catalytic core of rat tyrosine hydroxylase: terminal deletion analysis of bacterially expressed...

E L S E V I E R Biochimica et Biophysica Acta 1206 (1994) 113-119

BB, Biochi~ic~a et Biophysica A~ta

Catalytic core of rat tyrosine hydroxylase: terminal deletion analysis of bacterially expressed enzyme

Stephen J. Walker a Xuan Liu b Robert Roskoski c, Kent E. Vrana d,.

a Department of Biochemistry, Bowman Gray School of Medicine, Wake Forest University, Winston-Salem, NC 27157-1016, USA b Molecular Biology Institute, University of California at Los Angeles, 405 Hilgard St., Los Angeles, CA 90024, USA

c Department of Biochemistry and Molecular Biology, L SU Medical Center, 1100 Florida Ave, New Orleans, LA 70112, USA a Department of Physiology and Pharmacology, Bowman Gray School of Medicine, Wake Forest University, Winston-Salem, NC 27157-1083, USA

(Received 2 November 1993)

Abstract

Tyrosine hydroxylase (TH) catalyzes the rate-limiting step in catecholamine biosynthesis. This enzyme is hypothesized to consist of an amino-terminal regulatory domain and a carboxy-terminal catalytic domain. In the present studies, we have utilized recombinant DNA techniques to map the boundaries of the regulatory and catalytic domains of TH. We have isolated a full-length cDNA clone for rat pheochromocytoma TH and have expressed the enzyme in bacteria. Utilizing this bacterial expression system and polymerase chain reaction technology, we have constructed and subcloned genes for five amino-terminal deletion mutants (NA40, NA155, NA165, NA175 and NA200; NA denotes amino-terminal deletion and the numerical value denotes the number of amino acids deleted) and two carboxy-terminal deletion mutants (CA19 and CA50). The catalytic core of rat tyrosine hydroxylase has been identified to include the region from amino acid #165 to amino acid #479. The amino-terminal deletion mutants, NA40, NA155 and NA165 are from 1.85 to 2.5-fold more active than unmodified recombinant TH, while the removal of 19 amino acids from the C-terminus (CAI9) results in a 70% reduction in enzyme activity. Removal of additional sequences (ten more residues from the N-terminus [NA175]; or an additional 31 amino acids from the C-terminus [CA50]) results in protein that is totally without enzyme activity. As expected, removal of 40 (or more) N-terminal amino acids abolishes the ability of the catalytic subunit of the cAMP-dependent protein kinase to phosphorylate the recombinant enzyme; serine-40 is the phosphorylation site on TH for PKA. We conclude that the N-terminal boundary for the TH catalytic domain resides between residues 165 and 175 and that removal of this N-terminal domain (totally or partially) increases the activity of the enzyme. These findings confirm previous reports that proteolytic cleavage at amino acid #158 produces an active (and activated) catalytic fragment.

Key words: Bacterial expression; Catalytic domain; Deletion mutagenesis; Tyrosine hydroxylase

I. Introduction

Tyrosine hydroxylase (EC 1.14.16.2; TH) catalyzes the rate-limiting and committed step in the biosynthesis of the catecholamine hormones and neurotrans- mitters (dopamine, norepinephrine and epinephrine;

* Corresponding author. Fax: + 1 (919) 7167738. Abbreviations: BH4, tetrahydrobiopterin; /3-gal TH, recombinant beta-galactosidase/tyrosine hydroxylase fusion; CA, carboxyl terminal deletion; IPTG, isopropylthiogalactopyranoside; 6-MPH4, 6- methyl-5,6,7,8-tetrahydropterin; NA, amino-terminal deletion; PCR, polymerase chain reaction; PKA, cAMP-dependent protein kinase; SDS, sodium dodecyl sulfate; TH, tyrosine hydroxylase

0167-4838/94/$07.00 © 1994 Elsevier Science B.V. All rights reserved SSDI 0167-4838(94)00008-5

[1]). This pivotal enzyme is phosphorylated and activated by a number of protein kinases, including cyclic AMP-dependent protein kinase [2,3], ca lc ium/ca lmod- ulin-dependent protein kinase II [4], protein kinase C [5], and cyclic GMP-dependent protein kinase [6]. Sev- eral different serine residues of T H have been identified as potential phosphorylation sites [7,8]. More recent reports indicate that the enzyme is also a substrate for two microtubule-associated kinases [9], as well as a cell cycle-associated protein kinase [10]. Phos- phorylation has been postulated to be physiologically important in the regulation of T H in that it is associated with increases in the activity of the enzyme (see [111).

114 S.J. Walker et al. / Biochimica et Biophysica Acta 1206 (1994) 113-119

Limited proteolysis of TH has also been shown to increase enzyme activity [12-17]. In a recent study [17], Abate and Joh reported that trypsin hydrolysis of partially purified rat brain TH generates a 34 kDa activated fragment which contains the catalytic domain but not the N-terminal regulatory phosphorylation sites. Since this partially proteolyzed form of the enzyme is more active than the native enzyme, it appears that by removing the regulatory region, which includes the phosphorylation sites, the enzyme is released from inhibition and becomes activated.

One potentially useful strategy for the characterization of the post-translational regulation of TH lies in the generation of site-specific mutants in the enzyme and the expression and characterization of these mutants. Bacterial expression has many advantages for this type of analysis of TH [18,19]. Our approach has been to manipulate the TH gene through PCR-generated deletion and point mutation and then to express the protein in bacteria and subsequently characterize the recombinant enzyme. We herein report the isolation and engineering of a full-length rat TH cDNA clone and its subsequent expression as a /3-galactosidase fusion protein in E. coli. The recombinant enzyme can be phosphorylated by exogenous catalytic subunit of cyclic AMP-dependent protein kinase with a concomitant activation associated with a decrease in its K m for 6-MPH 4. We have utilized this expression system to generate and to characterize N-terminal and C-terminal deletion mutants in order to refine the boundaries of the catalytic core of rat TH.

2. Materials and methods

Methods cDNA isolation and characterization. A full-length cDNA clone for TH was isolated from

a rat pheochromocytoma library. Total RNA was isolated from frozen rat pheochromocytoma tissue [20], poly (A+)-containing RNA was selected (using messen- ger affinity paper; Amersham, Arlington Heights, IL) and this RNA was utilized to generate double-stranded cDNA molecules [21]. Decameric BamHI linkers were added to the cDNA, the molecules were size-selected (> 1200 bp), and the library was ligated into the BamHI site of pSP6-5 (Promega Corp., Madison, WI). Follow- ing the transformation of the HB101 strain of E. coli, TH-bearing clones were identified using a partial cDNA fragment (provided by Dr. D. Chikaraishi, Tufts Univ.; [22]). The full-length cDNA clone contained 29 bp of 5'-untranslated sequence, the entire open reading frame encoding TH, 285 bp of 3'-untranslated sequence containing 20 bp representing the poly-adeny- lated tail.

Generation of a clone encoding a ~-galactosidase fusion protein of TH.

The 1808 bp TH cDNA was engineered such that extraneous sequences at the 5' end of the gene were eliminated and a HindIII linker was added (for direc- tional cloning). This was accomplished through PCR amplification of the gene using a primer which coded for a 5' HindIII restriction site followed by the initiator methionine codon and the codons representing the initial six amino acids of rat TH (universal 5' primer; 5 ' - C C A A G C T T G A T G C C C A C C C C C A G C G C C C C C - 3'). The 3' end was simplified through the use of a primer which was designed such that it was the com- plementary DNA sequence to the final six codons (and the termination codon) of TH followed immediately by a BamHI restriction site (universal 3' primer; 3'- C G T G A C T C A C G G T A A T C G A T T C C T A G G C C - 5 ' ) . This DNA fragment, purified from agarose utilizing iodine-based extraction (GeneClean; Bio 101, La Jolla, CA), was then subcloned, using standard procedures [23], into a HindIII /BamHI-diges ted pUC9 expression vector (Fig. 1A). The resulting recombinant plasmid encodes the entire 498 amino acids of rat TH plus an additional eight N-terminal amino acids (Fig. 1B), and contains a 5' HindIII site and a 3' BamHI site. The plasmid contains the lacUV5 promoter, which can be induced by the non-hydrolyzable lactose analog IPTG, adjacent to a multiple cloning site.

A proteinase-deficient strain of E. coli, BL21, was used to prepare competent cells [24]. The bacterial B strain BL21 (F- , ompT, rBm s) has the advantage that it is deficient in the Ion proteinase and lacks the ompT outer membrane proteinase that can degrade proteins during purification [25]. Following transformation with the plasmid construct just described, the recombinant TH (referred to as /3-gal TH) was induced and ho- mogenates prepared as previously described [18].

Construction of mutants of ~-galactosidase TH. In order to identify the boundaries of the catalytic

core of TH, a series of deletion mutants was constructed (Fig. 3). This involved making a truncated DNA, in intact translational reading frame, from either the 5' (for amino-terminal deletions) or 3' (carboxyl terminal deletions) end of the gene and then expressing the mutant protein. The f~- mer was accomplished using specifically designed primers directed to the 5' end of the gene. For example, to generate an NA40 mutant (N denotes N-terminal deletion; the numeral denotes the number of amino acids which were deleted in this construct), a primer was designed such that it is identical to the universal 5' primer, up to and including the HindIII linker sequence, but the remainder of the primer codes for amino acids 41-47 (instead of the initiator methionine and amino acids 2-7). This primer, when used with the universal 3' primer and full-length ~-gal TH as template in a PCR amplification, gener-

S.J. Walker et al. / Biochimica et Biophysica Acta 1206 (1994) 113-119 115

ates a truncated TH gene. This recombinant still re- tains the additional eight 5' amino acids (five from fl-gal and three from the HindlII linker), but is shorter, by 120 bp (coding for the first 40 amino acids) than /3-gal TH. This gel-purified PCR product was digested with BamHI and HindlII, subcloned into pUC-9 and used to transform BL21 cells. The truncated enzyme was induced and prepared as described earlier. The other N-terminal deletion mutants were produced in an analogous fashion.

C-terminal deletion mutants (CA19 and CA50) were generated in much the same way as the N-terminal deletion mutants. In these cases, however, the intact universal 5' primer was utilized and the 3' primer was designed such that PCR amplification resulted in a gene truncated at the appropriate position (by inserting a stop codon for amino acid No. 480 in Czal9 or at the codon for amino acid No. 449 in CA50) followed by a BamHI restriction site.

Polyacrylamide gel electrophoresis and Western blot analysis.

Denaturing polyacrylamide gel electrophoresis (SDS-PAGE) was carried out on aliquots from bacterial extracts [26]. Following electrophoresis, the proteins were transferred to reinforced nitrocellulose membrane (Duralose; Stratagene, La Jolla, CA) in a semi-dry transblot apparatus (OWL Scientific, Cam- bridge, MA) according to the manufacturer's specifica- tions. After transfer, the membrane was soaked in blocking buffer (5% BSA in 10 mM Tris-C1 (pH 7.5), 0.05% Tween 20, 0.04% sodium azide and 0.9% sodium chloride) overnight at 4°C. The blocked filter was then incubated in the presence of immunopurified rabbit a-rat TH antibody (a 1 : 2500 dilution in blocking buffer; Pel-Freeze Biologicals, Rogers, AR) for one hour, with shaking, at room temperature. Immunoreactive TH bands were visualized by chemiluminescence on X-ray film following incubation with an enzyme-conjugated 2 nd antibody (donkey a-rabbit Ig conjugated to horseradish peroxidase; Amersham, Arlington Heights, IL) using a commercially available, non-radioactive Western blotting kit (ECL Western Blotting Analysis Systems; Amersham, Arlington Heights, IL). Alterna- tively, immunoreactive bands were visualized by autoradiography following incubation of the primary (anti- TH) complex with 125I-protein A (Amersham, Arling- ton Heights, IL).

TH specific activity determination. TH activity (expressed as nmol/h per ml) was deter-

mined using a 3H20 release assay (Ref. 27; at 100/xM tyrosine and 1 mM 6-MPH4). These values were con- verted to specific activities by normalization to the amount of immunoreactive TH as determined by quantitative Western blots (Figs. 2 or 4; the Western blot autoradiographs were scanned using a Microcomputer Imaging Device, MCID; Imaging Research, Ont.,

Canada). These values were then expressed as a per- centage of a /3-gal TH control activity within each independent experiment. This final correction was required to allow for comparison of values from independent experiments. The graphs represent the mean (and S.E.M.) of three independent experiments.

TH phosphorylation in vitro. The purified pheochromocytoma TH, partially puri-

fied fl-gal TH and NA40 mutant (which lacks the site for phosphorylation) were subjected to phosphorylation by the purified catalytic subunit of cAMP-dependent protein kinase (PKA C-subunit) under the following conditions. A premixture was prepared containing the C-subunit (250 /xg/ml) and MgCI 2 (250 mM) in the presence of [y-32p]ATP (10 /xCi/ml). TH-containing bacterial homogenate (23/xl) was added to 2/zl of the kinase premixture, placed on ice, and subsequently subjected to SDS-PAGE (as described previously). The phosphorylated proteins were detected by exposing 3zp-labeled proteins to X-ray film (following transfer to nitrocellulose). TH was visualized in parallel unlabeled samples by autoradigraphy of a Western blot (hy- bridized with 125I-protein A).

3. Results

Bacterial expression of recombinant rat TH A full-length TH cDNA clone was isolated from a

rat pheochromocytoma library. This gene is 1808 bp (containing 29 bp of 5'-untranslated sequence, the entire open reading frame of TH, and 285 3'-untranslated nucleotides containing a 20-residue poly-A tail) and was subcloned into a pUC expression plasmid. To minimize the number of extraneous nucleotides, while maintaining the translational reading frame of the /3- galactosidase leader residues, PCR was utilized to re- move the 5'-untranslated sequence of the TH cDNA and to replace the 5'-BamHI site with a HindlII site. Untranslated 3' sequences (including the poly (A ÷) tail) were similarly removed by PCR and replaced with a BamHI restriction site. We have therefore generated a fusion protein which contains the entire coding region of rat TH with eight additional N-terminal amino acids (Fig. 1A and B). This recombinant enzyme is expressed following induction of its transcription using IPTG (Figs. 1A and 2A). The recombinant TH is similar to purified pheochromocytoma in its physical structure (tetramer formation), and enzyme kinetic characteristics (most notably Km, pterin) (data not shown).

Cyclic AMP-dependent phosphorylation of recombinant TH

The recombinant TH was a substrate for phosphorylation by the cAMP-dependent protein kinase in

116 S.J. Walker et al. / Biochimica et Biophysica Acta 1206 (1994) 113-119

vitro. Fig. 2 A i l lus t ra tes the W e s t e r n blot prof i le of pur i f ied p h e o c h r o m o c y t o m a T H ( a d d e d to BL21 bac te - r ial extract; lane 2) and the bac te r ia l ly expressed re- combinan t T H (in a BL21 homogena t e ; lane 3). Fig. 2B d e m o n s t r a t e s tha t the r e c o m b i n a n t T H was phosphory la ted in vi tro using the pur i f i ed P K A catalyt ic subunit ( lane 3). F o r compar i son , lane 2 shows the phosphory la t ion of pur i f ied p h e o c h r o m o c y t o m a TH. This p ro t e in k inase fa i led to p r o d u c e r a d i o l a b e l e d bac te r i a l p ro te ins f rom cells which d id not con ta in the expression vec tor ( lane 1). Bac te r i a mus t t he re fo re lack significant amoun t s of subs t ra te for this eukaryo t i c kinase. Phosphory la t ion of the r e c o m b i n a n t enzyme p r o d u c e d a two-fold act ivat ion which is c o m p a r a b l e to tha t ob-

4. M M

¢1 ~ ¢¢ :J;

MWM A MWM B (kDa) 1 2 3 (kDa) 1 2 3

92 .5~ 9 2 . S ~

55 .0~ 55 .0~

Fig. 2. Expression and in vitro phosphorylation of recombinant rat TH. (A) Western blot of recombinant Ttl. Samples were elec- trophoresed by SDS-PAGE and electroblotted onto nitrocellulose membrane (described in Section 2). Immunoreactive bands were visualized by autoradiography following incubation of the primary (anti-TH) complex with lzSI-protein A. Lane 1:BL21 cell extract alone (no plasmid). Lane 2: Purified pheochromocytoma TH added to BL21 bacterial extract. Lane 3: Extract of BL21 cells expressing /3-gal TH. (B) In vitro phosphorylation. Using samples of the same preparations described above (with identical lane designations) the extracts were subjected to phosphorylation (with 32p-ATP), SDS- PAGE, electroblot transfer to nitrocellulose and autoradiography as described in Section 2.

. . . . . ' - - J ' I i I I

B beta-qalactosidase THCodXnq Sequence

ME___TT ~ MET ILE THR Dro se____Er leu MET PRO THR PRO SER ALA PRO - ATG ACC ATG ATT ACG CCA AGC TTG ATG CCC ACC CCC AGC GCC CCC - TAC TGG TAC TAA TGC GGT TCG AAC TAC GGG TGG GGG TCG CBG GGG -

Hind I l l

pUC9 pl asmid sequences

Fig. 1. Bacterial expression of rat TH. Panel A: the pUC-9 expression vector containing the full-length sequence of rat TH. The arrow at the top of the figure is a restriction map of the rat TH gene. This cDNA fragment was cloned into the HindlII/BamHI site of the multiple cloning site and places the cDNA in the 5' end of the /3-galactosidase gene ('lac Z"). In this map, 'lac I' is the lac repressor gene, 'ori' is the bacterial origin of replication, and 'Amp r' is the /3-1actamase, ampicillin-resistence gene. This recombinant construct was synthesized by PCR amplification of a TH cDNA to eliminate extraneous 5' and 3' sequences, as described in Section 2. Panel B: amino-terminal sequence of fl-gal TH. Capitalized amino acids represent authentic/3-galactosidase or TH sequences. Lower case amino acids are encoded by the pUC polylinker sequence. Bold nucleotides are derived from the TH cDNA and light nucleotides from pUC-9 plasmid.

served for the p h e o c h r o m o c y t o m a T H (da ta not shown). P h o s p h o p e p t i d e analysis (by H P L C ) fol lowing in vi tro phosphory la t i on and t rypsin d iges t ion of pur i f ied p h e o c h r o m o c y t o m a T H and the bac te r ia l ly expressed TH, as desc r ibed e l sewhere [7,8], d e m o n s t r a t e d that 32p was i nc o rpo ra t e d exclusively into ser ine-40 (da ta not shown).

Western analysis o f mutants Ful l - l eng th /3-galactosidase T H and seven de l e t ion

mu tan t s (five f rom the amino te rminus and two f rom the carboxyl te rminus ; Fig. 3) were p r o d u c e d in the E. coli s t ra in BL21. R e c o m b i n a n t p ro te ins were subjec ted to S D S - P A G E and t r ans fe r r ed to n i t roce l lu lose for W e s t e r n analysis (desc r ibed in Sect ion 2; Fig. 4). F r o m the W e s t e r n analysis, it is a p p a r e n t tha t each of the m u t a n t p lasmids p r o d u c e d p ro te in and tha t these re- combinan t p ro t e ins a re of the p r ed i c t ed mo lecu la r weights .

Enzyme activity analysis I n d e p e n d e n t cu l tu res of each of the T H de le t ion

mutan ts , as well as of /3-galactosidase TH, were pre- p a r e d and ana lyzed for the p roduc t ion of p ro t e in and for enzyme activity. A l t h o u g h the abso lu te activit ies vary f rom p r e p a r a t i o n to p r e p a r a t i o n (due to differ- ences in enzyme expression) , the specific activity of each r ecombinan t mu tan t d i sp layed a un ique relat ive va lue (Fig. 5). To accompl ish this analysis, the amoun t of specif ic T H pro te in in each p r e p a r a t i o n was de te r - m i n e d by quant i ta t ive W e s t e r n analysis. A f t e r normal - izing the enzyme activit ies to the W e s t e r n blot signals, the da t a were conver t ed to a ra t io with a /3-galacto- s idase T H s t anda rd p re sen t in each exper iment . This


"r'YROSlNE HYOROXYLASE

AMINO TERMINAL DELETION 8ERIER

~ NA40

m , g.li 1.(+

CARBOXYL TERMINAL DELETION SERIES

~ CA19

C ~ 0

Fig. 3. TH deletion series for determination of the catalytic core of the enzyme. PCR was used to selectively delete sequences from the amino and carboxyl termini of TH. In the case of the amino-terminal deletions, the deleting (5') primer was designed to introduce a HindllI site in the proper reading frame and to eliminate the indicated sequence. As a result, all of the amino-terminal deletions start with the same amino-terminal sequence (see Fig. 1B). In the case of the carboxy-terminal deletions, a 3' primer was designed to insert a stop codon (adjacent to a BarnHI linker) in such a manner as to produce premature termination of translation at specific amino-acid positions. The asterisks denote the four known phosphorylation sites for rat TH (serine residues 8, 19, 31 and 40). The @ indicates the position of a trypsin cleavage site which produces an activated fragment of TH (beginning with amino acid No. 158; Ref. 17). NA40 - Deletion of 40 residues which contain all of the post-translational phosphorylation sites of TH. NA155, NA165, NA175 are deletion series near the hypothetical boundary of the regulatory and catalytic domains. CA19 and CA50 are deletions in the carboxyl terminus. The solid regions represent a hypothetical catalytic domain based on published information (Ref. 33).

MWM 1 2 3 4 5 6 7 8 (kDa)

69

46

30

Fig. 4. Western blot of recombinant TH and deletion mutants. The deletion mutants described in Fig. 3 were grown in cultures of the proteinase-deficient bacterial strain BL21 and induced with the lactose analog IPTG. After harvesting the bacteria by centrifugation, the proteins were dissolved in SDS-containing dissociation buffer and subjected to SDS-PAGE, and then transferred to nitrocellulose. TH immunoreactive proteins were detected with a polyclonal antibody to rat TH coupled with chemiluminescence. Lane identification: (1) /3-galactosidase TH, (2) NA40, (3) NA155, (4) NA165, (5) NA175, (6) CA19, (7) CA50, (8) pheochromocytoma TH.

• ' r 3 0 0

m

200

o

g U 1

g.

/ /

#'-.gat IM NA40 NA155 NA165 NA175 CA19 CA50 (control)

Fig. 5. Specific activity of /3-galactosidase TH and recombinant deletion mutants. Using the TH enzyme assay data and quantitative Western blot analysis, TH specific activity was determined for/3-gal TH and each of the deletion mutants (as described in Section 2). Using/3-gal TH as the standard, an activity ratio was calculated for each of the mutants. The data are expressed as mean activity, as compared to/3-gal TH activity (_+ S.E.M.; based on three independent determinations). Statistical analyses (ANOVA followed by post- hoc analysis with a Newman-Keuls test) showed that activities of each of the first three deletion mutants were significantly different from those for the last three (P < 0.05).

normal ized the data for var ia t ions in Wes te rn blot

intensi t ies f rom exper iment to exper iment . Remova l of

40, 155 or 165 amino- te rmina l res idues (NA40, NA155,

or NA165, respectively) resul ted in an 85% to 150%

increase in enzyme activity. Rem ova l of just 10 addi-

t ional amino acids (NA175) p roduced a T H - i m m u n o -

react ive pro te in with no enzyme activity. The same

observat ion occur red when 200 res idues were r emoved

(NA200; data not shown). If 19 amino acids were

r emoved f rom the carboxyl te rminus (catalytic domain)

of TH, enzyme activity was reduced by approx. 70%

compared to /3-gal TH, and if 50 amino acids are

removed, the resul tant r ecombinan t T H (CA50) had no

activity. Fol lowing removal of the first 40 amino acids

( including Ser-40 which is the P K A phosphoryla t ion

site), the enzyme could no longer be phosphory la ted

(data not shown).

4. Discussion

We have described, in the present repor t , the isola-

t ion of a ful l - length c D N A clone for rat T H and the

subsequent man ipu la t ion of that c lone to crea te a

bacter ia l expression vector . This prokaryot ic vec tor di-

rects the expression of a fusion pro te in composed of

five amino acids encoded by bacter ia l /3-galactosidase,

th ree res idues encoded by a H i n d l I I rest r ic t ion site,

and the 498 amino acids of the rat T H enzyme (Fig.

1B). We have expressed the /3 -ga lac tos idase T H fusion

prote in in a p ro te inase-def ic ien t strain (BL21) of E.

coli . Weste rn blot analysis revea led very little degrada-

t ion of T H in the extracts (Figs. 2 and 4). The recombi-

118 S.J. Walker et aL / Biochimica et Biophysica Acta 1206 (1994) 113-119

nant TH enzyme displayed the predicted molecular weight for the fusion protein following SDS-PAGE and was expressed at appreciable levels (1% to 3% of total cellular protein; Fig. 2A and data not shown). The E. col i expression system also produced significant amounts of TH activity (Fig. 5).

Whereas recombinant TH can be phosphorylated in vitro by the purified PKA catalytic subunit (Fig. 2B, lane 3) bacterial extracts from cells which do not carry the plasmid show no phosphorylated protein (Fig. 2B, lane 1). However, recombinant TH is not phosphorylated by the bacteria in vivo (data not shown). This observation is consistent with the fact that bacteria do not contain the cAMP-dependent or calcium/calmod- ulin-dependent protein kinases [28] and is an important consideration in the selection of an expression system for TH, particularly for studies examining regulation by phophorylation. This is in contrast to the baculovirus system [19] where expressed TH contains stoichiometric amounts of phosphate and exhibits a K m for pterin which is characteristic of the a c t i v a t e d

enzyme. It appears, therefore, that the bacterial system may be well-suited for the study of post-translational phosphorylation of TH (e.g., [18,19] and this report). Our findings on the bacterial expression of TH are at slight odds with previously published reports [31,35]. Namely, we find that the recombinant TH has kinetic characteristics comparable to the wild type, pheochromocytoma enzyme and is activated by phosphorylation (data not shown) while the other investigators find that this is not the case unless dopamine is added to the reaction. Current hypotheses suggest that high-affinity binding of dopamine reverses a kinetic activation of the recombinant enzyme [31,35,36]. In the eukaryotic envi- ronment of catecholaminergic ceils, this complex is formed naturally and so may serve as a posttrans- lational regulatory mechanism. Although the reasons for the discrepancy between the present studies and others in the literature are not currently known, one possible explanation would be the fact that our enzyme contains an additional eight extra amino acids at the N-terminus. Although this does not interfere with other aspects of enzyme structure/function, it may compro- mize the natural high affinity binding of dopamine. Alternatively, our enzyme preparation is a crude bacterial homogenate which may contain small molecules capable of substituting for dopamine in reversing (or masking) the activation phenomenon.

One of the primary objectives of this work was to determine the precise boundary between the catalytic and regulatory domains in TH and to identify the minimal protein sequence necessary for enzyme activity. Since serine residues, known to be involved in the phosphorylation/activation of TH, occur in the amino- terminal end of the TH monomer, it has been sug- gested that the amino terminus contains a regulatory

domain and that the catalytic domain is housed within the carboxyl end of the molecule. It has been shown [13,29] that TH can be activated by trypsin proteolysis and that the resulting fragment cannot be further activated by phosphorylation [29,30]. This proteolytic product has been characterized further as a 34 kDa fragment which is catalytically active [16]. The amino- terminal boundary of this fragment was identified (by amino-acid analysis) as residue No. 158 and it was estimated that the C-terminal end of the proteolyzed monomer lacked approx. 45 amino acids (corresponding to 5 kDa; Ref. 17). More recently, Fitzpatrick and colleagues have expressed the C-terminal 330 amino acids of TH as a catalytically active fragment [34]. These findings concerning the phosphorylation/ activation sites and boundaries of the proteolyzed activated TH fragment, taken together, provided the ratio- nale for the choice of deletion mutants that have been generated and tested. The amino-terminal NA40 mutant TH should lack serine phosphoryl acceptors [7,8], yet should be active. The deletion mutants (NA155, NA165, NA175 and NA200) were generated to identify the amino-terminal boundary of active TH. The as- sumption was that the trypsin cleavage site (amino acid 158), characterized by Abate and Joh [17], did not have to exist precisely at the N-terminal boundary of the catalytic domain. Moreover, the 330 residue catalytic fragment of Daubner et al. [34], which corresponds to our NA155, was based on sequence homologies between TH, tryptophan hydroxylase and phenylalanine hydro~lase and also did not have to reside precisely at the boundary. The carboxyl terminal deletion mutants were chosen based on the putative boundary of the carboxyl end of the activated trypsin-digest fragment [171.

From the data in Fig. 5, we conclude that the N-terminal boundary of the TH enzyme catalytic domain resides between amino acid No. 165 and No. 175. This refines the values recently reported by Ribiero et al. [36] which placed the N-terminal boundary between residues 158 and 184. In addition to increasing the intrinsic activity of the recombinant protein, removal of 40 (or more, up to 165) N-terminal residues abolishes the ability of the enzyme to be phosphorylated by PKA (data not shown). The carboxyl boundary of the catalytic domain is between amino acid No. 448 and No. 479. The minimal sequence required for TH activity, therefore, is between amino acid residues 165 and 479.

We were suprised to find that the catalytic domain which we identified agreed so well with earlier information from Abate and Joh [17] on the trypsin diges- tion of the enzyme. We had not predicted that trypsin would prove to be such a sensitive probe since it has rather broad substrate specificity. Clearly, there is a prominant trypsin cleavage site very close to the N- terminal boundary of the catalytic domain. This sug-


gests that there may be a hinge region, conformation- ally available to trypsin, which separates the catalytic and regulatory domains. This region, when deleted by mutation or when cleaved by trypsin, effectively re- moves endogenous tonic inhibition such that the enzyme is now activated. In a similar manner, phosphorylation of this region or mutation of Ser-40 changes the conformation of the regulatory domain such that the enzyme is activated [31,32]. Finally, Woo and co- workers were remarkably accurate at predicting the boundaries of the catalytic domain based on amino-acid sequence homologies between phenylalanine, tryptophan and tyrosine hydroxylases [33]. Based on their analysis, the homologous catalytic domain for these three enzymes begins at amino acid 164 of TH.

In conclusion, we have established (within 10 amino acids) the amino-terminal boundary of the TH catalytic domain (between residues 165 and 175). This confirms the earlier identification of residue 158 as the site of a trypsin cleavage which generates an activated catalytic fragment [17]. We have also established that removal of the N-terminal 40 to 165 amino-acid residues increases the activity of the enzyme. This latter observation is consistent with the prevailing notion that the regulatory domain exerts a tonic inhibition on the catalytic domain.

Acknowledgements

The authors would like to thank Dr. Ali Ardekani for assistance with bacterial expression and Mr. Chris Kolanko for expert technical assistance. They also wish to thank Dr. Sheila Vrana for valuable comments on the manuscript. Aspects of this work were submitted in partial fulfillment of Ph.D. degrees from West Virginia University for S.J.W. and X.L. This work was sup- ported by NIH grants GM-38931 (K.E.V.) and NS- 15994 (R.R.).

References

[1] Levitt, M., Spector, S., Sjoerdsma, A. and Udenfriend, S. (1965) J. Pharmacol. Exp. Ther. 148, 1-7.

[2] Joh, T.H., Park, D.H. and Reis, D.J. (1978) Proc. Natl. Acad. Sci. USA 75, 4744-4748.

[3] Vulliet, P.R., Langan, T.A. and Weiner, N. (1980) Proc. Natl. Acad. Sci. USA 77, 92-96.

[4] Yamauchi, T. and Fujisawa, H. (1981) Biochem. Biophys. Res. Commun. 256, 5404-5409.

[5] Albert, K.A., Helmer-Matyjek, E., Nairn, A.C., Miller, T.H., Haycock, J.W., Green, L.A., Goldstein, M. and Greengard, P. (1984) Proc. Natl. Acad. Sci. USA 81, 7713-7717.

[6] Roskoski, R. Jr., Vulliet, P.R. and Glass, D. (1987) J. Neu- rochem. 48, 840-845.

[7] Vulliet, P.R., Woodgett, J.R. and Cohen, P. (1984) J. Biol. Chem. 259, 13680-13683.

[8] Haycock, J.W. (1990) J. Biol. Chem. 265, 11682-11691. [9] Haycock, J.W., Ahn, N.G., Cobb, M.H. and Krebs, E.G. (1992)

Proc. Natl. Acad. Sci. USA 89, 2365-2369. [10] Hall, F.L., Braun, R.K., Mihara, K., Fung, Y.K., Berndt, N.,

Carbonara-Hall, D.A. and Vulliet, P.R. (1991) J. Biol. Chem. 266, 17430-17440.

[11] Zigmond, R.E, Schwarzschild, M.A. and Rittenhouse, A.R. (1989) Annu. Rev. Neurosci. 12, 415-461.

[12] Musacchio, J.M., D'Angelo, G.L. and McQueen, C.A. (1971) Proc. Natl. Acad. Sci. USA 68, 2087-2091.

[13] Kuczenski, R. (1973) J. Biol. Chem. 248, 5074-5080. [14] Vigny, A. and Henry, J.P. (1981) J. Neurochem. 36, 483-489. [15] Togari, A., Ichikawa, S. and Nagatsu, T. (1986) Biochem. Bio-

phys. Res. Commun. 134, 749-754. [16] Abate, C., Smith, J.A. and Joh, T.H. (1988) Biochem. Biophys.

Res. Commun. 151, 1446-1453. [17] Abate, C. and Joh, T.H. (1991) J. Mol. Neurosci. 2, 203-215. [18] Wang, Y., Citron, B.A., Ribeiro, P. and Kaufman, S. (1991)

Proc. Natl. Acad. Sci. USA 88, 8779-8783. [19] Fitzpatrick, P.F., Chlumsky, L.J., Daubner, S.C. and O'Malley,

K.L. (1990) J. Biol. Chem. 265, 2042-2047. [20] Chirgwin, J.M., Przybyla, A.E., MacDonald, R.J. and Rutter,

W.J. (1979) Biochemistry 18, 5294-5299. [21] Gubler, U. and Hoffman, B. (1983) Gene 25, 263-269. [22] Lewis, E.J., Tank, A.W., Weiner, N. and Chikaraishi, D.M.

(1983) J. Biol. Chem. 258, 14632-14637. [23] Sambrook, J., Fritsch, E.F. and Maniatis, T. (1989) Molecular

Cloning: A Laboratory Manual; 2nd Edn., Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY.

[24] Hanahan, D. (1983) J. Mol. Biol. 166, 557-568. [25] Studier, F.W., Rosenberg, A.H., Dunn, J.J. and Dubendorff,

J.W. (1990) Methods Enzymol. 185, 60-89. [26] Ryrie, l.J. and Galagher, A. (1979) Biochim. Biophys. Acta 545,

1-14. [27] Reinhard, J., Smith, G. and Nichol, C. (1986) Life Sci. 39,

2185-2189. [28] Edelman, A.M., Blumenthal, D.K. and Krebs, E.G. (1987) Annu.

Rev. Biochem. 56, 567-613. [29] Vigny, A. and Henry, J.-P. (1982) Biochem. Biophys. Res. Com-

mun. 106, 1-7. [30] Holdtke, R. and Kaufman, S. (1977) J. Biol. Chem. 252, 3160-

3169. [31] Daubner, S.C., Lauriano, C., Haycock, J.W. and Fitzpatrick,

P.F. (1992) J. Biol. Chem. 267, 12639-12646. [32] Wu, J., Filer, D., Friedhoff, A.J. and Goldstein, M. (1992) J.

Biol. Chem. 267, 25754-25758. [33] Grenett, H.E., Ledley, F.D., Reed, L.L. and Woo, S.L.C. (1987)

Proc. Natl. Acad. Sci USA 84, 5530-5534. [34] Daubner, S.C., Lohse, D.L. and Fitzpatrick, P.F. (1993) Protein

Sci. 2, 1452-1460. [35] Ribiero, P., Wang, Y., Citron, B.A. and Kaufman, S. (1992) J.

Biol. Chem. 267, 9593-9597. [36] Ribiero, P., Wang, Y., Citron, B.A. and Kaufman, S. (1993) J.

Mol. Neurosci. 4, 125-139.

catalytic core of rat tyrosine hydroxylase: terminal deletion analysis of bacterially expressed...

Documents