Structural insight into Parkinson's disease treatment from drug-inhibited DOPA decarboxylase.

DOPA decarboxylase (DDC) is responsible for the synthesis of the key neurotransmitters dopamine and serotonin via decarboxylation of L-3,4-dihydroxyphenylalanine (L-DOPA) and L-5-hydroxytryptophan, respectively. DDC has been implicated in a number of clinic disorders, including Parkinson's disease and hypertension. Peripheral inhibitors of DDC are currently used to treat these diseases. We present the crystal structures of ligand-free DDC and its complex with the anti-Parkinson drug carbiDOPA. The inhibitor is bound to the enzyme by forming a hydrazone linkage with the cofactor, and its catechol ring is deeply buried in the active site cleft. The structures provide the molecular basis for the development of new inhibitors of DDC with better pharmacological characteristics.

Reaction of dopa decarboxylase with L-aromatic amino acids under aerobic and anaerobic conditions.

Analysis of the reaction of dopa decarboxylase (DDC) with L-dopa reveals that loss of decarboxylase activity with time is observed at enzyme concentrations approximately equal to the binding constant, K(d), of the enzyme for pyridoxal 5'-phosphate (PLP). Instead, at enzyme concentrations higher than K(d) the course of product formation proceeds linearly until complete consumption of the substrate. Evidence is provided that under both experimental conditions no pyridoxamine 5'-phosphate (PMP) is formed during the reaction and that dissociation of coenzyme occurs at low enzyme concentration, leading to the formation of a PLP-L-dopa Pictet-Spengler cyclic adduct. Taken together, these results indicate that decarboxylation-dependent transamination does not accompany the decarboxylation of L-dopa proposed previously [O'Leary and Baughn (1977) J. Biol. Chem. 252, 7168-7173]. Nevertheless, when the reaction of DDC with L-dopa is studied under anaerobic conditions at an enzyme concentration higher than K(d), we observe that (1) the enzyme is gradually inactivated and inactivation is associated with PMP formation and (2) the initial velocity of decarboxylation is approximately half of that in the presence of O(2). Similar behaviour is observed by comparing the reaction with L-5-hydroxytryptophan occurring in aerobiosis or in anaerobiosis. Therefore the reaction of DDC with L-aromatic amino acids seems to be under O(2) control. In contrast, the reactivity of the enzyme with L-aromatic amino acids does not change in the presence or absence of O(2). These and other results, together with previous results on the effect exerted by O(2) on reaction specificity of DDC towards aromatic amines [Bertoldi, Frigeri, Paci and Borri Voltattorni (1999) J. Biol. Chem. 274, 5514-5521], suggest a productive effect of O(2) on an intermediate complex of the reaction of the enzyme with L-aromatic amino acids or aromatic amines.

Ornithine and glutamate decarboxylases catalyse an oxidative deamination of their alpha-methyl substrates.

Ornithine decarboxylase (ODC) from Lactobacillus 30a catalyses the cleavage of alpha-methylornithine into ammonia and 2-methyl-1-pyrroline; glutamate decarboxylase (GAD) from Escherichia coli catalyses the cleavage of alpha-methylglutamate into ammonia and laevulinic acid. In our analyses, 2-methyl-1-pyrroline and laevulinic acid were identified by HPLC and mass spectroscopic analysis, and ammonia was identified by means of glutamate dehydrogenase. Molecular oxygen was consumed during these reactions in a 1:2 molar ratio with respect to the products. The catalytic efficiencies (k(cat)/K(m)) of the reactions catalysed by ODC and GAD were determined as 12500 and 9163 M(-1).min(-1) respectively. When the reactions were performed under anaerobic conditions, no ammonia, 2-methyl-1-pyrroline or laevulinic acid was produced to a significant extent. The formation of ammonia and O(2) consumption (in a 1:2 molar ratio with respect to ammonia) were also detected during the reaction of ODC and GAD with putrescine and gamma-aminobutyrate respectively. Taken together, these findings clearly indicate that ODC and GAD catalyse an oxidative deamination of their decarboxylation products, a reaction similar to that catalysed by dopa decarboxylase (DDC) with alpha-methyldopa [Bertoldi, Dominici, Moore, Maras and Borri Voltattorni (1998) Biochemistry 37, 6552-6561]. Furthermore, this reaction was accompanied by a decarboxylation-dependent transamination occurring for GAD, DDC and ODC with a frequency of approx. 0.24%, 1% and 9% respectively compared with that of oxidative deamination.

Preliminary X-ray analysis of a new crystal form of recombinant pig kidney DOPA decarboxylase.

DOPA decarboxylase is responsible for the synthesis of the key neurotransmitters dopamine and serotonin via decarboxylation of L-3, 4-dihydroxyphenylalanine (L-DOPA) and L-5-hydroxytryptophan, respectively. The crystals of recombinant DOPA decarboxylase differ from those previously reported for the enzyme purified from pig kidney. They belong to space group P622 with unit-cell dimensions a = b = 302.6, c = 178.1 A. Both the self-rotation function and the good diffraction quality of these crystals (2.5 A on a synchrotron source) suggest that there should be at least three protein dimers in the asymmetric unit. Diffraction data sets have been collected for the native enzyme and a heavy-atom derivative.

Aromatic amino acid methyl ester analogs form quinonoidal species with Dopa decarboxylase.

This study reports for the first time that binding of aromatic methyl ester analogs to Dopa decarboxylase in the native and inactive nicked forms causes the appearance of a dead-end quinonoidal species absorbing at 500 nm, in addition to an external aldimine absorbing at 398 nm. The equilibrium mixture of these species varies depending on both the analog structure and the enzyme form. The above mentioned intermediates are also characterized with respect to their CD properties and the equilibria for their formation are determined as a function of pH. The results have provided evidence that the establishment of proper contacts between the active site and hydroxyl groups of the ligand are indispensable in order to limit unwanted side reactions.

Cloning and expression of pig kidney dopa decarboxylase: comparison of the naturally occurring and recombinant enzymes.

L-Aromatic amino acid decarboxylase (dopa decarboxylase; DDC) is a pyridoxal 5'-phosphate (PLP)-dependent homodimeric enzyme that catalyses the decarboxylation of L-dopa and other L-aromatic amino acids. To advance structure-function studies with the enzyme, a cDNA that codes for the protein from pig kidney has been cloned by joining a partial cDNA obtained by library screening with a synthetic portion constructed by the annealing and extension of long oligonucleotides. The hybrid cDNA was then expressed in Escherichia coli to produce recombinant protein. During characterization of the recombinant enzyme it was unexpectedly observed that it possesses certain differences from the enzyme purified from pig kidney. Whereas the later protein binds 1 molecule of PLP per dimer, the recombinant enzyme was found to bind two molecules of coenzyme per dimer. Moreover, the Vmax was twice that of the protein purified from tissue. On addition of substrate, the absorbance changes accompanying transaldimination were likewise 2-fold greater in the recombinant enzyme. Examination of the respective apoenzymes by absorbance, CD and fluorescence spectroscopy revealed distinct differences. The recombinant apoprotein has no significant absorbance at 335 nm, unlike the pig kidney apoenzyme; in the latter case this residual absorbance is associated with a positive dichroic signal. When excited at 335 nm the pig kidney apoenzyme has a pronounced emission maximum at 385 nm, in contrast with its recombinant counterpart, which shows a weak broad emission at about 400 nm. However, the holoenzyme-apoenzyme transition did not markedly alter the respective fluorescence properties of either recombinant or pig kidney DDC when excited at 335 nm. Taken together, these findings indicate that recombinant pig kidney DDC has two active-site PLP molecules and therefore displays structural characteristics typical of PLP-dependent homodimeric enzymes. The natural enzyme contains one active-site PLP molecule whereas the remaining PLP binding site is most probably occupied by an inactive covalently bound coenzyme derivative; some speculations are made about its origin. The coenzyme absorbing bands of recombinant DDC show a modest pH dependence at 335 and 425 nm. A putative working model is presented to explain this behaviour.

Dissociation, unfolding and refolding trials of pig kidney 3,4-dihydroxyphenylalanine (dopa) decarboxylase.

The effect of guanidinium chloride (GuCl) on enzyme activity, hydrodynamic volume, circular dichroism, and fluorescence of 3,4-dihydroxyphenylalanine (Dopa) decarboxylase from pig kidney (pkDDC) was studied under equilibrium conditions. Unfolding proceeds in at least three stages. The first transition, occurring between 0 and 1 M GuCl, gives rise to a dimeric inactive species which has lost pyridoxal 5'-phosphate (PLP), and has a high tendency to aggregate, but retains almost all of the native spectroscopic characteristics. The second equilibrium transition, between 1 and 2.2 M GuCl, involves dimer dissociation, with some loss of tertiary and secondary structure. Additionally, gross conformational changes at or near the PLP microenvironment were detected by fluorescence of NaBH4-reduced enzyme. The third step, presumably representing complete unfolding of pkDDC, appears to be complete at 4.5 M GuCl, as indicated by the lack of further substantial changes in any of the signals being studied. Attempts at refolding resulted in the findings that: (1) partial reactivation is observed only starting from enzyme denatured at concentrations below 1.5 M GuCl, and (2) starting from completely denatured protein, the refolding process is apparently reversible down to concentrations of approx. 2 M GuCl. Taken together, this would seem to indicate that the monomer-dimer transition is impaired under the experimental conditions tested. A plausible model is presented for the unfolding/refolding of pkDDC.

Crystallization and preliminary X-ray analysis of pig kidney DOPA decarboxylase.

DOPA decarboxylase from pig kidney, an alpha 2 dimeric enzyme of Mr = 107,000, has been crystallized by the vapour diffusion method with ammonium sulphate as precipitant. The crystals belong to the space group P6(2) (or its enantiomer P6(4)) and have unit cell dimensions of a = b = 155.9 A, c = 87.7 A, alpha = beta = 90 degrees, gamma = 120 degrees. They diffract to 2.6 A resolution. There is one dimeric molecule per asymmetric unit. Rotation function studies have revealed the orientation of the non-crystallographic 2-fold axis of the dimer in the asymmetric unit.

Affinity labeling of pig kidney 3,4-dihydroxyphenylalanine (Dopa) decarboxylase with N-(bromoacetyl)pyridoxamine 5'-phosphate. Modification of an active-site cysteine.

Pig kidney 3,4-dihydroxyphenylalanine (Dopa) decarboxylase is inactivated by N-(bromoacetyl)pyridoxamine 5'-phosphate (BAPMP) in a reaction which follows first-order kinetics at pH 7.5 and 25 degrees C. The concentration dependence of inactivation reveals saturation kinetics with an apparent Ki of 0.16 mM and kinact of 0.086 min-1 at saturating inhibitor concentration. Enzyme can be protected from inactivation by pyridoxal 5'-phosphate. Inactivation of enzyme by [14C]BAPMP proceeds with the incorporation of a stoichiometric amount of labeled inhibitor. Proteolytic digestions of the radioactively labeled enzyme followed by high-performance liquid chromatography allow the isolation of the modified peptide corresponding to the sequence Ala-Ala-Ser-Pro-Ala-Cys-Thr-Glu-Leu in which cysteine (Cys111) is the modified residue. The conservation of this residue and also of an extended region around it in all Dopa decarboxylases so far sequenced is underlined. The overall conclusion of these findings is that Cys111 may be at, or near, the pyridoxal-5'-phosphate binding site of pig kidney Dopa decarboxylase and plays a critical role in the catalytic function of the enzyme. Furthermore, fluorescence studies of BAPMP-modified apoenzyme provide useful information on the microenvironment of the affinity label at its binding site.

Aromatic L-amino acid decarboxylase-immunohistochemistry in the cat lower brainstem and midbrain.

By indirect immunohistochemistry, the present study examined the distribution of neuronal structures in the cat medulla oblongata, pons, and midbrain, showing immunoreactivity to aromatic L-amino acid decarboxylase (AADC), which catalyzes the conversion of L-3, 4-dihydroxyphenylalanine (L-DOPA) to dopamine, and 5-hydroxytryptophan to serotonin (5HT). With simultaneous and serial double immunostaining techniques, immunoreactivity to this enzyme was demonstrated in most of the catecholaminergic and serotonergic neurons. We could also demonstrate AADC-IR cell bodies that do not contain tyrosine hydroxylase (TH-) or 5HT-immunoreactivity (called "D-type cells") outside such monoaminergic cell systems. At the medullo-spinal junction, very small D-type cells were found within and beneath the ependymal layer of the 10th area of Rexed surrounding the central canal. D-type cells were localized in the caudal reticular formation, nucleus of the solitary tract, a dorsal aspect of the lateral parabrachial nucleus, and pretectal areas as have been reported in the rat. Furthermore, the present study describes, in the cat brainstem, new additional D-type cell groups that have not been reported in the rat. Dense or loose clusters of D-type cells were localized in the external edge of the laminar trigeminal nucleus, dorsal motor nucleus of the vagus, external cuneate nucleus, nucleus praepositus hypoglossi, central, pontine, and periaqueductal gray, superficial layer of the superior colliculus, and area medial to the retroflexus. D-type cells were loosely clustered in the lateral part of the central tegmental field dorsal to the substantia nigra, extending dorsally in the medial division of the posterior complex of the thalamus and medial side of the brachium of the inferior colliculus. They extended farther rostrodorsally along the medial side of the nucleus limitans and joined with the pretectal cell group. Almost all these cells were very small and ovoid to round with 1-2 short processes with the exception of dorsal motor vagal cells. AADC-IR axons were clearly identified in the vagal efferent nerves, longitudinal medullary pathway, dorsal tegmental bundle rostral to the locus coeruleus. Serotonergic axons were identified not only in the central tegmentum field and lateral side of the central superior nucleus, but also in the ventral surface of the medulla oblongata. We describe principal densely stained fiber plexuses in the cat brainstem. The findings of the present study provide a morphological basis for neurons that decarboxylate endogenous and exogenous L-DOPA, 5HTP, and other aromatic L-amino acids.

Characterization of DOPA decarboxylase mRNA in rat pheochromocytoma.

Total poly (A+) RNA has been extracted from rat pheochromocytoma and translated in vitro by means of a reticulocyte lysate system. We show that two antisera, prepared against pig kidney DOPA decarboxylase (DDC) or rat pheochromocytoma DDC, immunoprecipitate an in vitro synthetized 50 kDa polypeptide identified as DDC by competition experiments with pure DDC. The proportion of specific mRNA has been calculated and represents 0.05% of total poly A+ mRNA. Its size has been established by electrophoresis in methylmercuric hydroxide containing agarose gel, corresponding to a 2.2 kb length mRNA.

Time-resolved extrinsic fluorescence of aromatic-L-amino-acid decarboxylase.

The coenzyme-linked fluorescence of aromatic-L-amino-acid decarboxylase decays non-exponentially. The decay of both native and NaBH4 reduced samples can only be fitted by two exponentials each roughly accounting for about half of the total fluorescence. Denaturation of the reduced protein with 8 M urea makes the fluorescence decay mono-exponential, like that observed for the reference compound pyridoxamine-5-phosphate. An extra pyridoxyl moiety can be bound to the enzyme after incubation with excess pyridoxal phosphate and reduction with NaBH4. This sample is almost twice as fluorescent and shows also two lifetimes. After denaturation only one fluorescence lifetime is observed. The presence of two non-equivalent pyridoxal sites in the native enzyme can be postulated. The heterogeneous decay behaviour of the pyridoxyl moiety in the enzyme together with the variability of lifetime shown, makes this fluorophore an even more interesting fluorescent probe for proteins.

Pig kidney dopa decarboxylase: inactivation by iodoacetamide and sequence of the carboxyamidomethylcysteine-containing peptide.

Pig kidney 3,4-dihydroxyphenylalanine (Dopa) decarboxylase is inactivated by iodoacetamide following pseudo-first order reaction kinetics. The apparent first order rate constant for inactivation is proportional to the concentration of iodoacetamide and a second order rate constant of 37 M-1 min-1 is obtained at pH 6.8 and 25 degrees C. Cyanogen bromide fragmentation of iodo(1-14C)acetamide - modified inactivated Dopa decarboxylase followed by trypsin digestion yields a single radioactive peptide. Automated Edman degradation reveals a heptapeptide sequence which contains labeled carboxyamidomethylcysteine. This finding and the results of the incorporation of the label from ido (1-14C)acetamide into the enzyme clearly indicate that the modification of 1 mol of SH per mol of enzyme dimer is responsible for the inactivation process. The labeled peptide, which was located by means of limited proteolysis on the fragment corresponding to the COOH-terminal third of the enzyme, has been aligned with a 7 amino acid stretch of Drosophila enzyme. Although this region appears highly conserved in the Dopa decarboxylase enzymes, the cysteinyl residue is not conserved. This observation together with the spectral binding properties of the iodoacetamide inactivated enzyme argue against a functional role for the modifiable cysteine in the mechanism of action of pig kidney enzyme. It is suggested that the loss of pig kidney decarboxylase activity produced by iodoacetamide modification might be attributable to steric hindrance. This could be due to the presence of the bulky acetamidic group on a cysteine residue at, or near, the active center or in a site of strategic importance to the maintenance of the active site topography.

Immunohistochemistry of aromatic L-amino acid decarboxylase in the cat forebrain.

The topographic distribution of aromatic L-amino acid decarboxylase (AADC)-immunoreactive (IR) neurons was investigated in the cat hypothalamus, limbic areas, and thalamus by using specific antiserum raised against porcine kidney AADC. The perikarya and main axons were mapped on an atlas in ten cross-sectional drawings from A8 to A16 of the Horsley Clarke stereotaxic plane. AADC-IR neurons were widely distributed in the anterior brain. They were identified in the posterior hypothalamic area, rostral arcuate nucleus of the hypothalamus, dorsal hypothalamic area, and periventricular complex of the hypothalamus, which contain tyrosine hydroxylase (TH)-IR cells and are known as A11 to A14 dopaminergic cell groups. AADC-IR perikarya were also found in the other hypothalamic areas where few or no TH-IR cells have been reported: the supramamillary nucleus, tuberomamillary nucleus, pre- and anterior mamillary nuclei, caudal arcuate nucleus, dorsal hypothalamic area immediately ventral to the mamillothalamic tract, anterior hypothalamic area, area of the tuber cinereum, retrochiasmatic area, preoptic area, suprachiasmatic and dorsal chiasmatic nuclei. We also identified them in the anterior commissure nucleus, bed nucleus of the stria terminalis, stria terminalis, medial and central amygdaloid nuclei, lateral septal nucleus, and nucleus of the diagonal band of Broca. AADC-IR neurons were localized in the ventromedial part of the thalamus, lateral posterior complex, paracentral nucleus and lateral dorsal nucleus of the thalamus, medial habenula, parafascicular nucleus, subparafascicular nucleus, and periaqueductal gray. Conversely, we detected only a few AADC-IR cells in the supraoptic nucleus whose rostral portion contains TH-IR perikarya. Comments are made on the relative localizations of the AADC-IR and TH-IR neurons, on species differences between the cat and rat, as well as on the possible physiological functions of the enzyme AADC.

Two classes of sympathetic nerves with different dopa decarboxylase immunoreactivities exist in dog vas deferens.

To determine whether dihydroxyphenylalanine (DOPA) decarboxylase (DDC) activity in the terminal regions of noradrenergic axons varies with axonal length, we compared the pattern of immunohistochemical staining for DDC in the 'short' terminal nerves of dog vas deferens with that in the 'long' nerves of spleen and atrium. The terminal nerves supplying the muscular coats of the vas deferens were, like those in spleen and heart, devoid of DDC immunoreactivity. The presence of this enzyme is therefore not characteristic of either short or long noradrenergic axons, in support of previous evidence that it is a specific marker for dopaminergic terminal nerves. Many axons supplying the mucosal epithelial cells in the vas deferens were DDC-positive, suggesting the existence of a dopaminergic innervation.

Tyrosine hydroxylase-immunoreactive neurons in the human cerebral cortex: a novel catecholaminergic group?

Tyrosine hydroxylase-like immunoreactive (TH-IR) neurons with morphological features of interneurons were found throughout the human cerebral cortex. Quantitative estimates in 14 different cytoarchitectonic areas revealed a specific regional distribution pattern, neurons being less dense in primary cortical areas and denser in higher order associative areas and some limbic related areas. A partial relationship was noted between the density of labeled neurons and that of the known dopaminergic innervation. The role of the cortical TH-IR neurons in catecholaminergic function, however, remains unclear since the presence of other catecholaminergic synthesizing enzymes, dopamine-beta-hydroxylase and DOPA decarboxylase, could not be demonstrated at their level. Similar neurons have been observed transiently in the rodent cortex during development; their persistence and topographical extension in the human brain warrants further study on their possible functional role.

DNA methylating activity in murine lymphoma cells xenogenized by triazene derivatives.

We investigated whether epigenetic rather than mutational events may be involved in the induction of novel immunogenicity ("xenogenization") in murine lymphoma treated with triazene derivatives. To this end, we assessed the DNA methylation pattern of tumor cells during the course of in vivo or in vitro xenogenization by 2 triazenes, and compared it with the changes induced by the DNA hypomethylating agent 5-azacytidine (5-aza). While all of the tested agents were able to increase tumor cell immunogenicity, their effects on DNA methylating activity were opposite. In particular, the novel immunogenicity conferred by 5-aza treatment correlated well with the extent of hypomethylation induced, whereas xenogenization by triazenes was accompanied by a limited increase in DNA methylating activity. Interestingly, the antimutagenic compound quinacrine, which is known to block the xenogenizing activity of triazenes, was incapable of hypermethylating effects, whereas the hypermethylating agent cytosine arabinoside (ara-C) was apparently unable to interfere with the induction of novel immunogenicity by triazenes.

DNA methylating activity in murine lymphoma cells treated with xenogenizing chemicals.

We investigated whether epigenetic rather than mutational events might be involved in the induction of immunogenicity by the triazene derivative 1-(p-chlorophenyl)-3,3-dimethyltriazene (DM-Cl). To this purpose, we assessed the DNA methylation pattern of murine lymphoma cells xenogenized by DM-Cl and compared it with the changes induced by the DNA hypomethylating agent 5-azacytidine (5-Aza), which is also capable of affecting tumor cell immunogenicity. Both agents were found to increase the immunogenic potential of the treated tumor but according to different modalities. In particular, the novel immunogenicity conferred by 5-Aza treatment correlated well with the extent of hypomethylation induced, as opposed to what was observed for tumor xenogenization by DM-Cl.

Dopaminergic and noradrenergic sympathetic nerves of the dog have different DOPA decarboxylase activities.

We have compared the pattern of neural catecholamine fluorescence with that of immunoreactivity for the catecholamine-synthesizing enzymes tyrosine hydroxylase (TH) and DOPA decarboxylase (DDC) in dog atrium, which is innervated by noradrenergic nerves, and in dog kidney, which is thought to be supplied by dopaminergic nerves as well. In both tissues the distribution of nerves containing catecholamine fluorescence was similar to that of nerves exhibiting TH-like immunoreactivity. By contrast, DDC-like immunoreactivity was present in some (but not all) of the nerves associated with the intrarenal blood vessels, but was not detectable in any atrial nerves. High DDC activity provides further confirmation of the existence of sympathetic dopaminergic neurons supplying the kidney.