| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
,1
* Department of Biomedicine, University of Bergen, Bergen, Norway;
Fachbereich Chemie, University of Konstanz, Konstanz, Germany; and Departments of
Chemical and Forensic Sciences,
Clinical and Experimental Dermatology/Department of Biomedical Sciences, University of Bradford, Bradford, UK
1Correspondence: Clinical and Experimental Dermatology, Department of Biomedical Sciences, University of Bradford, Bradford BD7 1DP, UK. E-mail: k.schallreuter{at}bradford.ac.uk
ABSTRACT
Pterin-4a-carbinolamine dehydratase (PCD) is an essential component of the phenylalanine hydroxylase (PAH) system, catalyzing the regeneration of the essential cofactor 6(R)-L-erythro-5,6,7,8-tetrahydrobiopterin [6(R)BH4]. Mutations in PCD or its deactivation by hydrogen peroxide result in the generation of 7(R,S)BH4, which is a potent inhibitor of PAH that has been implicated in primapterinuria, a variant form of phenylketonuria, and in the skin depigmentation disorder vitiligo. We have synthesized and separated the 7(R) and 7(S) diastereomers confirming their structure by NMR. Both 7(R)- and 7(S)BH4 function as poor cofactors for PAH, whereas only 7(S)BH4 acts as a potent competitive inhibitor vs. 6(R)BH4 (Ki=2.3–4.9 µM). Kinetic and binding studies, as well as characterization of the pterin-enzyme complexes by fluorescence spectroscopy, revealed that the inhibitory effects of 7(R,S)BH4 on PAH are in fact specifically based on 7(S)BH4 binding. The molecular dynamics simulated structures of the pterin-PAH complexes indicate that 7(S)BH4 inhibition is due to its interaction with the polar region at the pterin binding site close to Ser-251, whereas its low efficiency as cofactor is related to a suboptimal positioning toward the catalytic iron. 7(S)BH4 is not an inhibitor for tyrosine hydroxylase (TH) in the physiological range, presumably due to the replacement of Ser-251 by the corresponding Ala297. Taken together, our results identified structural determinants for the specific regulation of PAH and TH by 7(S)BH4, which in turn aid in the understanding of primapterinuria and acute vitiligo. —Pey, A. L., Martinez, A., Charubala, R., Maitland, D. J., Teigen, K., Calvo, A., Pfleiderer, W., Wood, J. M., Schallreuter, K. U. Specific interaction of the diastereomers 7(R)- and 7(S)-tetrahydrobiopterin with phenylalanine hydroxylase: implications for understanding primapterinuria and vitiligo
Key Words: Pterin 4a-carbinolamine dehydratase • NMR • isothermal titration calorimetry
THE AROMATIC AMINO ACID hydroxylases phenylalanine hydroxylase (EC 1.14.16.1, PAH) and tyrosine hydroxylase (EC 1.14.16.2, TH) and the tryptophan hydroxylases 1 and 2 (EC 1.14.16.4, TPH1 and TPH2) are highly homologous enzymes that all use molecular oxygen and (6R)-L-erythro-5,6,7,8-tetrahydrobiopterin [6(R)BH4], usually referred to as cofactor, as additional cosubstrates. PAH is the rate-limiting enzyme in the catabolic oxidation of the essential amino acid L-Phe. After the PAH catalytic cycle, the functional tetrahydro form is regenerated from the resulting pterin-4a-carbinolamine (4-OH-BH4) by the coupled action of pterin 4a-carbinolamine dehydratase (EC 4.1.2.96, PCD) that catalyzes the dehydration of 4-OH-BH4 to quinonoid dihydrobiopterin (q-BH2) and the NAD(P)H-dependent dihydropteridine reductase (EC 1.6.99.7 DHPR) that reduces q-BH2 back to 6(R)BH4 (Fig. 1
). The PAH system usually includes these two 6(R)BH4 regenerating enzymes in addition to PAH. The dimeric form of PCD is also a transcriptional activator functioning as dimerization cofactor for hepatic NF 1
. Intracellular 6(R)BH4 levels are maintained and regulated in mammals by de novo synthesis from GTP via the rate-limiting enzyme GTP cyclohydrolase 1 (EC 3.5.4.16; ref 1
).
|
Mutations in PAH resulting in impaired or absent enzyme activity are associated with phenylketonuria (PKU), an inborn error of L-Phe metabolism characterized by hyperphenylalaninemia (HPA) and accumulation of toxic breakdown metabolites of L-Phe leading to mental retardation in untreated patients. Moreover, decreased or absent PAH activity can lead to a deficiency of L-Tyr and its downstream products, including melanin, thyroxine, and the catecholamine neurotransmitters. Several forms of PKU or HPA are also associated with mutations in the enzymes that synthesize or recycle the cofactor 6(R)BH4 (1
, 2)
. Primapterinuria (OMIM 264070) is a mild (or atypical) form of PKU in which patients show both HPA and excretion of 7-substituted pterins including 7-L-biopterin (primapterin; ref 3
4
5
). Curtius et al. (6)
were the first to predict that this condition was due to an accumulation of 4-OH-BH4 during the PAH reaction in the absence of PCD activity, because its nonenzymatic dehydration to q-BH2 is not fast enough to maintain full regeneration of 6(R)BH4. The accumulated 4-OH-BH4 spontaneously produces 7-BH4 due to pyrazine ring opening and migration of the L-erythro side chain from C6 to C7 (7)
. Patients with primapterinuria due to mutations in the PCD gene (8
, 9)
are treated with 6(R)BH4 supplementation until HPA disappears and, where necessary, with a L-Phe restricted diet (5)
. An excess of 7-BH4 and its oxidation to 7-L-biopterin together with the production and accumulation of millimolar concentrations of H2O2 occur in the epidermis of patients with the depigmentation disorder vitiligo. In this context, it was shown that H2O2-mediated oxidation of Met and tryptophan (Trp) residues in the PCD structure significantly alters the active site of the enzyme, leading to deactivation and consequently to accumulation of 7-BH4 (10
, 11)
.
The 6(R)BH4-dependent PAH reaction is fully coupled, with a 1:1 ratio between L-Tyr hydroxylation and cofactor oxidation. Davis and Kafman (12
, 13)
reported that 7-BH4 functioned as a poor cofactor, uncoupling the reaction by 85% and as a potent inhibitor of rat PAH at concentrations as low as 1 µM in the presence of physiological concentrations of 6(R)BH4. The uncoupled reaction implies production of H2O2, causing further inactivation of PAH. Moreover, earlier investigations showed that 7-BH4 is a poor inhibitor of TH and TPH activities. Hence, it was concluded that the 7 isomer had no significant effect on these enzyme activities in vivo (13)
.
It is noteworthy that 6-BH4 presents two diastereomers, but only the 6(R) compound is enzymatically synthesized and functions as the natural cofactor for the aromatic amino acid hydroxylases and the NO synthases (14)
. Large differences are observed in the enzyme kinetic constants and the regulatory properties of PAH in the presence of 6(S)BH4 compared with 6(R)BH4. The latter diastereomer elicits regulatory inhibitory conformational changes in the absence of substrate, as well as substrate inhibition with high concentrations of L-Phe (15
16
17)
. The Km values are about half with the natural cofactor compared with the 6(S)BH4 diastereomer (16)
. 7-BH4 is produced as a 7(R,S) diastereomer mixture containing 57% (R) and 43% (S) as confirmed by 1H 13C NMR (18)
. Therefore, we hypothesized that each diastereomer exhibits different kinetic and conformational properties for PAH. To follow and characterize their specific effects on the PAH reaction, we synthesized and purified the two diastereomers. Our results showed that both act as cofactors for PAH but display lower affinities and catalytic efficiencies compared with 6(R)BH4. However, only 7(S)BH4 significantly inhibits the enzyme in the presence of physiological 6(R)BH4 concentrations. Molecular dynamics (MD) simulations were used to further elucidate the structure-function relationships for each of the diastereomers with PAH. Studies with the homologous enzyme TH did not yield significant inhibition by 7(S)BH4. This result is most likely due to the substitution of Ser-251 in PAH with Ala297 in TH.
MATERIALS AND METHODS
Materials
6(R)BH4, 7(R,S)-5,6,7,8-tetrahydro-L-primapterin [7(R,S)-BH4], and 7-L-biopterin were from Schircks Laboratories (Jona, Switzerland). All other chemicals and reagents were purchased from Sigma. Recombinant human PAH and recombinant human TH isoform 1 (TH1) were expressed and purified as tetramers as described previously (19
, 20)
.
Synthesis and separation of 7(R)- and 7(S)BH4 diastereomers
The synthesis and separation of 7(R) and7(S)BH4 are summarized in Fig. 2
where the individual steps for the pathways are numbered 1–6 and are indicated in the text with italics.
|
Synthesis of 2-N,N 5,1'-O,2'-O-tetraacetyl-7-(R,S)-5,6,7,8-tetrahydro-L-biopterin (2)
2-N,N5,1'-O,2'-O-Tetraacetyl-7-(R,S)-5,6,7,8-tetrahydro-L-biopterin (2) was synthesized from 7-L-biopterin (1; 0.24 g) by reduction with PtO2 (50 mg) reduced in H2O (25 ml) under H2 (2 equiv). The product was acetylated with Ac2O (10 ml) by shaking for 30 min. The catalyst was removed by filtration through Celite (Sigma). The filtrate was evaporated, and the final residue was treated with pyridine (2 ml) and Ac2O (15 ml) at 55°C for 15 min, followed by stirring at room temperature for 3 h. The resulting pale yellow solution was evaporated under vacuum, coevaporated with toluene (10 ml), AcOEt (10 ml), and CH2Cl2 (2x10 ml) yielding a solid foam. Thin layer chromatography (TLC; twice developed) in CH2Cl2/MeOH (9:1) gave two separate spots with Rf = 0.31 and Rf = 0.38.
Separation of 2N,N5,1'-O,2'-O-tetraacetyl-7(R)-5,6,7,8-tetrahydro-L-biopterin (3)
The crude mixture of 2N, N51'-O,2'-O-tetraacetyl-7(R,S)-5,6,7,8-tetrahydro-L-biopterin (2)
was treated under stirring with MeOH (8 ml), whereby a crystalline precipitate was formed. The colorless solid was collected and dried under vacuum yielding 0.114 g (28%) of the lower running diastereomer that was characterized by 1H 13C NMR as the 7(R) diastereomer. TLC silica gel showed Rf = 0.31 (twice developed). UV spectrum in MeOH showed maxima at 231 nm (4.53) and 297 nm (3.88) with a shoulder at 271 nm (3.79). The empirical formula was confirmed by analysis as C17H23N5O7 x 2H2O (445.4).
Separation of 2-N,N5,1'O,2'O-tetraacetyl-7-(S)-5,6,7,8-tetrahydro-L-biopterin (4)
The filtrate of the preceding procedure was evaporated, and the residue was dissolved in CHCl3 and applied to a silica gel column (8x2.5 cm) for chromatography with CHCl3/MeOH (50:1) followed by (50:2), which eluted the upper Rf (S) diastereomer. The combined product fractions were evaporated to give a colorless solid (0.15 g) that was dissolved in warm MeOH producing colorless crystals (45 mg). The residual filtrate was purified a second time by silica gel chromatography yielding a second 45 mg batch of crystals (total yield 22%). TLC gave an Rf of 0.38. This S diastereomer was characterized by 1H 13C NMR. The UV spectrum in MeOH showed maxima at 231 (4.58) and 294 nm (3.90) with a shoulder at 273 nm (3.83). The empirical formula was confirmed by analysis as C17H23N5O7 xH2O (427.4).
Synthesis of 7(R)-5,6,7,8-tetrahydro-L-biopterin dihydrochloride (5)
The fully acetylated 7(R) diastereomer (3) (95 mg) was treated with 1.25 M MeOH/HCl (5 ml) with stirring at room temperature for 24 h where it dissolved. The clear solution was evaporated, coevaporated with MeOH (2x10 ml) and Et2O (10 ml). The resulting solid was stirred with Et2O (15 ml), collected, and dried under vacuum yielding 48 mg. The filtrate was again evaporated to give another 17 mg of product. The total yield was 65 mg (90%). The structure of the R diastereomer was confirmed by 1H 13C NMR (Table 1A
). The Rf on cellulose sheets was 0.70 (solvent n-BuOH/ 5N AcOH 1:2).
|
Synthesis of 7(S)-5,6,7,8-tetrahydro-L-biopterin dihydrochloride (6)
The fully acetylated 7(S) diastereomer (4; 87 mg) was treated by the above procedure yielding 64 mg (97%) of product. TLC on cellulose sheets gave an Rf of 0.73. The structure of the S diastereomer was confirmed by 1H 13C NMR (Table 1B)
.
|
1H 13C NMR Spectroscopy
1H 13C NMR spectra were acquired at 30°C on a JEOL ECA600 NMR Spectrometer equipped with a 5 mm autotune (XH) multinuclear pulse-field gradient probe using JEOL Delta Software. 7(R,S)BH4 (Schircks) contained 57% (R) and 43% (S) diastereomer mixture (18)
. 7(R) and 7(S)BH4 were prepared by dissolution in 99.9% D2O at 1.0 mg/ml. 1H 13C assignments were made using a combination of 1D and 2D experiments including 1H-1H DQF-COSY, 13C(1H) DEPT135, HMQC (1JC-H correlations), and HMBC (2JC-H and 3JC-H correlations). Where it was not possible to detect long-range correlations, assignments were made with the assistance of a-chlorohydrin/Labs CNMR spectral prediction software. (1H 13C NMR assignments in Table 1
).
Enzyme assays
PAH activity was determined at 25°C, pH 7.0, for 1 min using 0.5 or 3 µg human tetrameric PAH with either 6(R)BH4, 7(R)BH4, or 7(S)BH4, respectively, as described previously (21)
, with quantitation of L-Tyr formed by HPLC and fluorimetric detection (22)
. Measurements were carried out in the presence of catalase (0.05 mg/ml final concentration in assay) to avoid accumulation of H2O2 with the 7-BH4 isomers as cofactors and with an incubation step with 1 mM L-Phe for 5 min (L-Phe activated enzyme) before triggering the reaction by adding the pterin cofactor [with 5 mM dithiothreitol (DTT)]. When indicated, the incubation with the substrate was omitted (nonactivated enzyme) and the reaction triggered by simultaneous addition of L-Phe and the corresponding 6- or 7-substituted BH4. The kinetic parameters were measured at 1 mM L-Phe with variable cofactor concentrations [0–200 µM for 6(R)BH4 and 0–800 µM for 7(R)- and 7(S)BH4]and at 75 µM 6(R)BH4 or 800 µM 7(R)- or 7(S)BH4 and variable L-Phe concentrations (0–2 mM). The experimental data were fit to Michaelis-Menten or Hill equations for dependences on cofactor or L-Phe concentration, respectively, using SigmaPlot 2000 (Statistical Packages for the Social Sciences). The inhibitory properties and Ki values for 7(R)- and 7(S)BH4 were measured by adding 5 µM of either 7(R)- or 7(S)BH4 together with different concentrations of 6(R)BH4 and were analyzed using standard equations for reversible inhibition. The inhibitory effect at physiological cofactor concentration was measured using 5 µM 6(R)BH4, 0.1–1 mM L-Phe, and 0–12.5 µM 7(R)BH4 or 7(S)BH4 as inhibitors.
The coupling efficiency of the PAH reaction, i.e., ratio L-Tyr formation to tetrahydropterin oxidation, was measured as previously reported (23)
using a reaction mixture similar to that used for the HPLC-based assays with 1 mM L-Phe but including 200 µM NADH and 0.4 U dihydropteridine reductase (from sheep liver, Sigma Aldrich). The reaction was started by adding 75 µM of the corresponding tetrahydropterin cofactor, and 
A340 was measured for 1 min to calculate the rate of NADH oxidation (pterin oxidation) by using 
340 nm = 6220 M–1·cm–1. Fifty microliter aliquots of the reaction were then taken to measure L-Tyr formation by HPLC and fluorimetric detection.
TH activity was measured at 25°C essentially as described previously (24)
, using a reaction mixture containing 0.2–0.6 µg purified TH1, 0.1% w/v BSA, 0.05 mg/ml catalase, 50 µM L-[3,5-3H]tyrosine (Amersham Pharmacia Biotech), and 20 µM ferrous ammonium sulfate. The reaction was initiated by adding tetrahydropterin [0–1 mM for 6(R)BH4 or 0–2 mM for the 7-BH4 diastereomers] with 5 mM DTT and stopped after 4 min with 1 ml of charcoal (7.5 w/v) in 1 M HCl. After centrifugation, supernatants were counted in a scintillation counter. The inhibitory properties of each of the 7-BH4 diastereomers, including Ki values, were studied in TH reactions triggered by adding 6(R)BH4 (0–1 mM) with 5 µM of either 7(R)BH4 or 7(S)BH4. Kinetic parameters were estimated by fitting to hyperbolic or Hill equations, using Sigma Plot 2000.
Isothermal titration calorimetry
Equilibrium binding experiments were performed using a VP-isothermal titration calorimetry (ITC) titration calorimeter (MicroCal). Titrations were performed in 20 mM Na HEPES, 200 mM NaCl, pH 7.0 under anoxic conditions by using a glucose (Glc)/glucose oxidase/catalase system as described previously (25)
. Spectrophotometric measurements indicated no significant oxidation of the 7-BH4 diastereomers in this anoxic environment. With 6(R)BH4 and 7(S)BH4, titration was typically carried out by loading tetrameric PAH fusion protein MBP-p-aminohippuric acid (25 µM subunit) and ferrous ammonium sulfate (0.5 mol/mol PAH subunit) into the sample cell and 0.5 mM of 6(R)BH4 or 7(S)BH4 into the syringe. For 7(R)BH4, 75 µM PAH subunit and 1.5 mM of the tetrahydropterin were used because of low binding heat (H
0). In each run, a first 4 µl injection followed by 32 injections of 8 µl of each tetrahydropterin were performed (4 min spaced) under 300 rpm constant stirring. The mean of the last three to five injections was used as experimental heat dilution and subsequently subtracted from the raw data. The corrected binding heats were fitted to a single set of identical sites using Microcal Origin v.7.0. software, obtaining the affinity constant (Ka), number of sites (n), binding enthalpy (H), entropy (S), and free energy (G) changes.
Fluorescence
Fluorescence measurements were performed on a Cary Eclipse fluorescence spectrophotometer equipped with a temperature controlled Peltier multicell holder (Varian). Samples contained PAH (2 µM subunit) in 20 mM Na HEPES, 200 mM NaCl, pH 7.0, 10 µM ferrous ammonium sulfate, 2 mM DTT, and 0–10 µM 6(R)BH4, 7(R)BH4, or 7(S)BH4. Intrinsic tryptophan fluorescence spectra were acquired at 25°C with excitation at 295 nm (5 nm slit) and emission at 300–400 nm (10 nm slit) and were corrected for blank emission. Thermal denaturation was monitored by following changes in ANS (8-anilino 1-naphtalene sulfonic acid; Sigma Aldrich) fluorescence emission (excitation at 395 and emission at 500 nm, 5.0 nm slit) at a 1.5°C/min in a 20–75°C range in samples similar to those used for the tryptophan fluorescence but using 200 µM tetrahydropterin instead and adding 100 µM ANS. Tm values were estimated as those temperatures at which half-denaturation occurred (fu=0.5) on fitting the experimental fluorescence (F) to the following equation:
 |
MD simulations, structure-energetics calculations, and molecular interaction fields
The structural models of full-length human PAH bound to either 6(R)BH4 [p-aminohippuric acid·6(R)BH4), 7(R)BH4 p-aminohippuric acid·7(R)BH4], or 7(S)BH4 [p-aminohippuric acid·7(S)BH4] were obtained by MD simulations as described earlier (17)
. Briefly, the models with the respective cofactor were prepared based on the dimeric structure of the rat enzyme (26
; PDB 2PHM) and aligned with the structure of the binary complex between the catalytic domain of human PAH and 6(R)BH4 (27)
to use the pterin ring as a template for appropriate attachment of the corresponding side chains at C7. Then, MD simulations using an implicit treatment of water were performed for 1 ns. The minimized average structures from the last 50 ps of the simulations were used to calculate the changes in polar and apolar solvent accessible surface area, ASApolar and ASAapolar, respectively, using GETAREA (28)
. The structure-based calculations of the thermodynamic parameters were then based on previously published formalisms (29
, 30)
, as described previously (25)
.
Grid calculations and preparation of the molecular interaction fields (MIFs) or maps were performed as reported previously, using the chemical probes O1, alkylhydroxy OH group, and C3, methyl CH3 group (27)
.
RESULTS
7(R,S)BH4 inhibits the PAH reaction
Over the past, several authors have provided evidence of a dual function for 7(R,S)BH4 serving both as a cofactor and as inhibitor of the rat PAH enzyme (7
, 12
, 13)
. We re-examined both functions using 7(R,S)BH4 and recombinant human PAH. Our results showed a Km value of 220 ± 41 µM and a Vmax of 0.52 ± 0.02 µmol L-Tyr/min/mg for this cofactor, with an 8-fold lower affinity and activity compared with 6(R)BH4. The kinetic parameters for L-Phe were not significantly different with 7(R,S)BH4 compared with 6(R)BH4, except for the 8- to 10-fold lower Vmax. Moreover, we confirmed that the oxidation of 7(R,S)BH4 is only 15–20% coupled to the formation of L-Tyr with human PAH, as previously reported for the rat enzyme (12)
, with substoichiometric formation of H2O2 with respect to 7(R,S)BH4 oxidation. 7(R,S)BH4 competitively inhibited the PAH reaction vs. 6(R)BH4 (Ki-value=7±1 µM). This result is in agreement with a Ki = 7–8 µM reported by other groups (7
, 13)
. The elucidation of the molecular mechanisms involving 7(R,S)BH4 in the inhibition of hepatic PAH in primapterinuria (13)
and of epidermal PAH in patients with vitiligo (31)
is complicated, because all results published were obtained with the diastereomer mixture. To provide more detailed insights into these two disorders and possible treatment strategies, we synthesized and successfully separated the 7(R) and 7(S) diastereomers of 7(R,S)BH4 to substantiate their specific effects on PAH (Fig. 2
; Table 1
).
Both 7(R)BH4 and 7(S)BH4 are cofactors for the PAH reaction
Our first attempt was to follow the kinetic properties of human PAH using each of the 7-BH4 diastereomers as cofactors. At saturating L-Phe concentrations (1 mM), PAH displayed similar specific activities and apparent affinities for both 7(R)- and 7(S)BH4. With these cofactors, the Vmax is 10-fold lower and the Km values are 5- to 7-fold higher than with 6(R)BH4 (Fig. 3
A; Table 2
). Although PAH is tightly coupled to L-Tyr formation and cofactor oxidation with 6(R)BH4 as cofactor (coupling efficiency 1.0; Table 2
), in the presence of both 7(R)BH4 and 7(S)BH4 a large degree of uncoupling is observed with coupling efficiencies of 0.4 and 0.2 for 7(R)BH4 and 7(S)BH4, respectively (Table 2)
. Clearly, this uncoupling could partially explain the low activity observed when the isolated 7 diastereomers or the 7(R,S)BH4 mixtures are used as cofactors.
|
|
The 6(R)BH4-dependent PAH reaction is characterized by specific regulatory properties due to activation by the substrate, as shown by an increase in specific activity on L-Phe preincubation and by positive cooperativity observed on the activity vs. substrate concentration curves (Hill coefficient, h, 2; ref 32
, 33
). These properties are maintained with either 7(R)BH4 or 7(S)BH4 as cofactors, but the S0,5 for L-Phe was slightly decreased with 7(S)BH4 (Fig. 3B
; Table 2
). Substrate inhibition of the PAH reaction was observed at high substrate concentrations (>1 mM L-Phe) with all three tetrahydropterin cofactors but more notably for 6(R)BH4 and 7(S)BH4 (data not shown).
PAH is inhibited by 7(S)BH4 but not by 7(R)BH4
To explore possible differential inhibition of PAH by each of the 7-BH4 diastereomers, we compared their effect on the 6(R)BH4-supported enzyme activity under different conditions (Fig. 3C-F
). 7(S)BH4 acts as a competitive inhibitor vs. 6(R)BH4 affecting the Km but not the Vmax, whereas no significant inhibitory effect was observed with 7(R)BH4. The inhibition was stronger for the L-Phe-activated enzyme, with Ki values of 2.3 and 4.9 µM on L-Phe preincubation and without preincubation, respectively. The inhibitory effect of 7(S)BH4 is more dramatic at high concentrations of L-Phe but still significant at lower, more physiological concentrations of the substrate. Thus, at 5 µM 6(R)BH4, an equimolar concentration of 7(S)BH4 causes inhibition of 55, 30, and 10% at 1, 0.25, and 0.1 mM L-Phe, respectively, for enzyme preincubated at the same concentration of L-Phe as in the assay (Fig. 3D, F
).
Equilibrium binding studies of 7(R)BH4 and 7(S)BH4 by ITC
The equilibrium binding of 6(R)BH4, 7(R)BH4, and 7(S)BH4 to PAH was studied by ITC at pH 7.0 and 25°C. Binding isotherms showed stoichiometric binding of the three tetrahydropterins (data not shown). The resulting binding thermodynamic parameters are compiled in Table 3
. 7(R)BH4 binding displayed very low enthalpy change (H=–0.6 kcal/mol), indicating an almost entirely entropically driven binding process. In addition, the binding affinity is low [(1/Ka)=Kd=16.1±3.9 µM, 20- and 3-fold higher than for 6(R)BH4 and 7(S)BH4, respectively]. 7(S)BH4 binding to PAH is entropically driven (–TS=–5.1 kcal/mol) and displays a modest and favorable enthalpic contribution (H=–2.0 kcal/mol), rendering a Kd 8-fold higher than that for 6(R)BH4. For 7(S)BH4, we were able to analyze the binding at temperatures ranging from 5 to 25°C (Table 3)
. From the slope in the linear fitting of H vs. temperature, a change in heat capacity (Cp) of –65 ± 16 cal/mol/K can be estimated for the binding of 7(S)BH4 to PAH, much lower than the Cp previously obtained for 6(R)BH4 (–357 cal/mol/K) and in fact very similar to that for 6M-PH4 binding (–63 cal/mol/K; ref 25
). For 7(R)BH4, the low binding heat precluded the reliable calculation of the Cp.
|
We also investigated the binding of both 7-BH4 diastereomers to PAH preincubated with 1 mM L-Phe. In contrast to 6(R)BH4 binding to PAH, which is strongly affected by preincubation with L-Phe (4-fold decrease in affinity and a 5 kcal/mol reduction in the favorable enthalpic contribution; Table 3
and ref 25
), L-Phe only slightly affected the binding parameters for 7(R)- and 7(S)BH4. These results indicate that the stronger inhibition of PAH by 7(S)BH4 on L-Phe preincubation (2-fold higher Ki; Fig. 3
) is probably more related to the decreased affinity of the L-Phe-activated enzyme for 6(R)BH4 (4-fold decreased affinity vs. the nonactivated enzyme) than to changes in the affinity for 7(S)BH4.
PAH was also titrated using simultaneous addition of both 7(R) and 7(S) diastereomers at equimolar concentrations (0.5 mM of each). The binding parameters obtained in this cotitration resembled those displayed when only 7(S)BH4 is injected (Table 3)
, indicating that under these experimental conditions, the binding isotherm is dominated by the binding of 7(S)BH4 [H is 3-fold larger for the binding of 7(S)BH4 over 7(R)BH4].
Modeled structure of PAH complexed with 7(R)BH4 and 7(S)BH4 and comparison of measured and calculated thermodynamic binding parameters
The available NMR-based and crystal structures of the complexes of 6(R)BH4 and the oxidized cofactor dihydrobiopterin (BH2) with N-terminal truncated forms of PAH show that the cofactor binds to PAH by interactions of the pterin ring with an invariant phenylalanine residue (Phe254 in human PAH) and a hydrogen bonding network involving positions N1, N2, N3, N5, and N8 of the pterin ring (34
35
36)
. Moreover, the hydroxyl groups in the side-chain at C6 establish hydrogen bonds with the carbonyl group of Ala322 and the hydroxyl group of Ser-251. To obtain a better understanding of both the structural and regulatory effects of the binding of the 7(R) and 7(S) diastereomers, we prepared models of full-length PAH, also including the N-terminal domain, bound to either 7(R)BH4 or 7(S)BH4 by combination of docking and MD as previously reported for the complexes of PAH with 6(R)BH4, 6(S)BH4, and 6M-PH4 (17
, 25)
. The MD-converged structures of unbound PAH and the PAH·6(R)BH4, PAH·7(R)BH4, and PAH·7(S)BH4 complexes are presented in Fig. 4
A–D. For bound 6(R)BH4 (Fig. 4B
), the pterin ring perfectly stacks to Phe254, and the position to be hydroxylated, i.e., C4a, and the iron are at 4.4 Å from each other (17
, 25
; Fig. 4B
). However, for both the PAH·7(R)BH4 (Fig. 4C
) and PAH·7(S)BH4 (Fig. 4D
) complexes, but notably for the former, a much less optimal stacking between the pteridine ring and Phe254 was observed. Moreover, slightly longer distances between C4a and the iron were measured in the PAH·7(R)BH4 (4.6 Å) and PAH·7(S)BH4 (4.8 Å) complexes, vs. 4.4 Å in the PAH·6(R)BH4 complex. These alterations in structure could explain the lower affinity and H, as well as the uncoupling of the reaction by the 7-BH4 diastereomers. In addition, high affinity binding of 6(R)BH4 induces conformational changes leading to a closure of the active site by rearrangement of the N-terminal autoregulatory sequence. These regulatory conformational change involve the concerted action of Ser-23, Ser-251, and the hydroxyls at the side chain at C6 of the natural cofactor (17
; Fig. 4B
). These conformational changes are not complete in the PAH·7(R)BH4 and PAH·7(S)BH4 complexes.
|
To further analyze the specific binding determinants at the cofactor site in PAH, we docked the simulated conformations of the 7-BH4 diastereomers into the GRID/MIF maps obtained earlier for this enzyme with the hydroxyl and the methyl probes (27)
. As seen in Fig. 5
A, the MIFs show that the hydroxyls at the dihydroxypropyl side chain at C7 in 7(S)BH4, and notably O1', are very close to an area of high affinity for the hydroxyl probe but does not interact with the methyl probe. This region of high affinity for the hydroxyl groups between the side chain of Ser-251 and the backbone carbonyl of Leu249 probably contributes to the higher affinity for the 7(S) diastereomer and the specific inhibition of the 7-substituted analogues for PAH compared with the other hydroxylases (Fig. 5A
and ref 27
) .
|
These simulated structures can also be applied to examine structure-energetic relationships. Thus, for the PAH·7(S)BH4 complex, for which an experimental Cp value was available, it was interesting to calculate a structural-derived Cp (Cp,cal). The MD-simulated structure provided ASApolar and ASAapolar values of –247 and –325 Å2, respectively, which gave a Cp,cal = –98 cal/mol/K. This value is slightly higher than the experimental Cp value (–65 cal/mol/K, see above). For the binding of PAH·6(R)BH4, the difference between the theoretical (–85 cal/mol/K) and the experimental (–357 cal/mol/K) Cp values is much higher (25)
. This difference has been associated to the large conformational change induced on 6(R)BH4 binding, which is not taken into account in the structure-energetics calculations due to the difficulty of estimating this contribution in the absence of high-resolution structures of both complexed and unliganded protein (30)
. Thus, these changes, which are notably associated to the reorganization of the N-terminal autoregulatory sequence (17
, 25
, 37)
, do not appear to be fully displayed on 7(S)BH4 binding, even less in the case of 7(R)BH4 binding for which the MD-structure derived calculations provided values of ASApolar, ASAapolar, and Cp,cal of –227 Å2, –350 Å2, and –83 cal/mol/K, respectively.
Conformational effects of 7(R)BH4 and 7(S)BH4 binding studied by fluorescence spectroscopy
Putative changes on the tertiary structure associated to 7(R)BH4 and 7(S)BH4 binding were investigated by Trp emission fluorescence spectroscopy. The Trp fluorescence of PAH is largely quenched on tetrahydropterin binding in a concentration dependent manner (38)
. The fluorescence intensity at 340 nm was reduced 75, 65, and 55% in the presence of a 5-fold molar excess of 6(R)BH4, 7(S)BH4, and 7(R)BH4, respectively (Fig. 6
A). The degree of quenching appears to be proportional to the saturation fraction of the pterin binding site (39)
. In fact, the concentration dependence of Trp fluorescence quenching observed in the presence of the pterins (Fig. 6A
, inset) is in agreement with the significantly higher affinity for 6(R)BH4 than for both 7-BH4 diastereomers as determined by ITC (Table 3)
. Pterin-induced Trp quenching has been associated to a Förster energy transfer process between Trp residues and the pterin (38)
. The quenched residue was later identified as Trp120 (39)
, which is located at 18 Å from the bound BH4 in the crystal structure (PDB 1KW0). At saturating conditions, all three pterins are able to quench Trp fluoresence to a similar extent (Fig. 5A
, inset), suggesting that these pterin quenchers share the same binding site at similar distances and/or orientations toward Trp120.
|
Further structural insights on the enzyme-tetrahydropterin complexes were inferred from thermal denaturation studies. The thermal unfolding of PAH was monitored by following the increase in ANS fluorescence on binding of this hydrophobic dye to unfolded and/or partially unfolded states (40)
. Since unfolding was irreversible, no further thermodynamic analyses were applied. However, at saturating concentrations of the ligands (200 µM vs. 2 µM subunit PAH), a clear stabilizing effect was observed only on 6(R)BH4 binding. The half denaturation temperatures (Tm) for the thermal unfolding of human PAH were 50.0, 53.0, 49.9, and 50.6°C for the unbound, 6(R)BH4-, 7(R)BH4-, and 7(S)BH4-bound PAH, respectively (Fig. 6B
). The 6(R)BH4-bound conformation of PAH can be defined as a "closed" structure, with a higher resistance toward proteolysis and increased conformational stability (23
, 25)
. Thus, our thermal dependent fluorescence results further corroborate the stronger, i.e., higher affinity, binding for 6(R)BH4 compared with the 7-BH4 compounds (40)
, and also suggested that the binding of 7(R)BH4 and 7(S)BH4 leads to a less closed PAH conformation (30)
. This is in agreement with the results from the MD simulations (see above).
Kinetic effects of 7(R)BH4 and 7(S)BH4 on the TH reaction
Earlier studies have shown a negligible inhibitory effect of 7(R,S)BH4 on the activity of either TH or TPH, with high IC50 concentrations (0.48–2.5 and 3.3 mM for the inhibition of the 6(R)BH4-dependent activity of rat TH and rabbit TPH2, respectively (13)
, and 3.8 mM for the inhibition of human TPH1 (27)
. As a cofactor, 7(R,S)BH4 also showed a very high Km (10- to 20-fold higher than that for 6(R)BH4; ref 13
) with 80% of reaction uncoupling and consequent H2O2 production (41)
. Thus, it has been inferred that, contrary to the effect on PAH, 7(R,S)BH4 will not have deleterious effects on the activity of the other aromatic amino acid hydroxylases in vivo.
We further investigated the kinetic properties of recombinant human TH1 with the two 7-BH4 diastereomers. Both 7(R)BH4 and 7(S)BH4 acted as cofactors for L-DOPA formation by TH1 but displayed a 9-fold lower specific activity than 6(R)BH4 (Fig. 7
A, Table 4
). Dependence of TH activity on 7(R)BH4 and 7(S)BH4 concentrations revealed an almost hyperbolic (h1), nonregulated behavior, in contrast to the negative cooperativity displayed with 6(R)BH4 (h0.6; Table 4
and ref 42
). The S0.5 value for 7(S)BH4 was similar to that observed for 6(R)BH4, whereas for 7(R)BH4 the affinity was lower. Unfortunately a direct comparison between apparent affinities for the 7-BH4 diastereomers and 6(R)BH4 is not straightforward [due to negative cooperativity for 6(R)BH4].
|
|
The dependence of TH activity on 6(R)BH4 concentration was also analyzed in the presence of low concentrations of the 7-BH4 diastereomers (Fig. 7B
, Table 4
). None of the 7-BH4 diastereomers seem to act as competitive inhibitors, whereas both affected the Vmax values. The inhibitory effect is larger for for 7(S)BH4 than for 7(R)BH4, i.e., 25 and 14% reduction in Vmax, respectively [Ki=17 µM for 7(S)BH4 and 48 µM for 7(R)BH4]. This noncompetitive inhibition may also be explained by an inactivation of TH1 in the presence of the 7-BH4 diastereomers due to uncoupling of the reaction (41)
.
DISCUSSION
Human PAH is inhibited by 7(S)BH4
The production of high concentrations of 7(R,S)BH4 from 6(R)BH4 in human health and disease is a relatively rare event (1)
. To date only two conditions have been reported where the accumulation of primapterin, the oxidation product from 7-BH4 occurs, i.e., primapterinuria and vitiligo (8
, 10
, 12
, 13
, 43)
. In normal healthy individuals, PCD is tightly coupled to PAH, increasing its activity 7-fold (6, 11–13, 44). Reduction of PCD activity thus results in a build up of 7-BH4 and L-Phe (5
, 6
, 31
, 45)
. The abiogenic mechanism for 7(R,S) formation does not confer stereoselectivity, and the 7-BH4 produced is a mixture of 7(R) and 7(S) diastereomers (43)
. Earlier 1H 13C NMR studies showed that the only commercially available 7-BH4 (Schircks) is a 57% (R), 43% (S) mixture (18)
, and consequently the cofactor and inhibitory roles on PAH have not been fully understood (10
, 13
, 43)
.
Here we have synthesized, separated, and characterized 7(R)BH4 and 7(S)BH4 (Fig. 2
, Table 1
). A detailed kinetic analysis showed that both diastereomers worked as inefficient cofactors for PAH compared with 6(R)BH4, displaying low coupling (20–40%) and catalytic efficiencies [Vmax/Km 2% of that for 6(R)BH4]. Only 7(S)BH4 was a potent inhibitor of PAH activity (Fig. 3
, Table 2
), acting competitively vs. 6(R)BH4. This inhibition increased in the presence of high L-Phe concentrations and after preincubation of the enzyme with the substrate. It also occurred at low micromolar concentrations of 7(S)BH4 and physiological concentrations of 6(R)BH4 and L-Phe (Fig. 3F
). Moreover, ITC and Trp fluorescence spectroscopy analysis corroborated that PAH binds the cofactor in the affinity order 6(R)BH4>7(S)BH4>7(R)BH4. Taken together, our results indicate that the inhibition of PAH by 7(R,S)BH4 is due to the effect of 7(S)BH4 on enzyme activity and not a combined effect of both diastereomers.
To get further insights into the interactions between 6(R)BH4 and the 7-BH4 diastereomers with PAH at the molecular level, we generated MD-simulated structures of their complexes and analyzed their putative interactions with areas of high affinity for polar and hydrophobic groups (Figs. 4
and 5A
). The modeled structures complement the experimental results vis-à-vis binding affinities and efficiencies as cofactors. The perfect 
-stacking interaction between Phe254 in PAH and the pterin ring of 6(R)BH4 is distorted in the complexes with the 7-BH4 diastereomers, most probably contributing to a reduction in their binding affinities and favorable enthalpic contributions to binding and positioning these pterin cofactors in a suboptimal conformation for an efficient coupled catalysis. Moreover, the slightly longer distances between the catalytic iron and the C4 position of the pterin ring in the 7-BH4 diastereomers, notably in the PAH·7(R)BH4 complex, may also add to the observed uncoupling and their lower catalytic efficiency as cofactors.
Our results on the effect of L-Phe on the specific inhibition of human PAH by 7(S)BH4 (Fig. 3F
) agree with results from rat PAH with 7(R,S)BH4 (13)
. The quantitative differences in the degree of inhibition between both studies might be explained by the different activations determined for L-Phe in the human and rat enzymes. Thus, the L-Phe-induced activating conformational change of PAH results in a 20–30-fold increase of the BH4-dependent activity of the rat enzyme but only a 3- to 4-fold activation for the human enzyme, which shows a higher level of basal activity prior to activation (33
, 46
, 47)
. This difference is probably related to the physiological responsiveness to blood concentration of L-Phe in the different species. The apparent affinity for L-Phe (S0.5) and the Hill coefficient for cooperativity (h=2) are nevertheless similar for rat and human PAH. We found that the activating conformational change increases the affinity of the enzyme for L-Phe, while it decreases the affinity for 6(R)BH4 [4-fold decrease in the Kd for 6(R)BH4 binding by L-Phe incubation], without affecting the affinity for 7(S)BH4. L-Phe thus appears to have the important effect of increasing the specific inhibition of PAH by 7(S)BH4 (25)
. We may infer that the increased activation of rat PAH by L-Phe may be related to a greater inhibition by L-Phe for this enzyme compared with the human enzyme, thus explaining the quantitative differences in the effect of L-Phe on PAH inhibition by 7- BH4 observed in our study and in previous results (13)
.
In summary, our results provide a kinetic, thermodynamic, and structural framework to explain the role of 7(S)BH4 accumulation in primapterinuria and acute vitiligo by inhibiting PAH at physiological concentrations of substrate and 6(R)BH4.
Differential effect of 7(R,S)BH4 on PAH, TH and TPH activities
One interesting feature of deficient PCD activity and accumulation of 7(R,S)BH4 as documented in primapterinuria and acute vitiligo is that there is no evidence for impaired TH and TPH activities in these patients (5)
. It is noteworthy that the affinities of TH, TPH1, and TPH2 for 7(R,S)BH4 as cofactor and competitive inhibitor vs. 6(R)BH4 are 1000 times lower than for PAH (13
, 27)
. Our results showed that both 7(R)- and 7(S)BH4 function as cofactors and inhibitors of recombinant human TH1. However, the enzyme displays lower catalytic efficiency using both diastereomers as cofactors, and by contrast to PAH, TH is not significantly inhibited with either of the diastereomers. The molecular basis for the unique specific inhibitory interaction between 7(S)BH4 and PAH appears to be related to concerted interactions between the dihydroxypropyl side chain at the C7 position of the pterin ring and key residues at the pterin binding site and the N-terminal autoregulatory sequence. The importance of a conserved Ser at the PAH active site (Ser-251, which is replaced by Ala in TH1(Ala297) and Pro in TPH1(Pro238), has been substantiated by the[PAH.7(S)BH4] structure (Fig. 4)
as well as by molecular interaction field analysis (27)
and Fig. 5
. Thus, the area of high affinity for the hydroxyl probe close to both of the hydroxyls on the side chain at C7 of 7(S)BH4 in PAH (Fig. 5A
) is absent in TH (Fig. 5B
) and TPH (27
; Fig. 5A, B
). Nevertheless, it is puzzling that TH and TPH activities are neither affected by accumulated 7(R,S)BH4 nor by the reduced 6(R)BH4 concentrations inferred from suboptimal plasma biopterin levels found in patients with primapterinuria (5)
. Moreover, PCD activity is essential to keep optimal 6(R)BH4 levels in the central nervous system (CNS) where both TH and TPH activities have been found (48)
, notably since 6(R)BH4 cannot efficiently cross the blood-brain barrier (49)
. A possible explanation for the lack of altered TPH and TH activities in primapterinuria patients could arise from the recent discovery of a new PCD isoform, namely DCoH. This enzyme is encoded by a different gene (50)
, but it displays similar catalytic properties to PCD (51)
. Complementation of genetically defective PCD activity by high expression of DCoH in tissues where TH and TPH activities are found could support efficient 6(R)BH4 recycling.
Physiological implications of 7-BH4 diastereomers
The question remains whether 7(R,S)BH4 has in addition to inhibition of PAH a genuine physiological function at the low concentrations that are always produced from the aromatic amino acid hydroxylases in healthy cells and tissues (45)
. Only recently it has been shown that 7(R,S)BH4 is formed in the pigment-forming cell, the melanocyte and in particular in the melanosomes, which are specific organelles of these cells. 7(R,S)BH4 can control melanogenesis via ß-melanocyte-stimulating hormone (ß-MSH) in a receptor independent manner, whereas -MSH does not bind 7(R,S)BH4 (52)
. This observation implicated 7(R,S)BH4 in the direct control of tyrosinase activity in human skin and hair under physiological conditions (52)
. The availability of the two 7-diastereomers opens the possibility to further investigate the molecular basis for the regulation of tyrosinase activity.
In summary, our findings implicate 7(S)BH4 as the causative diastereomer for inhibition of PAH in primapterinuria and in the epidermis of patients with vitiligo. This 7(S)BH4-induced inhibition of PAH causing hyperphenylalaninemia in patients with primapterinuria can be counteracted by increasing the concentration of 6(R)BH4 and decreasing the L-Phe concentration in the diet, overriding the inhibitory effect (13)
, whereas in vitiligo epidermal PCD activities are fully restored after removal of H2O2 leading to low 7(R,S)BH4 levels (10
, 11)
.
ACKNOWLEDGMENTS
The authors want to thank Randi M. Svebak and Ali Javier Sepulveda for expert technical assistance. This research has been supported by the Research Council of Norway (A. Martinez and K. Teigen), Stiefel International (K. U. Schallreuter), the University of Bradford/UK (K. U. Schallreuter, J. M. Wood, and D. J. Maitland), and private donations from patients with vitiligo.
Received for publication March 13, 2006. Accepted for publication May 22, 2006.
REFERENCES
), and 7(S)BH4 (
,
) were used as cofactors of L-Phe-preincubated (5 min preincubation, 25°C; closed symbols) or nonpreincubated (open symbols) PAH. In (A) 1 mM L-Phe and in (B) 800 µM 7-BH4 diastereomers was used. Lines are best-fits to the Michaelis-Menten equation (A) or the Hill equation (B) of the data for 7(R)BH4 (solid) and 7(S)BH4 (dashed). C and D) 6(R)BH4-dependent PAH activity in the absence (•) or presence of 5 µM of 7(R)BH4 (
) or 7(S)BH4 (
) with (closed symbols) and without (open symbols) preincubation at same concentrations; % activity is referred to as the activity in the absence of the 7-BH4 inhibitors. Data are mean ± SD of triplicate measurements.
), and 7(S)BH4 (
) 10 µM in A and 200 µM in B. A, inset) Quenching of Trp-fluorescence at 340 nm at increasing tetrahydropterin concentrations.