supplemental Fig.  S2). Purified isoforms A-C consistently exhibit higher molecular masses on SDS-PAGE (39, 45, and 47 kDa, respectively) than the calculated molecular masses (30,34, and 36 kDa, respectively) as noted previously for Drosophila (38), rat (46), and human (47,48) GTPCH subunits on SDS-PAGE. This discrepancy appears to be due, in part, to the N-terminal extensions of isoforms A-C, which exhibit apparent higher molecular masses (16,17, and 18 kDa, respectively) than expected (8, 12, and 13 kDa, respectively), whereas the mobility of the common region is consistent with its expected molecular mass at 24 kDa.
The recombinant isoforms were assayed for enzymatic activity using a kinetic microplate assay (27,44). Unlike the mammalian splice variants (30), all three Drosophila alternative isoforms exhibit enzymatic activity ( Table 1). The reported K 0.5 and V max values for the three isoforms are consistent with previously published values for purified Drosophila (34) and mammalian GTPCH enzymes (26,27,49,50). Inhibition by GTP substrate was also noted for all three isoforms at substrate concentrations exceeding 0.5 mM GTP (data not shown), which parallels previous observations with purified Drosophila GTPCH (34). Interestingly, a marked difference was noted in V max between isoform A and isoforms B and C in both Tris and phosphate buffers, with isoform A exhibiting a V max approximately twice that of isoforms B and C. Isoforms B and C, which differ only in a 16-amino acid addition to the N-terminal domain of isoform C, exhibit similar activity. The differences noted between isoforms A and isoforms B and C can be attributed to the N-terminal domain of isoform A, as it differs completely in sequence from those of isoforms B and C.
In accordance with previous observations of Drosophila GTPCH purified from crude head extracts, all three recombinant enzymes show positive cooperativity with Hill numbers exceeding 1.0 in Na ϩ /K ϩ phosphate buffer. In Tris buffer, marked positive cooperativity was noted for isoforms A and B, but not for isoform C. This observation is consistent with a buffer difference in positive cooperativity reported previously for GTPCH purified from adult Drosophila (34).
Drosophila GTPCH Isoforms Are Subject to Feedback Inhibition by BH 4 -Mammalian GTPCH is subject to feedback inhibition by BH 4 only in the presence of GFRP (19). To determine whether recombinant Drosophila GTPCH was similarly unresponsive to BH 4 , each purified isoform was assayed for activity in the presence of various concentrations of BH 4 by HPLC analysis (Fig. 1). Surprisingly, all three isoforms show inhibition by BH 4 (supplemental Fig. S3). In Tris buffer, enzymatic activity was nearly abolished by 1 M BH 4 . This BH 4 concentration had little effect on GTPCH in phosphate buffer. However, even in phosphate buffer, BH 4 strongly inhibited GTPCH activity at concentrations of 20 M or higher. This inhibition was unexpected because GFRP is not present in the Drosophila genome;

Isoform
Tris Phosphate

GFRP-independent Inhibition of Drosophila GTPCH Isoforms by DAHP Is Non-competitive and Requires N-terminal
Extensions-To further characterize this surprising inhibition of Drosophila GTPCH, we employed the prototypical GTPCH inhibitor, DAHP, in activity assays over a range of GTP concentrations (Fig. 2, A-C). All three isoforms exhibit non-competitive inhibition in the presence of DAHP with decreasing V max and unchanging K 0.5 as inhibitor concentration increased. As with BH 4 , the inhibition of the Drosophila enzymes by DAHP does not require GFRP. Nevertheless, this non-competitive inhibition is remarkably similar in nature to that observed with mammalian GTPCH in the presence of BH 4 or DAHP and GFRP (19,27).
Because the common region of Drosophila GTPCH is nearly identical to the mammalian protein, we hypothesized that the unique N-terminal extensions must facilitate the non-competitive inhibition, functionally replacing the mammalian regulatory protein. To test this hypothesis, we generated N-terminally truncated Drosophila GTPCH isoforms. We tested a truncated isoform C, lacking the first 58 amino acid residues (Ϫ58 isoform C truncation), as well as a protein lacking any N-terminal extension (common region, Ϫ116 amino acid residues for isoform C). The Ϫ58 isoform C truncation displayed non-competitive inhibition in the presence of DAHP similar to that of the full-length enzymes (Fig. 2D). In contrast, removal of the entire N-terminal extension resulted in the inability of DAHP to inhibit the enzyme (Fig. 2E). These results suggest that the unique N-terminal extensions are indeed responsible for the non-competitive inhibition. Moreover, for isoform C, amino acid residues 59 -116 appear to confer this regulatory property.
DAHP Inhibits Drosophila GTPCH in Vivo-In mammals, DAHP is capable of inhibiting GTPCH in vivo (24,25). In mammalian cells, GFRP is required to facilitate this inhibition (26). Recently, inhibition of swallowtail butterfly, Papilio xuthus, GTPCH by DAHP has also been demonstrated (51). To eliminate the possibility that the DAHP inhibition observed is an in vitro artifact, we tested whether DAHP is capable of inhibiting Drosophila GTPCH in vivo. To this end, 72 h post-eclosion, male Drosophila were fed either DAHP (10 or 50 mM) in 5% sucrose/5% Me 2 SO or 5% sucrose/5% Me 2 SO alone (control) for 72 h. Whole fly extracts were subsequently assayed for BH 4 by HPLC (Fig. 3A). BH 4 levels decreased with DAHP ingestion, consistent with GTPCH inhibition in vivo. We noted a potential threshold inhibition level for 72-h DAHP administration, with low variation between 10 and 50 mM DAHP treatments. Following the removal of Drosophila from DAHP-containing media, BH 4 pools returned to wild-type levels (Fig. 3B), suggesting that, in Drosophila, DAHP acts as a classic enzymatic inhibitor rather than eliciting permanent alteration in GTPCH activity. Thus, DAHP inhibition of GTPCH occurs in vivo in Drosophila as in mammals, despite the absence of a Drosophila GFRP homolog.
Truncation Analysis Reveals Native Negative Regulatory Region of Drosophila GTPCH-From the kinetic analysis of the three recombinant Drosophila GTPCH isoforms, it is evident that the V max of isoform A in phosphate buffer (60.89 Ϯ 2.82 nmol of H 2 NTP/ min/mg of protein) is approximately twice that of isoforms B and C (29.07 Ϯ 1.91 and 27.28 Ϯ 0.86 nmol of H 2 NTP/min/mg of protein, respectively). The only sequence differences among these three isoforms reside within the N-terminal extensions. The extension of isoform A is completely unique in sequence compared with the N-terminal domains of isoforms B and C. We hypothesized that specific residues or sub-domains within the N-terminal extensions of B and C act as negative regulators of GTPCH activity. To explore this possibility, various truncations of the N-terminal extension of recombinant isoform C were assayed for activity and compared with recombinant full-length isoforms and common region (Fig. 4). These experiments resulted in distinct and important observations. First, the common region alone (207 amino acid residues) exhibited significantly lower activity compared with the full-length isoforms; residues within the N-terminal extensions are therefore required for full enzymatic activity. Conversely, a truncation removing the initial 30 amino acid residues of isoform C had no significant effect on activity compared with full-length isoform C. This result suggests that amino acids 31-117 of the N-terminal extension of isoform C are required for full activity, whereas the first 30 amino acids are expendable in relation to activity.
Surprisingly, removal of the first 58 amino acid residues of isoform C resulted in a significant elevation in activity compared with full-length isoform C. In fact, the activity for the Ϫ58 isoform C truncation parallels that of full-length isoform A. This result suggests that a region of the N-terminal extension of isoform C, from amino acid residues 31 to 58, serves to negatively regulate the Drosophila enzyme and is thus responsible for the lowered enzymatic activity of isoform C compared with isoform A. This negative regulatory region is shared between isoforms B and C, suggesting that this regulatory property is common to both isoforms.
Potential Phosphorylation Sites Reside within the Candidate Negative Regulatory Domain of Drosophila GTPCH Isoforms B and C-It has been suggested previously that mammalian GTPCH is regulated via phosphorylation and that phosphoryl-  The initial 30 and 58 amino acids of the N-terminal extension of GTPCH isoform C were removed, and the remaining recombinant protein was tested for enzymatic activity using the kinetic microplate assay. Recombinant full-length isoforms A and C and common region were also assayed. For all assays, 0.05 M protein was used in Na ϩ /K ϩ phosphate buffer. Removal of the entire N-terminal extension (CR) results in significantly decreased activity. Excision of the first 30 amino acids (Ϫ30 C) results in no significant decrease from full-length isoform C. The removal of the first 58 amino acids of isoform C (Ϫ58 C) shows a 2-fold increase in activity over fulllength C. The activity of the Ϫ58 C truncation parallels that of full-length isoform A. Values are means Ϯ S.E. from two independent experiments, each with triplicate determinations. ***, p Ͻ 0.001 (one-way ANOVA with Bonferroni post test; failure to reach significance denoted as NS).
ation of the enzyme results in increased enzymatic activity (32,33). The candidate negative regulatory domain in isoform C defined by the above truncation study, amino acid residues 31-58 (NЈ-SRGRNNSVCSTSSTSGTSSLADRQQNQA-CЈ), is serine/threonine-rich and therefore likely to contain substrates for Ser/Thr kinases. We conducted a phosphorylation prediction analysis on the putative negative regulatory domain using NetPhos 2.0 (52) and found that seven serine residues within this domain are predicted phosphorylation targets (supplemental Table S1). Of these seven serine residues, Ser 37 had the highest phosphorylation prediction value (98.7%) and is predicted to be phosphorylated by both PKA and PKC.
To establish whether Drosophila GTPCH can be phosphorylated by these kinases, we utilized [ 32 P]ATP with PKA (Fig. 5) and PKC (supplemental Fig. S4). Full-length isoforms B and C are phosphorylated (lanes 6 and 7) by both PKA and PKC, whereas the common region alone (lane 4) is not phosphorylated by either kinase. Full-length isoform A is not phosphorylated by PKA, and only a faint band is observed with PKC (lane 5). The N-terminal extensions of isoforms B and C, but not of isoform A, are phosphorylated (lanes 1-3). These results suggest that full-length isoforms B and C serve as substrates for both PKA and PKC and that phosphorylation by these kinases occurs within the N-terminal extensions of these isoforms.
To determine the functional significance of this post-translational modification of GTPCH isoforms B and C, recombinant isoforms were treated with PKA and subsequently assayed for enzymatic activity (Fig. 6). Both isoforms exhibit an elevation in activity following PKA treatment, indicating that Drosophila GTPCH isoforms B and C are activated by phosphorylation as suggested by earlier studies in mammalian cell culture (32,33).
To explore the possibility that the serine-rich, candidate negative regulatory domain is regulated by phosphorylation, we modified Ser 37 to glutamic acid to mimic phosphorylation at this site. The enzymatic activity of S37E is significantly higher than wild-type GTPCH isoform C (Fig. 7). This suggests that phosphorylation of residues within the candidate negative regulatory domain may serve to alleviate the negative regulation imposed on the enzyme by the native domain.
We hypothesized that substituting alanine for Ser 37 would have no significant effect on enzymatic activity, because removal of the first 58 amino acids, including Ser 37 , results in higher activity as demonstrated by truncation analysis. Contrary to our expectations, the S37A substitution greatly reduced the activity of wild-type GTPCH isoform C (Fig. 7), suggesting that the candidate regulatory domain interacts with and plays a pivotal role at the active site of the enzyme. This regulatory interaction has not been reported previously for either GTPCH alone or GTPCH and GFRP.

DISCUSSION
Although it has been hypothesized that the N-terminal domain of GTPCH plays regulatory roles, we present here the first direct evidence of this phenomenon. Drosophila GTPCH is particularly amenable to this work, because the various alternative isoforms differ only in their N-terminal sequences, giving each isoform potentially distinct regulatory properties. The .5) or in kinase buffer alone, following ammonium sulfate precipitation from cleared cell lysate as described under "Experimental Procedures." Following 30-min incubation at 30°C, 500 l of 50% slurry of nickel-nitrilotriacetic acid resin was added to each sample, and recombinant GTPCH was purified. Proteins (0.05 M) were assayed in Na ϩ /K ϩ phosphate buffer for enzymatic activity using the kinetic microplate assay. Values are percent relative activity with 100% being activity of control samples. Values are means Ϯ S.E. from nine determinations. *, p Ͻ 0.05; ***, p Ͻ 0.001 (two-way Student's t test). recombinant Drosophila isoforms also exhibit kinetics comparable to the mammalian enzyme and parallel earlier observations of GTPCH purified from Drosophila.
We demonstrate that the three isoforms of Drosophila GTPCH have enzymatic activity unlike the mammalian alternative isoforms, which differ in their C-terminal sequences. In addition, the variation in the N-terminal sequences confers differential enzymatic activities and regulatory properties. The difference in activity and regulation is most evident between isoform A and isoforms B and C, which follows the disparity in the N-terminal sequences between these isoforms.
The N-terminal extensions of Drosophila GTPCH isoforms facilitate BH 4 and DAHP inhibition in the absence of an additional regulatory protein, such as GFRP. The truncation mapping of residues responsible for the non-competitive inhibition of the enzyme suggests that the N-terminal domains of isoform A and isoforms B and C, while sharing no similarity in sequence, are all capable of facilitating this regulation (Fig. 8A). Our in vivo studies indicate that this inhibition is not an in vitro artifact, because DAHP ingestion leads to lowered BH 4 pools in adult Drosophila. Moreover, BH 4 pools recover to wild-type levels post-DAHP treatment. Peculiarly, a threshold level of BH 4 reduction is achieved with DAHP feeding, because BH 4 pools could not be completely depleted even after ingestion of a high concentration of DAHP. The residual BH 4 (ranging from 30 to 60% of control levels) is not likely to be synthesized prior to enzyme inhibition, because the cofactor has a short half-life (53)(54)(55). Instead, we hypothesize that the residual BH 4 may be newly synthesized by GTPCH engaged in interactions that sequester DAHP binding sites, consistent with our previous observation that Drosophila GTPCH can be co-immunoprecipitated with tyrosine hydroxylase from head extracts (56). Because whole Drosophila were used in this analysis and because Drosophila has multiple active GTPCH isoforms, it is not possible to determine isoform-specific inhibition pat-terns in vivo by our methods. However, the finding that Drosophila GTPCH can be inhibited by DAHP ingestion in the absence of GFRP allows us to now manipulate GTPCH activity levels pharmacologically and to monitor the effect of GTPCH dysregulation in multiple disease models.
Drosophila GTPCH, like its human counterpart, exhibits buffer-dependent differences in inhibition properties. The differences that we observe, however, are primarily in degree of inhibition only, with the only exception being the lack of positive cooperativity of isoform C in Tris buffer. This difference may be a reflection of the alternative 16-amino acid insertion in the C-terminal portion of the N-terminal domain, which may have an impact on the catalytic core.
Surprisingly, we find that the N-terminal extensions of isoforms B and C provide an additional level of regulation in that they possess native negative regulatory domains. The disproportionately high V max of isoform A (60.89 Ϯ 2.82 nmol of H 2 NTP/min/mg of protein) compared with isoforms B and C (29.07 Ϯ 1.91 and 27.28 Ϯ 0.86 nmol of H 2 NTP/min/mg of protein, respectively) can be attributed to this native negative regulatory region within the N-terminal domains of isoforms B and C. The regulatory sequence predicted by truncation analysis (Fig. 8A) is serine/threonine-rich, suggesting that phosphorylation of residues within the sequence may modulate regulation. This idea is supported by site-directed mutation analysis within the candidate negative regulatory domain of isoform C. Altering Ser 37 to glutamic acid to mimic the phosphorylated state increases the maximum activity of the enzyme, which suggests that phosphorylation of residues within the negative regulatory domain may alleviate the negative regulation imposed on the enzyme.
Structural predictions for the N-terminal extensions are lacking. It has been assumed that the N-terminal domains of Drosophila GTPCH extend away from the core catalytic domain of the full enzyme, as seen in the crystal structures of human GTPCH (50). It is therefore interesting to note that a change of the aforementioned isoform C Ser 37 to alanine drastically lowers the V max , particularly because the complete removal of the negative regulatory region elevates the V max 2-fold. Such distinct effects of modulating a single amino acid within this regulatory region suggest that amino acids within the negative regulatory region may themselves be interacting at or near the active site of the full enzyme and thereby inhibiting the enzyme's activity.
Phosphorylation of residues within the negative regulatory region may alter the conformation of the protein, thereby changing the state of this interaction and leading to an activated enzyme (Fig. 8B). This model for regulation of the enzyme by the negative regulatory region is supported by various lines of evidence, including the truncation and site-directed mutation analyses, as well as the demonstrated phosphorylation of the N-terminal extensions of isoforms B and C by PKA and PKC. Post-translational modification of mammalian GTPCH by PKC has been demonstrated previously, with the phosphorylated enzyme exhibiting elevated activity (32,33). Likewise, we note a significant elevation in enzymatic activity for GTPCH isoforms B and C following treatment with PKA. Although we have not established that regulation by phosphorylation occurs in vivo, FIGURE 7. Site-directed mutagenesis analysis of candidate negative regulatory domain in Drosophila GTPCH isoform C. The serine residue at position 37 of the N-terminal extension of isoform C was mutated to alanine (S37A) and glutamic acid (S37E) and assayed for activity. S37A results in a significant decrease in activity compared with full-length C, whereas S37E results in a significant elevation in activity. Values are means Ϯ S.E. from two independent experiments, each with triplicate determinations. **, p Ͻ 0.01; ***, p Ͻ 0.001 (one-way ANOVA with Bonferroni post test).
our results parallel those obtained in the mammalian cell culture models. Ongoing studies of transgenic Drosophila and tissue extracts will address in vivo functions of phosphorylation.
The predicted site of phosphorylation by PKC for the mammalian enzymes lies within the core domain of the enzymes (Ser 167 , mouse GTPCH), which is conserved in Drosophila (Ser 249 , isoform C). However, we find that the N-terminal extensions of Drosophila isoforms B and C, and not the core catalytic region containing Ser 249 , are phospho-rylation targets for both PKC and PKA. Structure-function analyses will serve to further characterize this regulatory region and pinpoint the sites of phosphorylation within the N termini of the various isoforms. We note that Ser 19 in human GTPCH, which lies in the N-terminal arm of that enzyme, is a predicted target for PKC. It will be interesting to determine whether Ser 19 and Ser 167 in the mammalian protein serve regulatory functions.
Taken together, these data suggest that the various N-termi- The tentative amino acid sequences pertaining to specific regulatory functions have been mapped for each isoform, with each isoform drawn to scale. Sequences required for DAHP inhibition are indicated for each isoform. The serine-rich, candidate negative regulatory domain for isoforms B and C is also indicated. B, a model for N-terminal regulation of Drosophila GTPCH isoforms. The active site for the full GTPCH decameric enzyme is located at the interface between three adjacent subunits (depicted as spheres). The candidate negative regulatory domain (white arrowhead) in the N-terminal extensions (dotted line) of isoforms B and C interacts with the active site of the enzyme in the unphosphorylated state. This interaction results in lower activity for the unphosphorylated enzyme. Upon phosphorylation within the negative regulatory region, a change in conformation results in increased enzymatic activity. Truncation of the N-terminal domain (as with Ϫ58 isoform C) removes the negative regulatory domain, resulting in increased activity. Isoform A has a shorter N-terminal domain of different sequence and has high activity in the unphosphorylated state.
nal arms of Drosophila GTPCH have evolved to serve multiple roles in regulating BH 4 levels. The isoform-specific regulatory properties, most divergent between isoform A and isoforms B and C, help to explain the different spatiotemporal expression patterns of these isoforms during development. The necessity for acute regulation of cellular concentrations of BH 4 cannot be overemphasized. The precise and rapid regulation of GTPCH is a primary requirement for the synthesis and activity of three key neurotransmitters: dopamine, serotonin, and nitric oxide. GTPCH is therefore expected to be a sensitive target for complex mechanisms that modulate each of these neurotransmitter systems.
Although Drosophila lacks the regulatory protein GFRP, the Drosophila enzyme alone is capable of a broad spectrum of the regulatory interactions exhibited or hypothesized for the mammalian enzyme. Thus, despite structural disparities between the Drosophila and mammalian regulatory components, the functional convergence suggests that Drosophila can accurately model organismal consequences of GTPCH mutations. In particular, the capacity to efficiently generate transgenic animals in the Drosophila system suggests that it will be possible to model the effects of genetic variation within regulatory domains and to examine, in vitro and in vivo, functional interactions with downstream target pathway components of dopamine, serotonin, and nitric oxide synthesis.