Phenylalanine Hydroxylase Stimulator Protein Is a 4a-Carbinolamine Dehydratase*

Phenylalanine hydroxylase stimulator protein is shown to be an enzyme that catalyzes the dehydration of the 4a-carbinolamine, derived from the turnover of 6-methyltetrahydropterin with phenylalanine hydroxylase, to form p-quinonoid-6-methyldihydropterin. This conclusion is based on the facts that: 1) a linear dependence on phenylalanine hydroxylase stimulator concentration of the first order rate of decay at low levels of 4a-carbinolamine is observed; 2) the physical separation of phenylalanine hydroxylase from phenylalanine hydroxylase stimulator and 4a-carbinola- mine gives identical kinetic parameters with those obtained in its presence, thus indicating no catalytically significant protein-protein interaction; and 3) the 4a- carbinolamine analog, 5-deaza-4a-hydroxy-6-methyl-tetrahydropterin is an inhibitor of phenylalanine hydroxylase stimulator activity. The structure of the tetrahydropterin-derived intermediate formed by phenylalanine hydroxylase (EC 1.14.16.1) during catalysis has recently been shown to be the 4a-hydroxy adduct based on 13C NMR and UV data (1, 2). The 4a-CA' can dehydrate nonenzymatically to form the p-quinonoid dihydropterin (3), which is recycled to the tetrahydropterin by NADH and dihydropteridine reductase In some experiments physical separation of phenylalanine hydrox- ylase from the reaction mixture was accomplished by filtration via a 1-ml syringe through a Millex-PF (0.8 pM) cellulose ester membrane (Millipore). No phenylalanine hydroxylase activity is observed after filtration. At time zero, 11.8 p~ 6-MPH4 was added to 0.1 M Tris- HC1, pH 8.45, 1 mM L-phenylalanine and 57 pg of phenylalanine hydroxylase in 1-ml assay. After 50 s the mixture was filtered and at 80 s, 3 pg of PHS/ml were added and A z ~ ~ monitored.

f To whom correspondence should be addressed.
Phenylalanine hydroxylase stimulator protein, which has been purified and studied (4)(5)(6), was initially thought to influence the association-dissociation of phenylalanine hydroxylase (4). Subsequently, however, it was shown that PHS catalyzes the conversion to quinonoid dihydropterin of a pterin intermediate in the phenylalanine hydroxylase reaction ( 7 ) . The pterin intermediate was postulated to be the 4acarbinolamine derived from the corresponding tetrahydropterin ( 7 ) , a postulate that was later shown to be correct (1,2). The purpose of this paper is to present unequivocal evidence that PHS is a 4a-carbinolamine dehydratase.
Methods-UV spectra were obtained on either a Cary 118 or Cary 219 spectrophotometer. Phenylalanine hydroxylase was assayed by the method of Shiman et al. (8). PHS was assayed by measuring the increase in the first order rate of decay of 4a-CA to p-quinonoid-6-MPH2 at 244 nm. Assay conditions were as follows. A cuvette con- Inhibition studies using 5-deaza-4a-hydroxy-6-MPH4 were carried out by adding equivalent amounts to blank and sample cuvettes containing 0.1 M Tris-HC1, pH 8.45, 1 mM L-phenylalanine, and 72 pg of phenylalanine hydroxylase, both with and without 50 pg of catalase, at time zero, followed by addition of 12 p~ 6-MPH4 at t = 90 s and 3 pg of PHS at t = 130 s. The amount of inhibitor present at t = 150 s, where kobs was determined, was calculated to be 73% of the initial concentration due to the instability of 5-deaza-4a-hydroxy-6-MPH4 ( k = 0.0021 s-l) under these conditions. The amount of 4a-CA at 150 s was estimated to be between 1.7 and 5.5 p M depending upon inhibitor concentration. The product of the decomposition of 5-deaza-4a-hydroxy-6-MPH4 (10) does not inhibit PHs. 90% trans; 10% cis by 'H NMR.
The specific activity of phenylalanine hydroxylase in 0.1 M Tris, pH 8.45, is two-thirds that in 0.1 M KPi, pH 6.8 (8).

P H s Is a 4a-Carbinolamine Dehydratase
In some experiments physical separation of phenylalanine hydroxylase from the reaction mixture was accomplished by filtration via a 1-ml syringe through a Millex-PF (0.8 p M ) cellulose ester membrane (Millipore). No phenylalanine hydroxylase activity is observed after filtration. At time zero, 11.8 p~ 6-MPH4 was added to 0.1 M Tris-HC1, pH 8.45, 1 mM L-phenylalanine and 57 pg of phenylalanine hydroxylase in 1-ml assay. After 50 s the mixture was filtered and at 80 s, 3 pg of PHS/ml were added and A z~~ monitored.

RESULTS
The effect of P H S is most evident in the reaction progress curve of 6-MPH4 with phenylalanine hydroxylase at pH 8.45, where no intermediate is observed in its presence (Fig. lC). 4 If P H S is added after the 4a-CA is generated by phenylalanine hydroxylase (see "Methods") an increase in the first order rate of decay (k&) of 4a-CA is observed (Fig. 1B). These results with the 4a-CA derived from 6-MPH4 are in agreement with those previously reported on the effect of PHS on the 4a-CA derived from tetrahydrobiopterin (7). Assuming that the 4a-CA concentration is below its K,,, for PHS and 6-MPH4 2 0,6 which should be valid late in the reaction, then k,& = ko + kpHs where k, is the first order nonenzymatic rate of decay (= 0.010 s") and kpHs = (kc,,[PHS]/K,,,) for P H s . Catalase has no effect on kobs.
As can be seen from Fig. 2 there is a strict dependence of kpHs on PHS concentration. That PHS is acting as an enzyme to catalyze dehydration of 4a-CA and not in conjunction with phenylalanine hydroxylase is clearly demonstrated by the fact that the physical separation of phenylalanine hydroxylase from 4a-CA by filtration through a Millex-PF membrane and subsequent addition of PHS (3 pg/ml) gives an identical rate (kohs = 0.56 s-') with that observed in the presence of phenylalanine hydroxylase. In addition there is only a very slight effect (-7%) on k, upon filtration of phenylalanine hydroxylase from the reaction mixture.
Further evidence consistent with the 4a-carbinolamine dehydratase activity of P H S is derived from the observation that 5-deaza-4a-hydroxy-6-MPH4 is an inhibitor of the PHS reaction. Assuming that 4a-CA is below the K, for PHS (see  "Methods") and that inhibition of the deaza analog is competitive against 4a-CA, then: Therefore a plot of l/kPHS versus I should be linear with KI = intercept/slope. The data in Fig. 3 give a value of KI = 43 p~. No inhibition is observed by 5-deaza-4a-hydroxy-6-MPH4 (100 p~) when assayed with phenylalanine hydroxylase and 6-MPH, (12 p~) either at pH 6.8 or pH 8.45. Also, there is no inhibition by 6-phenyl-5-thiatetrahydropterin (100 PM), a potent competitive inhibitor of phenylalanine hydroxylase (K, = 0.2 p~ (ll)), when assayed with PHs and 6-MPH4 (12 p~) .

DISCUSSION
The evidence that PHS is a 4a-carbinolamine dehydratase is based on the facts that: 1) a linear dependence on PHS concentration of the first order rate of decay at low concentrations of 4a-CA is observed; 2) there is no catalytically significant protein-protein interaction between phenylalanine hydroxylase and PHs; and 3) the 5-deaza-4a-hydroxy analog is an inhibitor of PHS activity. It has now been demonstrated that both tetrahydrobiopterin (7) and 6-MPH4 (1) form a 4acarbinolamine during phenylalanine hydroxylase turnover which is more stable a t higher pH values. The relative stability of 4a-CA at higher pH values is not surprising since carbinolamine dehydrations are acid-catalyzed reactions (13). Since the 4a-CA derived from both tetrahydropterins serves as a substrate for PHs, the side chain at C-6 is not involved in the chemical mechanism of the dehydration.
It is interesting to note that inhibitors of phenylalanine hydroxylase are not detectible inhibitors for PHS and vice versa even though the product of phenylalanine hydroxylase is the substrate for PHs. This implies that the equilibrium phenylalanine hydroxylase. 4a-CA + phenylalanine hydroxylase + 4A-CA favors dissociation and that 4a-CA is not a strong product inhibitor. Consequently, PHS stimulation of phenylalanine hydroxylase turnover should be viewed as maintenance at higher steady state levels of the tetrahydropterin pool during the recycling assay (6).