Two apparent molecular weight forms of human and monkey phenylalanine hydroxylase are due to phosphorylation.

Two-dimensional polyacrylamide gel analyses of purified human and monkey liver phenylalanine hydroxylase reveal that the enzyme consists of two different apparent molecular weight forms of polypeptide, designated H (Mr = 50,000) and L (Mr = 49,000), each containing three isoelectric forms. The two apparent molecular weight forms, H and L, represent the phosphorylated and dephosphorylated forms of phenylalanine hydroxylase, respectively. After incubation of purified human and monkey liver enzyme with purified cAMP-dependent protein kinase and [gamma-32P]ATP, only the H forms contained 32P. Treatment with alkaline phosphatase converted the phenylalanine hydroxylase H forms to the L forms. The L forms but not the H forms could be phosphorylated on nitrocellulose paper after electrophoretic transfer from two-dimensional gels. Phosphorylation and dephosphorylation of human liver phenylalanine hydroxylase is not accompanied by significant changes in tetrahydrobiopterin-dependent enzyme activity. Peptide mapping and acid hydrolysis confirm that the apparent molecular weight heterogeneity (and charge shift to a more acidic pI) in human and monkey liver enzyme results from phosphorylation of a single serine residue. However, phosphorylation by the catalytic subunit of cAMP-dependent protein kinase does not account for the multiple charge heterogeneity of human and monkey liver phenylalanine hydroxylase.

the H forms contained s2P. Treatment with alkaline phosphatase converted the phenylalanine hydroxylase H forms to the L forms. The L forms but not the H forms could be phosphorylated on nitrocellulose paper after electrophoretic transfer from two-dimensional gels. Phosphorylation and dephosphorylation of human liver phenylalanine hydroxylase is not accompanied by significant changes in tetrahydrobiopterin-dependent enzyme activity. Peptide mapping and acid hydrolysis confirm that the apparent molecular weight heterogeneity (and charge shift to a more acidic PI) in human and monkey liver enzyme results from phosphorylation of a single serine residue. However, phosphorylation by the catalytic subunit of CAMP-dependent protein kinase does not account for the multiple charge heterogeneity of human and monkey liver phenylalanine hydroxylase.
Mammalian liver phenylalanine hydroxylase (phenylalanine 4-monooxygenase, EC 1.14.16.1) catalyzes the rate-limiting step in the hepatic conversion of phenylalanine to tyrosine (for reviews see Refs. [1][2][3]. Purified phenylalanine hydroxylase from rat (4), human, and monkey (5) livers has two different molecular weight polypeptides, designated H (apparent M, = 50,000) and L (apparent M, = 49,000), as revealed by polyacrylamide gel electrophoresis in the presence of SDS.' In addition to the molecular weight heterogeneity, two-dimensional polyacrylamide gel electrophoresis of human and monkey liver enzyme demonstrates that each apparent molecular weight class consists of three major isoelectric forms (6, 7). We have investigated the multiple forms of phenylalanine hydroxylase polypeptides to determine whether the polypep-* This work was supported in part by grants from the Victorian Mental Health Fund and the National Health and Medical Research Council of Australia. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "aduertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
' The abbreviation used is: SDS, sodium dodecyl sulfate. tides derive from post-translational modification of a single gene or from different genes. Previously, the L form was considered to be a degradative product of the H form, since preparations of the human enzyme which were low in activity contained more of the L apparent molecular weight class of polypeptides (6). However, recent evidence from this laboratory has demonstrated that the two apparent molecular weight forms of rat liver phenylalanine hydroxylase are encoded by two different mRNAs and hence may be products of different genes (8).
Phenylalanine hydroxylase can be phosphorylated (9) and it has been suggested that differences in the extent of phosphorylation may be responsible for the chromatographic heterogeneity of the enzyme on calcium phosphate (10, 11). In this paper we examine the influence of phosphorylation and dephosphorylation on the two-dimensional polyacrylamide gel electrophoretic patterns of human and monkey phenylalanine hydroxylase and demonstrate that the higher apparent molecular weight forms are phosphorylated and that the lower apparent molecular weight forms can be converted to the higher apparent molecular weight form by phosphorylation, but this change is not accompanied by significant alteration of enzyme activity.

EXPERIMENTAL PROCEDURES
Protein Purification-The catalytic subunit of CAMP-dependent protein kinase, hereafter referred to as protein kinase, was isolated from bovine heart muscle by the procedure of Sudgen et al. (12).
Human and monkey liver phenylalanine hydroxylase were purified from crude liver extracts by affinity chromatography using a monoclonal antibody (PH1-1) immunoaffinity column (13). The recovery of human and monkey enzyme in three recent preparations for each was 41, 59, and 52% and 50, 63, and 62%, respectively. The yield of human enzyme varied from 0.53 to 0.90 mg/20 g of liver in 3 recent experiments. The purified enzymes were approximately 95% homogeneous based on densitometric scans of stained patterns obtained from analytical SDS-polyacrylamide gel electrophoresis.
Phosphorylation of Phenylalanine Hydroxylase-The phosphorylation of phenylalanine hydroxylase by the protein kinase was carried out at 30 "C in a reaction mixture (0.1 ml) containing 10 mM Tris-HCI, pH 7.6, 10% (v/v) glycerol, 0.1 mM EDTA, 10 mM magnesium acetate, 0.1% (v/v) 2-mercaptoethanol, 0.1 mM [y-32P]ATP (4000 cpm/pmol), and 16.5 pg of protein kinase. From the reaction mixture, 5-pl aliquots were withdrawn, placed on phosphocellulose ion exchange filter paper squares, and washed as described previously (14). jnP incorporation into protein was determined by liquid scintillation counting. The stoichiometry of the acid-stable phosphorylation was calculated using a subunit molecular weight of 50,000 for phenylalanine hydroxylase. When activity measurements were to be made after phosphorylation, 2-mercaptoethanol was omitted from the reaction mixture, 0.1 mM ATP replaced radioactive ATP, and protein kinase was reduced to 7.5 pg.
The products of the phosphorylation reactions were isolated either by precipitation with 7% (w/v) trichloroacetic acid at 4 "C or by immunoprecipitation, essentially as described previously (8).
Enzymatic Dephosphorylation of Phenylalanine Hydroxylase-Preparations of phenylalanine hydroxylase (10 pg) were incubated at 30 'C with 0.4 unit of calf intestinal alkaline phosphatase (Boehringer Mannheim, Grade I suspension), 1 mM lysolecithin, 0.3 mM MgCl2, 0.1 mM phenylalanine in the presence of 0.03 mM phenylmethylsulfonyl fluoride and 0.03 mM eaminocaproic acid. The control samples were incubated without alkaline phosphatase. After 30 min, the reaction was stopped in the presence of 10 mM phenylalanine, and aliquots (5 pg) were removed for two-dimensional gel analysis (see text below), and the remainder was purified from alkaline phosphatase using the PHI-1 immunoaffinity column (13). The phenylalanine hydroxylase was concentrated and phosphorylated to 30 "C by the protein kinase as described above, and the stoichiometry of acidstable radioactive incorporation of phosphate into purified phenylal anine hydroxylase was measured.
When activity measurements were to be made after dephosphorylation lysolecithin and phenylalanine were omitted.
Enzyme Assays-Phenylalanine hydroxylase activity, after phosphorylation in the presence of protein kinase as above but replacing radioactive phosphate with 100 p~ cold ATP or dephosphorylation in the presence of alkaline phosphatase as described above, was measured by following the conversion of ["Clphenylalanine to ["C] tyrosine at 25 "C as described (6).
Gel Electrophoresis-One-dimensional electrophoresis was as described @), but gels were run at constant current of 20 mA for 4 h.
Two-dimensional isoelectric focusing/SDS-polyacrylamide gel electrophoresis was performed as described (13). Isoelectric focusing was carried out at room temperature at a constant voltage (450 V) for 16 h followed by a further 800 V for 1 h. Second dimension SDS gels were 9% (w/v) polyacrylamide containing 0.1% (w/v) SDS (15). Electrophoresis was at a constant current of 25 mA for 5-6 h in 0.1% (w/v) SDS running buffer (16). All gel patterns are shown with the basic end of the isoelectric focusing dimension at the left and the lower molecular weight region of the second dimension at the bottom. For the location of 32P-polypeptides, gels were autoradiographed after staining with 0.05% (w/v) Coomassie Brilliant Blue in metha-no1:water:acetic acid, 5 5 1 (v/v), destaining in water:ethanol:acetic acid, 8l:l (v/v), and vacuum drying. The dried gels were then exposed to Kodak XAR-5 film using a DuPont Lightning Plus intensifying screen at -70 "C for autoradiography.
The pH gradient in the isoelectric focusing dimension was determined by removing the gels from their tubes, freezing them, and cutting 1-mm slices. The pH of the gel slices was measured after the slices were placed in 2 ml of deionized degassed water and allowed to equilibrate for 2 h.
Transfer of Protein to Nitrocellulose and Phosphorylation-Phenylalanine hydroxylase was transferred from two-dimensional SDS slab gels onto nitrocellulose paper according to the method of Towbin et al. (17). Electrophoresis was carried out in 25 mM Tris, 192 mM glycine, and 20% (v/v) methanol for 2 hat 4 "C. After protein transfer, the appropriate section of nitrocellulose paper containing the phenylalanine hydroxylase polypeptides was cut out and incubated overnight at room temperature in 2 ml of 10 mM Tris-HCl, pH 7.4, 0.15 M NaC1, and 3% (w/v) serum albumin.
The phosphorylation of phenylalanine hydroxylase polypeptides on nitrocellulose paper by the protein kinase was carried out under the same conditions as described above. After 30 min of incubation, the nitrocellulose papers were washed 3 times and stored overnight in 75 mM phosphoric acid. The washed nitrocellulose papers were then air-dried and autoradiographed with Kodak XAR-5 film using a DuPont Lightning Plus intensifying screen. The upper and lower apparent molecular weight forms of the phenylalanine hydroxylase were located using rabbit anti-rat liver phenylalanine hydroxylase serum and lSI-protein A after decay of 32P as described below. The nitrocellulose papers were incubated overnight at room temperature in 3 ml of 10 mM Tris-HCI, pH 7.4, 0.15 M NaCI, 3% (w/v) serum albumin containing 25 pl of rabbit anti-rat liver phenylalanine hydroxylase serum. The papers were then washed at room temperature once with 50 ml of 10 mM Tris-HC1, pH 7.4,0.15 M NaCI, twice with 50 ml of 10 mM Tris-HCl, pH 7.4, 0.15 M NaCl, 0,2% (v/v) Nonidet P-40 on a gyratory shaker, and once with 50 ml of 10 mM Tris-HCI, pH 7.4, 0.15 M NaCI. The washed nitrocellulose papers were then incubated for 60 min at room temperature with 3 ml of 10 mM Tris-HCl, pH 7.4, 0.15 M NaCl, 3% (w/v) serum albumin containing 0.5 pCi/ml (3.7 X lo' cpm/ng) of 1251-protein A (Amersham International). The nitrocellulose papers were washed as described above, air dried, and autoradiographed.

RESULTS
Two-dimensional Gel Pattern of Human and Monkey Phenylalanine Hydroxylase-The two-dimensional gel pattern of both purified human and monkey liver phenylalanine hydroxylase exhibited two apparent molecular weight forms, each of which contained three isoelectric forms (Fig. 1, A and B)?
The most basic isoelectric species were predominant. This pattern is essentially the same as those previously described in this laboratory (6.7). The relative distribution of the charge forms is similar in both mammalian enzymes (see Table I).
Phosphorylation of Phenylalanine Hydroxylase by the Protein Kinase-When the purified human and monkey liver enzyme were incubated with [ Y -~~P ] A T P and M e in the presence of the protein kinase the 32P-polypeptides co-migrated with the H form polypeptides (apparent M , = 50,000). Each of the charge species of the H form was labeled with 32P ( Fig. 1, C and D). There    hydroxylase was phosphorylated (Fig. 3A). No 32P was associated with the H form polypeptides (apparent M, = 50,000).
The location of the L and H forms of the enzyme was determined by probing the nitrocellulose filters with rabbit anti-phenylalanine hydroxylase serum (6) and IBI-protein A (Fig. 3, C and D)  These results indicate that only the L form (apparent M, = 49,000) and not the H form (Mr = 50,000) of phenylalanine hydroxylase is the substrate for the protein kinase.
Stoichiometry of Phosphorylutwn-The interconversion of the L and H forms of phenylalanine hydroxylase was studied further by examining the stoichiometry of the phosphorylation reaction. Densitometric scans of gels indicated that purified human and monkey liver phenylalanine hydroxylase contained approximately 70 and 50%, respectively, in the H form (Fig. lA). The protein kinase catalyzed the incorporation of 0.2-0.3 mol of [32P]phosphate per mol of phenylalanine hydroxylase (Fig. 4). Following alkaline phosphatase treatment, the immunopurified phenylalanine hydroxylase was phosphorylated to a greater extent by the protein kinase (0.67 and 0.91 mol of 32P per mol of phenylalanine hydroxylase monomer for human and monkey, respectively (Fig. 4)). The lower level of incorporation of 32P into the human enzyme is consistent with the incomplete conversion of the H form to the L form prior to phosphorylation of partially phosphorylated human liver enzyme (see Fig. 2).
Tryptic peptide maps of the [32P]phenylalanine hydroxylase from human and monkey liver revealed a single 32P-phosphopeptide for both enzymes (Fig. 5). ["P]Phosphoserine was the only phosphoamino acid detected following partial acid hydrolysis (data not shown).
The results obtained support the conclusion that the protion was monitored by the withdrawal of aliquots (5 pl of reaction mixture containing 100 pg/ml of protein) a t intervals, and the 32P incorporated was determined as described under "Experimental Procedures." Activity of Human Liver Phenylalanine Hydroxylase-It was of interest to test the influence of phosphorylation and human phenylalanine hydroxylase activity since studies have shown that the rat counterpart is activated by phosphorylation (9)(10)(11). We have found no reproducible stimulation of human phenylalanine hydroxylase activity following phosphorylation when tested over a range of tetrahydrobiopterin cofactor concentrations (0.1-50 p~) .
There was a small decrease in phenylalanine hydroxylase activity (26%) with removal of phosphate by alkaline phosphatase treatment.

DISCUSSION
The results presented here clearly demonstrate that the L form (apparent M, = 49,000) and H form (apparent M, = 50,000) of human and monkey phenylalanine hydroxylase represent the dephosphorylated and phosphorylated forms of the enzyme, respectively. The H and L forms can be interconverted with CAMP-dependent protein kinase and alkaline phosphatase. The difference in SDS gel mobility leading to the apparent shift in molecular weight results from the incorporation of a single phosphate. The molecular basis of this effect is not known although it has been observed in several systems by others (18-20). The effect of phosphorylation on mobility in SDS is opposite to that expected for the simple addition of two negative charges. One interpretation of the results is that phosphorylation leads to a reduction in the amount of SDS bound by the enzyme and thereby reduces its electrophoretic mobility. In addition to the apparent molecular weight change, the phosphorylation of the phenylalanine hydroxylase was also associated with a charge shift to a more acidic PI on two-dimensional gels. Similar changes in the PI following phosphorylation have been observed by others (16, 18, 21, 22). Phosphorylation of human and monkey phenylalanine hydroxylase on a single serine residue by the CAMPdependent protein kinase is analogous to the findings for rat liver phenylalanine hydroxylase (9,23).
The mechanism generating the two apparent molecular weight forms in primate phenylalanine hydroxylase should be contrasted with that found with the rat enzyme. All three enzyme preparations show two apparent molecular weight forms of which the lower molecular weight forms were previously assumed to be due to proteolysis (6). In this paper we demonstrate that the two molecular weight forms of human and monkey phenylalanine hydroxylase result from protein phosphorylation. In the rat, however, we have previously demonstrated that the two molecular weight forms are iso-enzymes encoded by different mRNAs (8). More recent analysis has revealed that the two proteins of different apparent molecular weight are coded by allelic genes (24). Preliminary results indicate that both of these forms of the rat phenylalanine hydroxylase can be pho~phorylated.~ Our finding that phosphorylation of human phenylalanine hydroxylase activity was not altered by protein phosphorylation was surprising in the light of the results obtained with the rat enzyme (8)(9)(10)(11). This negative result, however, does not exclude the possibility that under different assay conditions or in vivo, phosphorylation may have an influence.
In contrast to the results obtained here, Abita et al. (25) reported that purified human phenylalanine hydroxylase was not phosphorylated by the CAMP-dependent protein kinase. There are major differences in the purification procedures used by Abita et al. (25) and those used here. However, it is not clear why the form isolated by these workers could not be phosphorylated. One possibility is that there are allelic forms (24) that differ in their capacity to act as substrates. Preliminary amino acid analysis of the preparations used in the study indicate that the level of methionine is lower than the 10.8 residues reported by Abita et al. (25). Furthermore, treatment of the human phenylalanine hydroxylase with cyanogen bromide resulted in only 4 peptides as indicated by SDS-gel electrophoresis. ' The approach used in this study of testing the capacity of phenylalanine hydroxylase to act as a substrate for the protein kinase following transfer from a denaturing slab gel to nitrocellulose paper was unexpectedly successful. Although immunological identification of transferred proteins is regularly used, the use of a protein kinase as a probe to screen nitrocellulose bound protein has not been reported previously. We expect this approach will be applicable to other proteins but should be cautiously applied since the method may mask phosphorylation sites or alternatively expose sites that are not normally phosphorylated. In the experiments reported here no new sites were exposed on the enzyme bound to the nitrocellulose; otherwise both molecular weight forms of the enzyme would have been phosphorylated.
Our results demonstrate that the multiple charged forms of both apparent molecular weight forms (H and L) do not result from charge heterogeneity following phosphorylation by the CAMP-dependent protein kinase. The molecular basis of the different charged forms of human and primate phenylalanine hydroxylase remains to be determined.