The mechanism of potent GTP cyclohydrolase I inhibition by 2,4-diamino-6-hydroxypyrimidine: requirement of the GTP cyclohydrolase I feedback regulatory protein.

Inhibition of GTP cyclohydrolase I (GTPCH) has been used as a selective tool to assess the role of de novo synthesis of (6R)-5,6,7,8-tetrahydro-L-biopterin (BH4) in a biological system. Toward this end, 2,4-diamino-6-hydroxypyrimidine (DAHP) has been used as the prototypical GTPCH inhibitor. Using a novel real-time kinetic microplate assay for GTPCH activity and purified prokaryote-expressed recombinant proteins, we show that potent inhibition by DAHP is not the result of a direct interaction with GTPCH. Rather, inhibition by DAHP in phosphate buffer occurs via an indirect mechanism that requires the presence of GTPCH feedback regulatory protein (GFRP). Notably, GFRP was previously discovered as the essential factor that reconstitutes inhibition of pure recombinant GTPCH by the pathway end product BH4. Thus, DAHP inhibits GTPCH by engaging the endogenous feedback inhibitory system. We further demonstrate that L-Phe fully reverses the inhibition of GTPCH by DAHP/GFRP, which is also a feature in common with inhibition by BH4/GFRP. These findings suggest that DAHP is not an indiscriminate inhibitor of GTPCH in biological systems; instead, it is predicted to preferentially attenuate GTPCH activity in cells that most abundantly express GFRP and/or contain the lowest levels of L-Phe.

Four decades ago, DAHP was identified as a potent inhibitor of biopterin-dependent growth of the protozoan Crithidia fasciculata (21). Investigations of biopterin excretion in rats suggested the presence of a de novo biopterin synthesis pathway that is potently inhibited following DAHP treatment (22). Subsequently, biosynthesis of 7,8-dihydrobiopterin from GTP was demonstrated in rat brain, and the responsible multienzyme pathway was found to be inhibited by DAHP (23,24). DAHPinduced BH4 deficiency was later used to generate an animal model for hyperphenylalaninemia (25) as well as a model for neurological BH4 deficiency (26). Additionally, DAHP has been widely used to determine the impact of BH4 availability on NO synthesis in cells in culture (27)(28)(29)(30)(31)(32) and in animals (33,34).
Because of the overt structural homology between DAHP and the GTPCH substrate GTP, the ability of DAHP to inhibit GTPCH has been assumed to result from competition for substrate binding. However, DAHP is also structurally similar to BH4 ( Fig. 1) and, inasmuch as GTPCH is regulated by end product feedback inhibition, DAHP could potentially inhibit GTPCH activity by co-opting the endogenous feedback inhibitory mechanism. Feedback inhibition by BH4 is mediated by the formation of a ternary complex between BH4, GTPCH, and a protein termed GTPCH feedback regulatory protein (GFRP) (35,36). GTPCH is a stable decamer, and in the presence of BH4, feedback inhibition is engaged by the binding of one GFRP pentamer to each of the two pentameric faces of GTPCH. Interestingly, millimolar level concentrations of L-Phe have been shown to reverse this feedback inhibition without causing dissociation of the GFRP⅐GTPCH complex (37). In the absence of BH4, L-Phe triggers GFRP binding to GTPCH resulting in increased GTPCH activity at low levels of GTP substrate (35,38). Because L-Phe is potentially neurotoxic, the ability of L-Phe to stimulate GTPCH activity may have evolved to provide an abundant supply of BH4 cofactor for optimal activity of phenylalanine hydroxylase. A previous observation that DAHP-mediated inhibition of GTPCH in rat aortic smooth muscle cells is partially reversed by addition of L-Phe suggests that the mode of inhibition by DAHP may be GFRP-dependent (39). Nonetheless, DAHP has been widely assumed to selectively inhibit GTPCH activity via competition for binding substrate GTP.
Utilizing a novel kinetic assay for GTPCH activity and purified recombinant rat GTPCH and GFRP from prokaryotic expression systems, we have now conducted a rigorous in vitro analysis of the mechanism of GTPCH inhibition by DAHP. Results confirm that potent GTPCH inhibition by DAHP requires GFRP and recapitulates the endogenous mechanism by which BH4 inhibits its own synthesis.

EXPERIMENTAL PROCEDURES
Expression and Purification of Recombinant GFRP-GFRP cDNA was amplified by PCR from a rat aortic smooth muscle cell cDNA library (Invitrogen). To facilitate insertion into the pET15b expression vector (Novagen), primers for PCR were designed to include the desired restriction enzyme cut sites. The GFRP forward primer (5Ј-CCAGC-CACTCCATATGCCCTAAA-3Ј) contained a site for NdeI digestion, and the GFRP reverse primer (5Ј-AAGTCAGCTCACTCGAGTCATTCC-3Ј) contained an XhoI cut site. PCR was performed by standard methods using 2 l of DNA template and the following schedule: 95°C for 5 min followed by 35 cycles of denaturation (94°C for 45 s), annealing (55°C for 1 min), and elongation (72°C for 1 min), terminating with extension at 72°C for 10 min. The resulting PCR products were separated on a 1% agarose gel containing ethidium bromide and visualized by UV fluorescence. The predicted GFRP cDNA-containing product (255 bp) was observed and gel-purified (Bio 101, GeneClean). The purified PCR product and pET15b plasmid were both digested with NdeI and XhoI and resolved by electrophoresis. Following repurification from the gel, the GFRP insert was ligated into the plasmid using T4 DNA ligase (Invitrogen) at 24°C overnight. On insertion of GFRP cDNA between the NdeI and XhoI sites, the resulting pET15b expression construct encodes a stretch of six His residues fused to the N terminus of GFRP. Competent DH5␣ Escherichia coli (Invitrogen) were transformed with the ligation product, and positive clones were selected using LB agar plates containing ampicillin (100 g/ml). Correct orientation of GFRP cDNA in pET15b was confirmed by restriction enzyme digestion analysis. Plasmid DNA from a positive clone was purified and used to transform the E. coli expression host strain Tuner(DE3)pLysS (Novagen). A clone transformed with the expression construct was grown in LB medium containing ampicillin (100 g/ml) and chloramphenicol (35 g/ml) at 37°C with shaking at 250 rpm. After the bacteria reached an A 600 of 0.5-0.6, expression of the GFRP fusion protein was induced with 1 mM isopropyl-1-thio-␤-D-galactopyranoside (IPTG) at room temperature in the dark for 18 h. Bacteria were harvested by centrifugation, and the resulting pellet was stored at Ϫ70°C until needed. Frozen pellets were thawed and resuspended in lysis buffer (20 mM Tris-HCl, pH 7.9, 500 mM NaCl, 5 mM imidazole, 5 mM ␤-mercaptoethanol, 0.25% Tween 20) containing protease inhibitors (1 g/ml pepstatin A, 1 g/ml leupeptin, 100 M phenylmethylsulfonyl fluoride). Bacteria were lysed by sonication, and the lysate was cleared by centrifugation at 39,000 ϫ g. The supernatant was applied to an affinity chromatography column of 3 ml nickel-nitrilotriacetic-agarose (Qiagen). The column was washed with 50 ml of lysis buffer without protease inhibitors followed by 30 ml of wash buffer (20 mM Tris-HCl, pH 7.9, 300 mM NaCl, 20 mM imidazole). His 6 -GFRP was eluted from the column with elution buffer (20 mM Tris-HCl, pH 7.9, 300 mM NaCl, 500 mM imidazole). Elution fractions were analyzed for purity by SDS-PAGE and Coomassie Blue protein staining. Fractions containing a single strong band at the predicted molecular mass of the fusion protein (11.8 kDa) were combined and subjected to buffer exchange using Sephadex G-25 PD-10 columns (Amersham Biosciences). The fusion protein was exchanged into a buffer containing 20 mM Tris-HCl, pH 7.4, 150 mM NaCl, 550 mM L-Arg. The presence of L-Arg in this buffer was necessary to prevent rapid precipitation of His 6 -GFRP.
Expression and Purification of Recombinant GTPCH-Recombinant rat GTPCH was expressed as a maltose-binding protein (MBP) fusion protein and purified from E. coli as described previously (39). In brief, DH5␣ E. coli were transformed with a pMAL-P2 expression plasmid engineered to express GTPCH-MBP fusion protein. GTPCH-MBP expression was induced by growing bacteria in LB broth with 1 mM IPTG at 37°C for 3 h. Bacteria were harvested by centrifugation and then lysed by sonication. The lysate was cleared by centrifugation and applied to a column of amylose resin for affinity capture of GTPCH-MBP. GTPCH-MBP was eluted from the column with 10 mM maltose and analyzed for purity using SDS-PAGE and Coomassie Blue protein staining. The purified fusion protein exhibited a native level of GTPCH activity at Ϸ40 nmol of product/min/mg protein.
Kinetic Microplate Assay for GTPCH Activity-Formation of dihydroneopterin triphosphate (H 2 NTP), the product of GTPCH, was continually monitored using a spectrophotometry-based kinetic assay (40). A 200-l reaction mixture consisted of the desired additions of the following components: 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), 0.1 M GTPCH, 0.5 M His 6 -GFRP (5-fold molar excess over GTPCH-MBP), 10 mM GTP, 1.5 mM DAHP, and 3 mM L-Phe. The accumulation of H 2 NTP was assessed from the rate of increase in A 340 , measured at 15-s intervals over a 1-h period at 37°C using a Thermomax microplate spectrophotometer (Molecular Devices). From a total of 241 data points collected, 100 were used for computerassisted determination of the GTPCH activity V max .
Determination of Extinction Coefficient and Standard Curve for H 2 NTP-Because H 2 NTP is neither commercially available nor readily amenable to synthesis, standards were produced by complete enzymatic conversion of the desired concentrations of GTP and a fixed excess of GTPCH (40). Assuming the final concentration of H 2 NTP was equal to the initial concentration of GTP, A 340 values were plotted against H 2 NTP concentration to generate a standard curve. Beer's law was used to calculate a H 2 NTP working extinction coefficient of ⑀ 340 ϭ 1820 M Ϫ1 cm Ϫ1 for the assay. The standards and extinction coefficient were used to establish a relationship between A 340 and GTPCH activity. The extinction coefficient for GTP was also calculated (⑀ 340 ϭ 0.091 M Ϫ1 cm Ϫ1 ) and used to determine the extent to which absorbance of GTP interfered with the measurement of H 2 NTP. This absorbance by GTP was determined to be negligible because complete consumption of GTP during the reaction would cause the measured A 340 to decrease by Ͻ0.1% of the total increase in OD.
HPLC-based Assay for GTPCH Activity-The activity of 0.1 M GTPCH activity was measured by spectrometry at 37°C as described for the microplate assay above, and after 60 min, a 100-l aliquot was withdrawn for product analysis using the conventional HPLC-based assay for H 2 NTP (41). To this end, oxidation of H 2 NTP in the sample was achieved by treatment with 20 l of 1 M HCl and 10 l of 0.1 M KI/I 2 for 1 h at 37°C in the dark. Samples were then decolorized with 10 l of freshly prepared ascorbic acid, neutralized with 20 l of 1 M NaOH, and dephosphorylated for 1 h at 37°C by treatment with alkaline phosphatase (20 l of 16 units/ml dissolved in 50 mM MgCl 2 ). After a 2-min microcentrifugation at 13,000 rpm, samples were subjected to C18 reverse phase high performance liquid chromatography, and neopterin content was determined by comparison with a pure standard.
Materials-Restriction enzymes, TaqDNA polymerase, and LB broth base were purchased from Invitrogen. BH4 was from Shircks Labs (Jona, Switzerland). DAHP, L-Phe, GTP, and all other chemicals were obtained from Sigma and were of the best grade available. Oligonucleotides were synthesized by Invitrogen.

Development of a Novel Kinetic Microplate Assay for GTPCH
Activity-H 2 NTP has a markedly greater A 340 than the substrate GTP (42), and this feature was used for the development of a high throughput kinetic microplate assay for GTPCH activity (40). An increase in A 340 was a linear function of time and GTP concentration, in the presence of a fixed amount of recombinant GTPCH enzyme ( Fig. 2A). Activity was also dependant on GTPCH concentration and, after a brief lag phase, was linear for at least 60 min or until substrate GTP became rate limiting (Fig. 2B). A standard curve was generated based on complete conversion of GTP to H 2 NTP (Fig. 3A) as described under "Experimental Procedures." Notably, a linear relationship was established between A 340 and H 2 NTP production (Fig. 3B).
Analysis of GTPCH Activity in the Presence of DAHP Indicates that Inhibition Is Weakly Competitive-To investigate the extent to which DAHP inhibits GTPCH activity via competition for substrate occupancy, recombinant rat GTPCH activity was analyzed over a range of GTP concentrations in the presence and absence of 1.5 mM DAHP, a concentration that effectively inhibits BH4 synthesis in cells (Fig. 4) (39). In the absence of DAHP, the K m and V max for GTP utilization were 75.52 Ϯ 9.78 M and 46.31 Ϯ 1.23 nmol H 2 NTP/min/mg protein, respectively, in accord with published values (39,43,44). On addition of DAHP, the K m for GTP was modestly increased (100.70 Ϯ 17.34 M), and V max was unchanged (47.48 Ϯ 1.89 nmol H 2 NTP/min/mg protein) as determined by nonlinear regression analysis. The very modest DAHP-induced increase in K m was considered insufficient to explain the potent attenuation of BH4 synthesis observed with 1.5 mM DAHP in cells.
Inhibition of GTPCH by DAHP Is Achieved Primarily through an Indirect GFRP-dependent Mechanism-Based on the structural similarity between DAHP and BH4 (see Fig. 1), it was considered that DAHP may be acting as a BH4-mimetic to engage GFRP-mediated feedback inhibition of GTPCH (39). To test this possibility, recombinant GFRP was purified from a prokaryotic expression system (Fig. 5) and characterized with reconstituted components of the feedback inhibition system in vitro. GTPCH activity was measured in the presence of a fixed GTP concentration (10 mM) with increasing concentrations of DAHP in the presence or absence of GFRP. As shown in Fig. 6, DAHP elicited a significant and concentration-dependent inhibition of GTPCH only in the presence of GFRP, in accord with a BH4-mimetic mechanism of GTPCH inhibition. Notably, GTPCH activity was reduced by only 6.5% with DAHP (1.5 mM) alone, whereas the addition of GFRP increased inhibition to 81.4%. Conferral of enhanced GTPCH inhibition by GFRP in vitro suggests that DAHP recruits the feedback inhibitory system for potent inhibition of GTPCH in biological systems.
Reversal of DAHP/GFRP-mediated Inhibition by L-Phe-As discussed above, potent inhibition of GTPCH was DAHP concentration-dependent and required GFRP. The addition of increasing concentrations of L-Phe to DAHP-inhibited GTPCH resulted in complete restoration of enzymatic activity (Fig. 7). The concentration of L-Phe needed for half-maximal reversal of GTPCH inhibition was Ϸ1.5 mM. Notably, no detectable reversal of GTPCH inhibition was found with other aromatic amino acids, Tyr, or Trp at concentrations up to 6 mM. 2 Similarly, BH4-mediated GFRP-dependent inhibition of GTPCH has been found to be reversed by L-Phe and by no other naturally occurring amino acids (35, 36). Verification of the GFRP Dependence and L-Phe Reversibility of GTPCH Inhibition by DAHP-H 2 NTP is a labile compound that undergoes dephosphorylation to dihydroneopterin. Inasmuch as the microplate assay for GTPCH activity relies on spectral properties of H 2 TNTP for quantification, we considered the possibility of an artifact because of a possible DAHP-induced alteration in H 2 NTP stability. Thus, we sought to verify that GFRP is required for inhibition of GTPCH by DAHP using an independent assay method. Results shown in Fig. 8A confirm that DAHP inhibition is indeed GFRP-dependent and L-Phe-reversible as demonstrated by the conventional assay for GTPCH (i.e. relying on HPLC separation and fluorescencebased detection of neopterin in chemically oxidized samples). The Ϸ30% decrease in total H 2 NTP quantified by the HPLC assay compared with the microplate assay (Fig. 8B) 5. Purification of His 6 -tagged rat GFRP from an E. coli expression system. His 6 -GFRP was expressed in bacteria and purified by nickel chromatography as described under "Experimental Procedures." Expression and purification were analyzed by SDS-PAGE on a 10-20% Tris-glycine gel. Protein was visualized by staining with Coomassie Blue. Lane 1, crude lysate prepared from E. coli transformed with the GFRP/pET15b expression plasmid and induced with 1 mM IPTG; lane 2, flow-through after application of crude lysate to a nickel affinity column; lane 3, elute from nickel affinity column on addition of 500 mM imidazole. Note the presence of an IPTG-inducible highly purified protein in the column elute at the predicted molecular mass of His 6 -GFRP (11.8 kDa).
FIG. 6. Influence of GFRP on the ability of DAHP to inhibit GTPCH. The activity of recombinant rat GTPCH (1.44 g, 0.1 M) was measured using the microtiter plate assay as described under "Experimental Procedures." The inhibitory potency of DAHP was assessed in the absence (Ϫ) or presence (ϩ) of recombinant rat GFRP (1.15 g, 0.5 M). The effect of 3 mM L-Phe alone or in combination with GFRP was also determined. Note that concentration-dependent inhibition of GT-PCH by DAHP is achieved only in the presence of GFRP. Reversal of inhibition by L-Phe is in accord with a GFRP-dependent mechanism akin to that established for feedback inhibition by BH4. Values are means Ϯ S.E. from two experiments, each using triplicate determinations for each data point.
observed and attributed to inefficiencies in sample processing for HPLC analysis (i.e. sample oxidation, dephosphorylation, and recovery).
The Molar Ratio of GFRP to GTPCH Determines the Efficacy of DAHP as a GTPCH Inhibitor-To investigate the GFRP dependence of GTPCH inhibition by DAHP, the activity of a fixed amount of GTPCH was studied in the presence of 1.5 mM DAHP and increasing amounts of GFRP. As shown in Fig. 9, the potency of DAHP was enhanced as the ratio of GFRP to GTPCH increased until a ratio of 1:1 at which point inhibition was nearly maximal. These data support the hypothesis that the extent of GFRP protein expression would dictate the degree of GTPCH inhibition by DAHP in cells and tissues. DISCUSSION We describe a novel kinetic assay for quantification and characterization of GTPCH activity in microtiter plate wells. This assay directly and continually measures formation of H 2 NTP in real time, providing more detailed information on GTPCH reaction kinetics than available previously. The microplate format is conservative with respect to reagent consumption and allows for high throughput analysis of GTPCH activity simultaneously, under multiple experimental conditions. A limitation is that the assay can only be used effectively with pure or partially purified GTPCH because of interference by UV-absorbing contaminants in crude biological extracts. This constraint does not hinder our application of the assay to probe the mechanism for potent inhibition of GTPCH by DAHP.
DAHP is a specific inhibitor of GTPCH activity and has been widely used to inhibit de novo BH4 synthesis in biological systems. Despite extensive use of DAHP as an inhibitor of BH4 biosynthesis, the mechanism of inhibition has never been defined experimentally; it has been assumed that inhibition is competitive based on a structural similarity of DAHP with the GTPCH substrate, GTP. It is established that in vivo GTPCH activity is subject to feedback inhibition by the pathway end product BH4. The mechanism of feedback inhibition involves the induced formation of an inhibitory complex between GTPCH and the regulatory protein GFRP. L-Phe reverses this inhibition without disrupting the inhibitory complex. The apparent structural similarity of DAHP and BH4 as well as results from previous studies conducted in rat aortic smooth muscle cells (39) and in vitro (45) suggested a possible role for GFRP in the mechanism of DAHP-mediated inhibition. Based on careful analyses using pure recombinant proteins, our findings confirm that a major mode of DAHP inhibition of GTPCH is indeed dependent on GFRP. In the presence of GFRP, inhibition of GTPCH by DAHP was observed to be potent, concentration-dependent, and completely reversed by L-Phe, which are features that typify the GFRP-dependent GTPCH feedback inhibitory system. GTPCH assays described herein were routinely performed in phosphate buffer (100 mM, pH 7.8). Interestingly, when the mode of inhibition of GTPCH by DAHP was investigated in a non-phosphate buffer system (i.e. 100 mM Tris-HCl, pH 7.8), it was found to be largely competitive. 2 Thus, in contrast to the present findings, inhibition of GTPCH could be observed in the absence of GFRP and was not found to be reversed by L-Phe addition. However, a very recent report (46) shows that prolonged treatment of rats with DAHP results in a decrease in blood pressure by a mechanism that was reversed on cotreatment with L-Phe, in accord with the view that in vivo inhibition of GTPCH by DAHP is indeed GFRP-dependent. The essentially exclusive observation of GFRP-dependent inhibition of GTPCH by DAHP in phosphate buffer may be explained by the binding of phosphate itself to the GTPCH active site, limiting the binding of DAHP as a competitive inhibitor (GTP-mimetic) and switching the mode of inhibition to GFRP-mediated (BH4mimetic). An unpublished structure of GTP-free E. coli GTPCH revealed bound inorganic phosphate that coincides with the position of the GTP ␥-phosphate (47). The contribution of physiological levels of phosphate to GTPCH regulation and interaction with GFRP awaits investigation.
Data presented here strongly support the model in which, under biological conditions, DAHP is functioning as a BH4mimetic to engage GFRP-mediated feedback inhibition of GTPCH. These results may explain in vivo disparities in the efficacy of DAHP as an inhibitor of BH4 synthesis in various tissues. The relative potency of DAHP will predictably depend on the cell content of GFRP and L-Phe.