Proteolytic Modification of the Amino-terminal and Carboxyl-terminal

THE JOURNAL OF BIOLOGICAL CHEMISTRY Vol. 261, No. 5, Issue of February 15, pp. 2051-2056,1986 Printed in U.S.A.

Proteolytic Modification of the Amino-terminal and Carboxyl-terminal Regions of Rat Hepatic Phenylalanine Hydroxylase*

(Received for publication, June 7,1985)

Masayoshi Iwaki$, Robert S. Phillips& and Seymour Kaufman From the Laboratory oi Neurochemistry, National Institute of Mental Health, Bethesda, Maryland 20205

Activation of rat liver phenylalanine hydroxylase by limited proteolysis catalyzed by chymotrypsin was investigated with the use of sodium dodecyl sulfate- polyacrylamide gel electrophoresis and high pressure gel filtration. Both activation and proteolysis were decreased by the addition of the natural cofactor, (6R)- tetrahydrobiopterin. From chymotryptic digests of the hydroxylase carried out in the presence and absence of (6R)-tetrahydrobiopterin, several different enzyme species were isolated by high pressure gel filtration. One species (subunit M, = 47,000) with unchanged hydroxylase activity was isolated from the chymotryptic digest in the presence of (6R)-tetrahydrobiopterin; it was derived from the native enzyme (Mr = 52,000) by cleavage of the COOH-terminaI M, = 5,000 portion of the native enzyme. In the absence of (6R)-tetrahydrobiopterin, another species (subunit M, = 36,000) was isolated. In addition to modification at the COOH- terminal end of the molecule, this species also had lost a M, = 11,000 fragment from the NH2-terminal end of the hydroxyIase. The M, = 11,000 fragment was shown to include the phosphorylation site of the enzyme. This M, = 36,000 species was 30-fold more active than the native phenylalanine hydroxylase when assayed in the presence of tetrahydrobiopterin. These results suggest that the regulatory domain that inhibits hydroxylase activity in the basal state may be located at the NH, terminus of the phenylalanine hydroxylase subunit.

The activity of phenylalanine 4-hydroxylase (EC 1.14.16.1) from rat liver can be regulated by its substrate, phenylalanine, and cofactor, tetrahydrobiopterin, as well as by various other effectors. Preincubation of the enzyme with L-phenylalanine results in a 20-30-fold increase of the initial activity, which is reversible upon removal of the substrate (1,2). The enzyme can also be activated by phospholipids such as lysolecithin (3), by sulfhydryl modification with N-ethylmaleimide (4), and by phosphorylation both in vitro and in vivo (5-7). The full increase in hydroxylase activity resulting from these modifications can be observed only when the enzyme activity is measured in the presence of tetrahydrobiopterin (BH,)l.

* The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore he hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

$Present address: Dept. of Biochemistry, Shiga University of Medical Science, Ohtsu, Shiga 520-21, Japan.

3 Present address: Dept. of Chemistry, University of Georgia, Ath- ens, GA 30602.

The abbreviations used are: BH,, tetrahydrobiopterin; GMPH,, D~-6-methyl-5,6,7,8-tetrahydropterin; SDS, sodium dodecyl sulfate; PAGE, polyacrylamide gel electrophoresis; Pipes, 1,4-piperazineeth- anesulfonic acid HPGF, high pressure gel filtration,

l___-__”

Activation of the pure hydroxylase by phosphorylation, e.g. increases the BH4-dependent activity by 3-4-fold (5-8), whereas it increases the 6MPH4-dependent activity by only

In additions to activations of this type, phenylalanine hydroxylase can also be activated by limited proteolysis with chymotrypsin (3). Recently, we reported the effects of phenylalanine and BH, on the rate of limited cleavage of the hydroxylase (9). We found that BH4 inhibited the chymotryptic activation of phenylalanine hydroxylase, whereas phenylalanine stimulated the activation, although the effect of phenylalanine was not a marked one. From these results, we postulated that binding of BH, might drive the enzyme molecule into a deactivated form with a compact or globular conformation that is resistant to proteolysis, whereas phenylalanine might act in the opposite manner. We also suggested that limited proteolysis could be a useful technique in studying changes in the conformational state that might be associated with activation of the enzyme.

In the present study, we have investigated the proteolytic activation of the enzyme in greater detail in order to get information that might be useful in correlating changes in the structure of phenylalanine hydroxylase with the regulation of its activity.

10-15% (8).

EXPERIMENTAL PROCEDURES

Materials-Phenylalanine hydroxylase was isolated and purified from the livers of male Sprague-Dawley rats as previously described (IO) with the use of phenyl-Sepharose CL-GB (Pharmacia) chromatography (11). a-Chymotrypsin was purchased from Worthington Biochemical. Phenylmethylsulfonyl fluoride and Dt-6-methyl-5,6,7,8- tetrahydropterin (6MPH4) were from Calbiochem-Behring. (6R,S)- 5,6,7,8-Tetrahydrobiopterin (BH,) was obtained from Dr. Schirks’ laboratory (Switzerland) and resolved into (6R) and (65’) diastereo- isomers according to the method of Bailey and Ayling (12).

Methods-SDS-polyacrylamide gel electrophoresis (SDS-PAGE) was performed according to the method of Weber and Osborn (13) with the use of 10% acrylamide (1-mm-thick) slab gels. Electropho- resis was carried out with a constant current of 40 mA/cm2 for 6 h. Details of the cyanogen bromide cleavage of proteins (14) and Edman degradation for NHP-terminal analysis of peptides (15) were described in the previous paper (10).

Phenylalanine hydroxylase activity was measured with a continu- ous assay in which the rate of NADH oxidation in the presence of either BH, or 6MPH4 was monitored as previously described (4). 32P- phosphorylated phenylalanine hydroxylase was prepared according to the method of Phillips and Kaufman (16).

Limited Proteolysis by Chymotrypsin-Proteolysis by chymotrypsin was carried out essentially according to the method previously described (9). The typical incubation mixture contained, per ml, 50 pmol of Pipes (pH 6.8), 150 pmol of KCI, 500 pg of purified phenylalanine hydroxylase, 10 pg of catalase, 1 pmol of dithiothreitol, 25 pg of a-chymotrypsin, and, when present, 2-10 nmol of (6R)-BH,. For the assay of hydroxylase activity, aliquots were removed from the mixture after determined times of incubation and immediately added to the assay mixture. For SDS-PAGE, aliquots were treated with 25 pg/ml phenylmethylsulfonyl fluoride and then frozen, lyophilized,

2051

2052 Proteolytic Modification of Phenylalanine Hydroxylase

and subjected to electrophoresis. When preparative scale proteolysis the products A ( M , = 47,000), B ( M , = 40,000), and C ( M , = (1-25 mg of Phenylalanine hydroxylase) was carried out, the chyme- 36,000) were present in approximately equal amounts. After tryptic digest was treated with 25 ride, cooled, and concentrated with a Minicon B-15 (Amicon Corp.) or a Diaflo cell equipped with a PM-10 membrane (Amicon Corp.) to bands and A had disappeared and and were the a concentration of approximately 10 mg/ml. two major bands. Further digestion increased the ratios of

atingconditions was carried out essentially according to the procedure In the presence of (6R)-BH,, the rate of chymotr@ic reported by Horiike et al. (17) with the use of a Gilson Model 302 activation was decreased ( ~ i ~ . 1). The addition of 2 pM (6R)- high pressure liquid chromatography system with a 1 l l B ultraviolet BH, inhibited the initial rate of chymotryptic activation about detector (280 nm). The column was a TSK G 3000 SW (7.5 X 600 mm, Altex) and the eluant was 0.1 M potassium phosphate buffer, pH go%; approximately PM (6R)-BH4 was required for maxi- 6.8. Protein samples (1-10 mg/ml) were applied in a volume of 20- mum protection. It should be noted that the amount of 150 pl a t a flow rate of 1 ml/min at 25 "C. The column was calibrated hydroxylase used in this experiment was also about 10 pM. with the following standard proteins as previously described (10): The proteolytic pattern in the presence of 10 p~ (6R)-BH, is catalase (& = 232,000), aldolase ( M , = 158,000), lactate dehydrogen- shown in Fig. 2 ~ . Partial inhibition of proteo~ysis is evident. ase (MI = 14O,OOO), bovine Serum albumin (Aft = 66,300), ovalbumin As proteo~ysis proceeded, product A ( M , = 47,000) accumu-

phenylmethylsulfonyl flue- 30 min, when the activity had reached a maximum,

High Pressure Gel Filtration (HPGF)-HPGF under nondenatur- bands C and D (Mr = 29,500).

( M , = 43,000), and cytochrome c ( M , = 12,400). HPGF in the presence of6 M guanidine hydrochloride was carried out as previously described kited with very little h k a t i o n of the formation of degrah- (10). tion products (B, C, and D). After 1 h, when hydroxylase

activity had increased approximately 3-fold (from a specific

(N) had almost disappeared and the fragment A was the only significant band detected on electrophoresis.

Isolation of Chymotrypsin-modified Phenylulanine Hydrox- yhe-The patterns of the SDS-PAGE and the estimated

digestion in the presence and absence of (611)-BH, yields the

RESULTS activity of 0.07 to a specific activity of 0.21), the native subunit

Chymotrypsin Digestion of Phenylalanine Hydroxylase with and without (6R)-Tetrahydrobiopterin-As reported previously (9), there is evidence that in the presence of the natural cofactor, (6R)-BH,, the hydroxylase molecule is in a "tight"

investigate the effects of (6R)-BH4 on hydroxylase activity or proteolysis-resistant conformation* It was of interest to molecular weights ofthe products indicated that chymotryptic

and on the size of the hydroxylase molecule during limited Same proteo1yzed products, although the amounts and time proteolytic degradation. ~ i ~ ~ . 1 and 2 show the time courses course of appearance and disappearance are different. Because of the chymotryptic activation in the presence and absence of more than One product was found to be generated by Vote- (6R)-BH4, followed by measurement of hydroxylase activity olysis, it was necessary to isolate each one to investigate its and by SDS-PAGE. properties and structure. Fig. 3 shows the HPGF patterns of

native phenylalanine hydroxylase and the c h y m o t ~ ~ t i c di-

creased the hydroxylase activity rapidly until the maximum performed under nondenaturing conditions with the use of a value (33-fold increase over the initial activity) was reached TSK after 30 min ( ~ i ~ . 1). upon SDS-PAGE, followed by protein The native enzyme showed two peaks on gel filtration which

proteolyzed products as indicated by in ~ i ~ . 2 ~ . ~ f t e ~ were considered to be the tetrameric and dimeric forms of the a 15-min digestion, the native subunit (N, M, 52,000) and subunit ( M r = 52,000), respectively. It should be noted that

more than 70% of the native enzyme applied to the gel was recovered as a tetrameric species.

after a 60-min incubation in the presence of 10 p~ (6R)-BH,. Three peaks with hydroxylase activity were detected. The major peak indicated by a bar ( A ) was collected. From the calibration curve with standardproteins, the molecular weight of this peak was calculated to be 92,000. Fraction A showed a single band at M, = 47,200 on SDS-PAGE (Fig. 4, slot A ) indicating that it corresponded to the band A shown on SDS- PAGE of the chymotryptic digest in Fig. 2A. These results indicated that product A existed as a dimer of the chymotrypsin-modified subunits, in sharp contrast to the structure of the native enzyme whose major oligomeric form is a tetramer when estimated by gel filtration under the same conditions. No tetramerization of product A could be detected by rechro-

TIME (mid mg/ml to 8 mg/ml). The results also showed that no signifi-

droxylase by a-chymotrypsin with and. without (6R)-B%. FIG. 1. Time course of the activation of phenylalanine hy- cant amount of a-chymotrypsin Or the cleaved small pep-

Phenylalanine hydroxylase (o.5 mg/ml) was incubated with a-chy- tide(s) remained associated with the product. The overall yield

of 10 PM (&"€I), 5 PM (W), and 2 p~ ( O " - O ) (6R)-BH4 as was 65% (w/w, X 100). described under "Methods." At the indicated times, aliquots (20 pl) The elution profile of the digest without (6R)-BH4 is shown were removed for activity measurements. The activity was assayed in Fig. 3C. The elution pattern was rather complicated but essentially according to Ref. 4 as described under "Methods." VelOC- the hydroxylase activity was associated with the first two ities were measured by monitoring the rate of NADH oxidation. Tetrahydrobiopterin (40 p ~ ) and phenylalanine (0.1 mM) were used minor peaks and the next major peak* The major fraction* in 1.0 ml of total assay volume. The activity of the untreated enzyme indicated bY a bar (c)7 Was collected and found to account for was 0.07 pmol/min/mg of protein. 19% (w/w, x 100) of the total protein. The molecular weight

ln the absence of (~R)-BH,, chymotryptic digestion chymotrypsin/phenylalanine hydroxylase, 1:20, w/w) in- gests in the Presence and absence of (6R)-BH4. HPGF was

3000 sw

staining, the digest under the above conditions showed four corresponded M r = 210,000 and 112,000 (Fig- 3A)* They

I I I I I I Fig. 3B shows the elution profile of the chymotryptic digest

0 1 0 # ) 3 0 4 0 5 0 8 0 matography on HPGF after a 20-fold concentration (from 0.4

motrypsin (25 pg/ml) in the absence (c-".) and in the presence of the product A from the native Phenylalanine hydroxylase

Proteolytic Modification of Phenylalanine Hydroxylase 2053

/N

< -A

‘D

Lu

B

S 0 5 1 0 1 5 3 0 4 5 6 0 TIME (mid

S 0 5 1 0 1 5 30 45 60 TIME (mid

FIG. 2. SDS-PAGE patterns of time courses of the proteolysis of phenylalanine hydroxylase by a- chymotrypsin in the absence ( A ) and presence ( B ) of (6R)-BH4. The proteolysis was carried out under the same conditions as in Fig. 1 without ( A ) and with ( B ) l O p ~ (6R)-BH,. At the indicated times, 20-4 aliquots were withdrawn, added to 0.5 pg of phenylmethylsulfonyl fluoride, lyophilized, and then subjected to SDS-PAGE as described under “Methods.” The slot on the left was for standard proteins: phosphorylase b (M, = 94,000), bovine serum albumin (M, = 66,300), ovalbumin (M, = 43,000), carbonic anhydrase ( M , = 30,000), trypsin inhibitor (M, = 20,100), and a-lactalbumin ( M , = 14,400). Bunds A, B, C, and D represent proteolytic degradation products. See text for details.

/N -A

0 10 20 TIME lminl

7 c

0 10 20 TIME (mid

0 10 TIME Iminl

20

FIG. 3. HPGF of the native enzyme ( A ) and the chymotryptic digests in the presence ( B ) and absence (C) of (6R)-BH,. Conditions for HPGF are described under “Experimental Procedures.” For each chromatography, 250-p1 fractions were collected and 6MPH,-dependent phenylalanine hydroxylase activity (GMPH,, 0.15 mM; Phe, 5 mM) was measured using 20 pl of each fraction as described in Ref. 4. Each closed circle in the lower profile in the figures expresses the total activity in each fraction (250 pl). A , the native enzyme (1.1 mg in 50 pl) was applied. B, the native enzyme (0.5 mg/ml) was digested with a-chymotrypsin (final concentration, 25 pg/ml) in the presence of 10 p~ (6R)-BH4 for 60 min as described under “Experimental Procedures.” The digest was concentrated 20-fold and 110 p1 of the concentrate was applied. C, the condition was exactly the same as B except that the digestion was performed in the absence of (6R)-BH4.

of the product estimated by the elution volume from HPGF was 51,000, whereas the SDS-PAGE pattern showed a single band, M, = 36,300 (data not shown). HPGF in the presence of 6 M guanidine hydrochloride also gave a similar molecular weight of 36,000. We have previously found (3) that hydroxylase that had been fully activated by limited proteolysis with chymotrpysin is a dimer of M, = 67,000 composed of two M, = 35,000 subunits (see “Discussion”). It should be noted that the hydroxylase activities shown in Fig. 3 were all determined in the presence of GMPH,. Thus, even though the hydroxylase is markedly activated during its conversion to enzyme C (see Fig. l), this activation, as has been reported previously (3), is only expressed when the enzyme is assayed with BH4 and not with GMPH, (see also Table I).

Other Products and Inter-relation of Modified Enzymes- Two chymotrypsin-modified enzymes A and C were purified. It should be noted that enzyme A is composed of subunits, M, = 47,000, which are designated as band A in Fig. 2 and that enzyme C is composed of subunits M, = 36,000, which are designated as band C in Fig. 2. Attempts were also made to isolate product B shown in Fig. 2 as an active enzyme species by changing the conditions for chymotryptic digestion of the native phenylalanine hydroxylase. These attempts were not successful (see “Discussion”). The protein corresponding to product D ( M , = 29,500) shown in Fig. 2 was isolated by HPGF from the 75-min digest without (6R)-BH,, but no phenylalanine hydroxylase activity could be detected indicating that the product D had undergone overdigestion and


-

-

0 0.2 0.4 0.6 0.8 1.0 4 12 16

L-PHENYLALANINE ImM) TIME h i n l FIG. 4 (left). Phenylalanine saturation curves for the native and the chymotrypsin-modifiedenzymes.

Velocities were measured by monitoring the rate of NADH oxidation in the coupled assay in the presence of dihydropteridine reductase. 40 p~ BH, was used for each assay. The velocity was expressed as mol of NADH oxidized per min per mol of the enzyme subunit. Phenylalanine hydroxylase concentrations in 1.0 ml of total assay volume were: native enzyme, 35 pg (U); enzyme A, 43 pg (o"0); enzyme C, 8.8 pg (U).

FIG. 5 (right). HPGF of the CNBr digests of the native enzyme and the chymotrypsin-modified enzymes in the presence of 6 M guanidine hydrochloride. Column, TSK G-3000 SW (7.5 X 600 mm); elution buffer, 10 mM potassium phosphate buffer (pH 6.8) containing 1 mM EDTA and 6 M guanidine hydrochloride; flow, 1 ml/min at 25 "C. The column was calibrated using standard proteins as described in Ref. 10. After CNBr digestion (IO), the samples were lyophilized and preincubated with the elution buffer containing 0.1 M dithiothreitol at 37 "C for 1 h and applied to HPGF. N , 160 pg of the digest of native phenylalanine hydroxylase; A, 70 pg of the digest of the enzyme A; C, 130 pg of the digest of the enzyme C.

TABLE I Kinetic characteristics of the chymotrypsin-modified and native

phenylalanine hydroxylases

Enzyme species BH, assay" GMP%

assaf

K , for Pheb ullEnl' UllEol' mM min" min"

Native 0.25 4.3 354.2 A 0.28 3.0 340.1 C 0.14 52.2 343.0

a The enzyme activity was measured following the phenylalanine- dependent NADH oxidation spectrophotometrically as described previously (4). The final concentrations of BH, and GMPH, were 40 p~ and 150 pM, respectively.

Apparent values. The value for the native enzyme is the phenylalanine concentration for half-maximum velocity.

e Values are expressed as mol of NADH oxidized per min per mol of subunit. Phenylalanine concentrations were 0.2 mM for BH, and 5 mM for GMPH, assays.

inactivation. To clarify the relationship of these products, the following experiment was carried out. Purified enzyme A was treated with chymotrypsin in the absence of (6R)-BH4 (a- chymotrypsin/enzyme A, 1:20, w/w) as described under "Methods". A rapid increase in the hydroxylase activity was observed similar to the proteolytic activation of the native phenylalanine hydroxylase. SDS-PAGE (data not shown) showed that the enzyme C (Mr = 36,000) was produced but, unlike the case of the native enzyme, no stained band was observed corresponding to the band B shown in Fig. 2. After further digestion, band D was also detected (data not shown).

Kinetic Properties of Chymotrypsin-modified Enzyme-As mentioned above, conditions to prepare the two different chymotryptic products were established. It was found that the native enzyme and both of the modified enzymes had nearly

identical molecular activity when assayed with GMPH,. Since most types of activation of this enzyme can be observed easily only when BH, is used as a cofactor (3, 4, 8, 18), initial velocity analyses of the native and modified enzymes were carried out in the presence of BH, (Fig. 4). Their kinetic characteristics are summarized in Table I. As reported previously (3, 4), the native enzyme showed a sigmoidal initial velocity versus phenylalanine concentration curve with a Hill coefficient of 2 (19). The half-maximal concentration of phenylalanine was approximately 0.25 mM. The modified enzyme A (subunit M, = 47,000), however, showeda typical hyperbolic saturation curve with a K, of 0.28 mM. The maximal activity per subunit was approximately 60% that of the native enzyme. These results showed that the enzyme A had not undergone activation, although it had lost a M, = 5,000 portion(s). The modified enzyme C (subunit M, = 36,000), on the other hand, showed a dramatic increase in activity. The phenylalanine saturation curve was hyperbolic (apparent K, = 0.14 mM) and was also characterized by marked substrate inhibition at phenylalanine concentrations above 0.2 mM. The results showed that the enzyme C was one of 'the chymotrypsin- activated species which had been reported previously but not been identified thoroughly as a distinct molecular species (3). The kinetic behavior is essentially the same as that of the lysolecithin-activated (3) and sulfhydryl-modified (4) enzymes.

Existence of the Phosphorylation Site on Chymotrypsin- modified Enzyme-We have previously reported (5-8,20) that in vitro phosphorylation catalyzed by cyclic AMP-dependent protein kinase activates phenylalanine hydroxylase several- fold. Furthermore, the activation by chrymotrypsin and by phosphorylation were not additive because the phosphorylation site was removed during the course of activation by chymotrypsin ( 5 ) . These data suggested that the phosphoryla-

Proteolytic Modification of Phenylalanine Hydroxylase 2055

Native Acetyl-NH-ala Met COOH Enzyme

Mr=52000 " Mr=33000 Mr=20000

FIG. 6. Scheme showing the conversion of native phenylalanine hydroxylase to modified species A and C through the action of chymotryp-

-Mr=5000 from COOH-terminus

sin. As mentioned in the text, there-is I also evidence for an alternate pathway A ~ ~ ~ ~ ~ - N H - ~ ~ ~ for the conversion of native enzyme to C, not shown in the figure, which in- volves the formation of modified species B.

Met COOH

-Mr=llOOO from NH2-terminus

NH,-SlY Met COOH

tion site may be on the region whose removal was responsible for the activation of this enzyme. To investigate this possibil- ity in greater detail, in vitro phosphorylated phenylalanine hydroxylase was prepared using [32P]ATP and CAMP-dependent protein kinase (16); the phosphorylated product was subjected to proteolysis with chymotrypsin in the presence and absence of (6R)-BH4. The proteolyzed products were isolated by HPGF in the same manner as described above. The pattern of proteolyzed products was essentially the same as that seen on digestion of the nonphosphorylated enzyme. Phosphoryl- ated species corresponding to the enzymes A and C were collected. Approximately 0.6-mol of 32P was incorporated per mol of the subunit of the native phenylalanine hydroxylase by the protein kinase system. Enzyme A obtained by the chymotrypsin digestion of 32P phosphorylated enzyme contained 0.57-mol of 32P per subunit ( M , = 47,000), whereas no significant amount of radioactivity was detected in the species corresponding to, C.

Location of Modified Regions-It was important to deter- mine the terminal(s) from which the proteolytic modification occurred. As stated above, in the process of chymotryptic activation, the phosphorylation site was cleaved off. Since we have previously shown that the phosphorylation site is located within the NH2-terminal ( M , = 33,000) portion (IO), it seemed likely that the proteolytic modification might occur from the NH,-terminal end of the subunit. To obtain more detailed information on this point, cyanogen bromide cleavage was carried out on the proteolytically modified enzymes. We have previously reported (10) that there is only one methionine residue per subunit; therefore, cyanogen bromide cleavage of the subunit gave two peptides (NH2-terminal side, CB1, M, = 33,000; COOH-terminal side, CB2, M, = 20,000). The native enzyme and the two chymotrypsin-modified species were treated with cyanogen bromide and the digests were applied to HPGF in the presence of 6 M guanidine hydrochloride (10). Fig. 5 shows the elution patterns of the enzymes A and C compared to the native enzyme ( N ) . A typical pattern on HPGF of a cyanogen bromide digest of the native enzyme

Modified Enzyme A Mr=47000

Modified Enzyme C Mr=36000 (activated)

(Fig. 5) shows three peaks, uncleaved enzyme ( M r = 52,000; retention time, 11 min), peptide CB1 ( M , = 33,000; time, 12.4 min), and peptide CB2 (MI = 20,400; time, 14.4 min). Enzyme species A also gave three peaks, uncleaved A ( M r = 47,000; time, 11.4 min), intact peptide CB1 ( M , = 33,000; time, 12.4 min), and shortened peptide CB2 ( M , = 16,000; time, 15.0 min). The chromatogram of the digest of enzyme species C showed uncleaved C ( M I = 36,000; time, 12.2 min) and an asymmetric peak at M, = 20,000-15,000, but no intact CB1 was observed. The results indicated that enzyme C had undergone proteolysis from the NH, terminus.

From the above results, we assume that during chymotryptic activation, the COOH-terminal M, = 5,000 fragment is first liberated to form enzyme A. This species is next con- verted to C, the activated species, by the loss of the NH,- terminal M, = 11,000 fragment. Added support for this scheme was provided by a determination of the amino acid of the modified enzymes with the Edman method (15). The NH, terminus of enzyme A could not be detected as in the case of the native enzyme (10). Enzyme C, however, had a single free NH, terminus, glycine (yield 67%, mol/mol of subunit). The scheme shown in Fig. 6 summarizes the key structural changes involved in the conversion of native enzyme to forms A and C.

DISCUSSION

Phenylalanine hydroxylase catalyzes the conversion of phenylalanine to tyrosine, a reaction that is the initial and rate-limiting step in phenylalanine catabolism. The activity of this enzyme, therefore, must be carefully regulated to avoid the wasteful consumption of the phenylalanine pool of the body. It has been proposed that the phenylalanine hydroxylase molecule may be in a relatively inactive conformation in the basal state (2, 3, 16) and that its regulation may involve a change in its conformation to one or more forms which have increased hydroxylase activity. Indeed, as mentioned previously, a variety of activations in vitro as well as in vivo are known. On the basis of the characteristics of the activated


enzyme, the activations that have been reported can be grouped into three different types. One of them is exemplified by activation by lysolecithin and would also include proteolytic activation (3) and activation by sulfhydryl modification (4). This type of activation is characterized by a 20-30-fold increase of BH,-dependent phenylalanine hydroxylating activity at low concentrations (0.1-0.2 mM) of the substrate, decreased K , for phenylalanine, and broadened substrate specificity (21). Another type of activation is that mediated by phosphorylation which increases the maximum activity 3- 4-fold without changing the kinetic characteristics (5-7). “Spontaneous” activation (18) might be included in this type. Substrate activation is the third type of activation (1, 2). Preincubation with phenylalanine as well as some other amino acids (21) increases the initial velocity 20-30 fold.

The structural changes that underlie these different types of activation have not been identified. Fisher and Kaufman (3) proposed an “inhibitory peptide” model for the enzyme in order to explain various types of activation. According to this proposal, activation of the enzyme by lysolecithin leads to a reversible displacement of the inhibitory peptide, whereas activation by limited proteolysis leads to a more drastic change resulting in the release of the peptide itself from the catalytic domain.

We have now shown that during chymotryptic digestion, the COOH-terminal ( M , = 5,000) region is first liberated from the native subunit of phenylalanine hydroxylase generating enzyme A. This species of enzyme was found to have a dimeric structure when estimated by HPGF. Details of the mechanism of oligomerization of the subunits have not been clarified, but it seems likely that the cleaved M, = 5,000 region may be essential for the formation of tetramers from two dimers. Subsequently, the NHP-terminal ( M , = 11,000) portion, which includes the phosphorylation site, is released; in this step, the enzyme is activated more than 30-fold when activity is measured in the presence of BH,. The apparent K,,, value for phenylalanine is decreased and substrate inhibition is observed (Fig. 4 and Table I). The molecular weight of the active enzyme C determined by HPGF is 51,000 and the subunit molecular weight determined by SDS-PAGE and HPGF with guanidine hydrochloride is around 36,000. The discrepancy between the value of 51,000 and our previously reported value of 67,000 (3) for this species may be explained by the increased hydrophobicity of enzyme C that might retard its elution during HPGF analysis. We tentatively assume that the enzyme may have a dimeric structure. From the observation that enzyme A accumulates during chymotryptic digestion in the presence of (6R)-BH, (Figs. 2B and 3B), the protective effect of (6R)-BH4 appears to be exerted mainly in the step involving the conversion of A to C. As mentioned earlier, the cleaved M, = 11,000 region may correspond to a domain or a hypothetical “inhibitory peptide” which regulates the hydroxylase activity negatively (3).

Many phosphoproteins share a common structural feature, namely,, that the phosphorylation site is located near either end of the polypeptide and this phosphorylation site is readily cleaved by limited proteolysis. In muscle phosphofructokinase (23), the phosphorylated region is located on a terminal peptide of 33-35 amino acids that is liberated by subtilisin digestion of the protein. In the presence of ligands such as ATP, MgC12, and citric acid, the enzyme is in an inhibited conformational state that also inhibits liberation of the phosphorylation site. Glycogen phosphorylase b‘, the proteolytic product of phosphorylase a produced by the limited action of trypsin (24), has been shown to have lost the NH,-terminal phosphorylation site (Serls from the NH2 terminus).

Glycogen phosphorylase shares some other structural characteristics .with phenylalanine hydroxylase. Glycogen phosphorylase b’ (24), the proteolyzed product mentioned above, has also lost the ability to associate into tetramers under the conditions in which the undigested phosphorylase readily associates. This change in behavior is similar to that which occurs with the chymotrypsin-modified hydroxylase. Further- more, the NH, terminus of glycogen phosphorylase is acety- lated (25) like phenylalanine hydroxylase (10). The physio- logical role of the acetylation of phenylalanine hydroxylase is unclear but it might function to protect the putative regulatory domain from the attack of exopeptidases in vivo and by this means prevent the unregulated increase of phenylalanine hydroxylase activity.

We have demonstrated that modified enzyme C is produced from the enzyme A. It should be emphasized that starting with enzyme A, product B (Fig. 2, M, = 40,000) is not detectable. From this finding, it seems likely that product B is produced by liberating the M, = 11,000 fragment from the NH2 terminus of the native enzyme as the first event. If that is the case, the chymotryptic digestion may have two path- ways: (native + A + C) and (native + B + C). We have not established conditions to isolate product B as a single species, perhaps because it has a tetrameric structure and therefore may be difficult to isolate by HPGF.

1. 2.

3.

4.

5.

6.

7.

8.

9.

10.

11.

12.

13. 14. 15.

16.

17.

18.

19. 20.

21.

22.

23.

24.

25.

REFERENCES Tourian, A. (1971) Biochim. Biophys. Acta 242,345-354 Shiman, R., and Gray, D. W. (1980) J. Biol. Chem. 255 , 4793-

Fisher, D. B., and Kaufman, S. (1972) J. Biol. Chem. 248,4345-

Parniak, M. A., and Kaufman, S. (1981) J. Biol. Chem. 2 5 6 ,

Abita, J.-P., Milstein, S., Chang, N., and Kaufman, S. (1976) J.

Donlon, J., and Kaufman, S. (1978) Biochem. Biophys. Res.

Donlon, J., and Kaufman, S. (1978) J. Bwl. Chem. 253 , 6657-

Kaufman, S., Hasegawa, H., Wilgus, H., and Parniak, M. (1981)

Phillips, R. S., Iwaki, M., and Kaufman, S. (1983) Biochem.

Iwaki, M., Parniak, M. A., and Kaufman, S. (1985) Biochem.

Shiman, R., Gray, D. W., and Pater, A. (1979) J. Biol. Chem.

Bailey, S. W., and Ayling, J. E. (1978) J. Biol. Chem. 2 5 3 , 1598-

Weber, K., and Osborn, M. (1969) J. Biol. Chem. 244,4406-4412 Prahl, J. W., and Porter, R. R. (1968) Biochem. J . 107 , 753-763 Blomback, B., Blomback, M., Edman, P., and Hassel, B. (1966)

Phillips, R. S., and Kaufman, S. (1984) J. Biol. Chem. 259,2474-

Horiike, K., Tojo, H., Iwaki, M., Yamano, T., and Nozaki, M.

Hasegawa, H., and Kaufman, S. (1982) J. BWl. Chem. 257,3084-

Hill, A. J. (1913) Bwchem. J. 7,471-480 Milstien, S., Abita, J.-P., Chang, N., and Kaufman, S. (1976)

Kaufman, S., and Mason, K. (1982) J. Bid. Chem. 257, 14667-

Fukushima, T., and Nixon, J. (1980) Anal. Biochem. 1 0 2 , 176-

Riquelme, P. T., and Kemp, R. G. (1980) J. Biol. Chem. 255,

Graves, D. J., Mann, S. A. S., Philip, G., and Oliveira, R. J. (1968)

Koide, A., Titani, K., Ericsson, L. H., Kumar, S., Neurath, H.,

4800

4353

6876-6882

BWl. Chem. 251,5310-5314

Commun. 78,1011-1017

6659

Cold Spring Harbor Conf. Cell Prolit. 8 , 1391-1406

Biophys. Res. Commun. 110,919-925

Bwphys. Res. Commun. 126,922-932

254,11300-11306

1609

Biochim. Biophys. Acta 115 , 371-396

2479

(1982) Biochem. Int. 4,477-483

3089

Proc. Natl. Acad. Sci. U. S. A. 73, 1591-1593

14678

188

4367-4371

J. Biol. Chem. 243,6090-6098

and Walsh, K. A. (1978) Biochemistry 17,5657-5672

Proteolytic Modification of the Amino-terminal and Carboxyl-terminal

Documents

Transcript of Proteolytic Modification of the Amino-terminal and Carboxyl-terminal