Decarboxylation of 3,4-Dihydroxyphenylalanine (DOPA) by Erythrocytes: a Reaction Promoted-By Methemoglobin and Other Ferriheme P

Proc. Nat. Acad. SCi. USA Vol. 69, No. 9, pp. 2505-2508, September 1972

Decarboxylation of 3,4-Dihydroxyphenylalanine (DOPA) by Erythrocytes: A Reaction Promoted-by Methemoglobin and Other Ferriheme Proteins (human/dopamine/Parkinson's disease/hemoglobin/myoglobin) SURESH S. TATE, JOSEPH ORLANDO, AND ALTON MEISTER Department of Biochemistry, Cornell University Medical College, New York, N.Y. 10021 Contributed by A1tot Meister, July 3, 1972 ABSTRACT The decarboxylation of DOPA by erythro- 20-24 hr after death. The minced tissue was homogenized cyte hemolysates differs from DOPA decarboxylation with 10 volumes (w/v) of ice-cold phosphate buffer containing catalyzed by aromatic aminoacid decarboxylases that con- tain vitamin B6 in several significant respects. The ability 5 mM 2-mercaptoethanol and 10,gM pyridoxal 5'-phosphate of erythrocyte hemolysates to decarboxylate DOPA is in a Potter-Elvehjem homogenizer equipped with a Teflon associated with interaction between DOPA and methemo- pestle. The homogenate was centrifuged at 12,000 X g for globin; the ferriheme protein is reduced and DOPA is de- 30 min, and the supernatant solution was used as a source of carboxylated, probably after oxidation to a quinone intermediate. An analogous reaction takes place between xvaromatic aminoacid decarboxylase. DOPA a-nd other ferriheme proteins, such as metmyo- The decarboxylation of DOPA was measured essentially as globin and cytochrome c. This phenomenon may be of described (1). The assay solution (2 ml) contained 0.05 M po- significance in relation to the side effects observed in tassium phosphate buffer (pH 7.2) containing 1 mM EDTA patients with Parkinson's disease who are treated with and 1 mM -DOPA, to which [1-14C]D-DOPA was added very large doses of DOPA. to give a final specific activity of 1.7 X 105 cpm/jimol of Recent studies in this laboratory showed that DOPA is DOPA. Incubations were usually performed for 15 min at decarboxylated by-hemolysates of human erythrocytes (1). 37°. Controls were simultaneously run containing all of the We have further investigated this phenomenon and have reagents except hemolysate or heme proteins. found that this reaction is different from that catalyzed by The activity of human kidney extracts was measured as the typical aromatic aminoacid decarboxylases found in described above, except that the assay solution also con- liver, kidney, and brain with respect to heat stability, optical tained 5 mM 2-mercaptoethanol and 10 /AM pyridoxal 5'- specificity, inhibition by typical DOPA decarboxylase in- phosphate. Controls contained heat-inactivated extracts. hibitors, and activation by pyridoxal 5'-phosphate. In the The spectra were determined with a Cary model 15 spectro- present work, evidence has been obtained that the ability of photometer equipped with a water-jacketted cuvet com- erythrocyte hemolysates to decarboxylate DOPA is associated partment. Dopamine was determined as described by Hinten- with an interaction between DOPA and methemoglobin in berger (2). Chromatographic separation of DOPA, dopamine, which methemoglobin is reduced and DOPA is decarboxyl- and other products derived from DOPA was on Whatman ated. That the reaction may be of more general significance No. 3 MM paper strips (2.5 X 55 cm) in a solvent consisting is indicated by the occurrence of an analogous reaction be- of n-butanol-acetic acid-water 25:4: 10; DOPA and dopamine tween DOPA and other ferriheme proteins. had Rf values of 0.18 and 0.51, respectively. The strips were scanned for radioactivity with a Nuclear-Chicago Actigraph EXPERIMENTAL III model 1002 strip scanner. Materials. L-DOPA and N1-(D-seryl)-N2-(2,3,4-trihy- droxybenzyl)-hydrazine (Ro 4-4602) were gifts from Hoff- Preparation of [1-14C]D-DOPA. [1-14C]D-DOPA was pre- mann-La Roche and D-a-hydrazinomethyl-3,4-dihydroxy- pared from [1-'4C]D-DOPA by decarboxylation of the phenylalanine (MK 486) was, a gift from Merck. D-DOPA, L-enantiomer with mouse-liver homogenate (1). The homog- dopamine, [3H]-DOPA (generally labeled), sperm-whale enate was centrifuged at 12,000 X g for 30 min, and the myoglobin, horse-heart cytochrome c, and horse-radish supernatant solution was dialyzed for 18 hr against 200 vol- peroxidase were purchased from Schwarz-Mann. Human umes of phosphate buffer containing 5 mM 2-mercapto- methemoglobin, and beef-liver catalase were obtained from ethanol and 10 /AM pyridoxal 5'-phosphate. Decarboxylation Sigma. [1-'4C] and [2-14C]DL-DOPA, [1-14C]-tyrosine, and was performed in a standard Warburg flask. The side- generally labeled [14C]-histidine were purchased from New arm contained 0.2 ml of 1 N HCl, and the center well was England Nuclear Corp. Hog-kidney D-aminoacid oxidase was supplied with 0.1 ml of 1 N KOH. The reaction mixture a gift from Dr. Daniel Wellner. (2 ml) contained 1.8 ml of phosphate buffer containing 5 mM 5'-phosphate, 0.2 mg of human erythrocytes were pre- 2-mercaptoethanol, 10 &IM pyridoxal Methods. Hemolysates of [1-"4C]DL-DOPA, and 0.2 ml of dialyzed extract from pared as described (1), except that 0.05 M potassium phos- The flask was and incubated at 370 for buffer mM EDTA (referred to below as mouse liver. stoppered phate (pH 7.2)-i 90 min. The reaction was terminated by tipping in HCl, and "phosphate buffer") was used. The hemolysate was centri- then on a shaker for for 15 min at and the supernatant the flask was placed reciprocating fuged at 8000 X g 40, 60 min. An aliquot of the KOH solution was mixed with liquid solution was used in the studies described here. scintillation fluid and counted. About 48% of the total Samples of human kidney cortex were obtained at autopsy radioactivity added as [14C]DL-DOPA was found as 14CO2; 2505 Downloaded by guest on October 1, 2021 2506 Biochemistry: Tate et al. Proc. Nat. Acad. Sci. USA 69 (1972)

TABLE 1. Decarboxylation of - and i-DOPA by human the [1-14C]iDOPA in the solution thus obtained was 1.45 X erythrocyte hemolysates and kidney extracts 105 cpm/,umol. Decarboxylase activity RESULTS Enzyme source -DOPA D-DOPA Characteristics of the DOPA Decarboxylase Activity of Human Erythrocytes. The ability of hemolysates of human erythro- Human erythrocyte hemolysate 7.1 7.6 cytes to decarboxylate DOPA was relatively resistant to Human kidney extract 41.7 <1 heating at 1000. Thus, although the DOPA decarboxylase activity of human kidney extracts was completely abolished Extracts were prepared and assayed as described in the text. for 1 treat- Activity is expressed as nmol of DOPA decarboxylated per by incubation of the extracts at 1000 min, similar 15 min per ml of hemolysate or per mg of kidney-extract protein. ment of erythrocyte hemolysates destroyed only 40-70% of the initial activity. Addition of pyridoxal 5'-phosphate(10- thus, virtually all of the [14C]i-DOPA was decarboxylated. 100 MM) had no effect on the decarboxylation of DOPA by The reaction mixture was placed in a centrifuge tube, which erythrocyte hemolysates; on the other hand, the aromatic was immersed in a boiling-water bath for 1 min. After the L-aminoacid decarboxylases from other sources, e.g., rat tube was cooled and centrifuged, 2 ml of the clear supernatant liver (3), rat kidney (4), bovine adrenal medulla (5), and solution was withdrawn and D-DOPA was added to yield a hog kidney (6), are significantly stimulated by pyridoxal specific activity of the 5'-phosphate. Prior incubation of human kidney extracts final concentration of 0.01 M. The Ro [1-14C]D-DOPA in the solution thus obtained was 1.08 X 105 with the aromatic L-aminoacid decarboxylase inhibitors 4-4602 and MK 486 (final concentration, 10 MM; see ref. 7 for cpm/MUmol. abolished Preparation of [1-14C]L-DOPA. [1-14C]DL-DOPA (0.3 mg) a review of these and other inhibitors) completely DOPA decarboxylase activity; such treatment, however, had was dissolved in 0.4 ml of 0.01 N HCl; 0.3 ml of 0.1 M sodium hemol- pyrophosphate buffer (pH 7.2, containing 1 Mumol of 2- no significant effect on the activity of erythrocyte ysates. mercaptoethanol) was added and the pH was adjusted to decar- A mixture (0.3 ml) containing 0.5 mg of hog-kidney D- As indicated in Table 1, erythrocyte hemolysates 7.2. i- at about the same rate, but aminoacid oxidase, 10 ug of FAD, and 3400 units of catalase boxylated and D-DOPA human kidney extract exhibited activity only toward L- was incubated at 370 for 3 hr. The pH of the solution was the de- and after DOPA. Erythrocyte hemolysates did not promote then adjusted to 5.5 by addition of acetic acid, or addition of 5 Mmol of 2-mercaptoethanol, the solution was carboxylation of histidine, tyrosine, 5-hydroxytryptophan. immersed in a boiling-water bath for 1 min. The denatured Decarboxylation of DOPA by Hemoproteins. The experi- protein was centrifuged off, and the supernatant solution was ments described above indicate that the decarboxylation of then applied to a Dowex-1 (acetate) column (0.5 X 2 cm); DOPA by erythrocyte hemolysates differs markedly from L-DOPA was eluted with 3 ml of water. Sufficient L-DOPA that catalyzed by tissue preparations containing aromatic was added to the eluate to yield a final concentration of aminoacid decarboxylase activity. When erythrocyte hemol- 0.01 M. The eluate contained about 54%O of the radioactivity ysates were fractionated by addition of increasing concentra- originally added as [1-14C]DiDOPA. The specific activity of tions of ammonium sulfate, it was found that the fractions that promoted the decarboxylation of DOPA were the same TABLE 2. Decarboxylation of L-DOPA by hemoglobin, myo- as those that contained hemoglobin. When the hemolysates globin, and other heme-containing compounds were treated with potassium ferricyanide to convert the

Exp. DOPA decarboxylation no. Heme compound (mg/ml)* nmol/15 min 1 Erythrocyte hemolysate (1.5) 0.20 2 Erythrocyte hemolysate (0.75)t 4.2 3 Methemoglobin (1.0) 11.2 4 Methemoglobin (2.0) 15.6 5 Methemoglobin (2. 0)t 2.4 ;z 6 Methemoglobin (2. 0)§ 1.5 7 Metmyoglobin (2.0) 14.0 cn 8 Metmyoglobin (2.0)§ 1.2 cK 9 Cytochrome c (0.5) 12.0 10 Peroxidase (2.0) 2.6

* The assay solution (final volume, 2 ml) contained phosphate buffer (pH 7.2), L-DOPA, and heme compound as indicated, incubated for 15 min at 37°. Other details are given in the text. t The hemolysate (1 ml) was treated with 1 ml of 0.01 M 500 550 600 650 potassium ferricyanide. After standing at 250 for 15 min, the WAVELENGTH, nm solution was dialyzed exhaustively against phosphate buffer. FIG. 1. Effect of L-DOPA on metmyoglobin. The solutions $ The ferriheme protein was treated with 5 mM 2-mercapto- contained phosphate buffer (pH 7.2) and 1.5 mg/ml of sperm- ethanol, then dialyzed exhaustively against phosphate buffer. whale metmyoglobin (Curve 1). Curve 2 gives the spectrum after § The ferriheme protein was treated with 0.25 mM dithionite, metmyoglobin was incubated for 30 min at 370 in air with 1 mM then dialyzed exhaustively against phosphate buffer. i-DOPA. Downloaded by guest on October 1, 2021 Proc. Nat. Acad. Sci. USA 69 (1972) Decarboxylation of DOPA by Erythrocytes 2507 hemoglobin to methemoglobin, and thoroughly dialyzed, there was about a 40-fold stimulation of DOPA decarboxylation (Table 2, Exp. 2). A preparation of commercially availa- E able human methemoglobin promoted the decarboxylation c of DOPA at a rate that was about 80 times greater than that observed with the hemolysate (Table 2, Expts. 3 and 4). Treatment of methemoglobin with 2-mercaptoethanol (Exp. 0 5) or with dithionite (Exp. 6) decreased decarboxylation greatly. Sperm-whale myoglobin, shown by spectral analysis to be primarily metmyoglobin, also promoted the decarboxylation of DOPA at a rate similar to that observed with methemoglobin (Exp. 7); treatment with dithionite also greatly diminished this effect (Exp. 8). Other hemoproteins tested, such as horse-heart cytochrome c and peroxidase also promoted DOPA decarboxylation. MINUTES Fig. 1 shows the effect of L-DOPA on the absorption spec- FIG. 3. Correlation between the reduction of metmyoglobin by trum of metmyoglobin. Curve 1 gives the spectrum of met- DOPA and the decarboxylation of DOPA. The solutions con- myoglobin; after incubation of metmyoglobin with DOPA tained (final volume, 1 ml) phosphate buffer, 1.5 mg (84 nmol) of in air, a spectrum characteristic of oxymyoglobin was ob- sperm-whale metmyoglobin, and 1 jmol of either iDOPA (-) or tained (Curve 2). The findings indicate that metmyoglobin is [1-14C]ILDOPA (0). The absorption spectra were recorded at 370 reduced to the ferrous form, which reacts rapidly with oxygen. with a Cary model 15 spectrophotometer. The concentration of Human methemoglobin gave substantially the same results metmyoglobin was computed from absorbance changes at 580 nm. in studies performed under similar conditions. In the experiment described in Fig. 2, ferricytochrome c (Curve 1) was 4 mol of iDOPA were decarboxylated per mol of methe- incubated with DOPA; the spectrum obtained (Curve 2) is moglobin, or about 1 mol of L-DOPA per mol of heme. characteristic of ferrocytochrome c. In the experiment de- These findings indicate that there is equivalence between the scribed in Fig. 3, sperm-whale metmyoglobin was incubated reduction of the ferriheme proteins and the amount of DOPA with L-DOPA; the reduction of metmyoglobin to the ferrous decarboxylated. The effect of ferriheme protein concentra- form and the decarboxylation of DOPA were determined. tion on the rate of DOPA decarboxylation was examined Under these conditions, there was a slow decarboxylation of (Fig. 4); the rate of decarboxylation increased with increased DOPA in the absence of metmyoglobin; this occurred at protein concentration, then leveled off. The findings suggest about 15% of the rate found in the presence of metmyoglobin. that the decarboxylation may occur in at least two steps, as The data, thus, indicate that the presence of metmyoglobin discussed below. increased the rate of DOPA decarboxylation by about 6-fold. We found that the decarboxylation of DOPA by metmyo- As indicated in Fig. 3, the extent of the rapid decarboxylation globin and methemoglobin is markedly inhibited by i- produced by metmyoglobin is about equivalent, within experi- ascorbate. Thus, as indicated by the data given in Table 3, mental error, to the conversion of metmyoglobin to the 0.02 mM L-ascorbate inhibited decarboxylation by more than ferrous form. Similar results were obtained in experiments 70%, and 0.1 mAM L-ascorbate decreased decarboxylation by with 1.0 and 2.0 mg of metmyoglobin. In these experiments, about 97%. It seems unlikely that the inhibition of DOPA formation of CO2 was slower than the reduction of metmyo- decarboxylation by ascorbate under these conditions is due to globin, suggesting that an intermediate is formed that is de- reduction of ferriheme protein to ferroheme protein; thus, carboxylated in a subsequent step. Analogous experiments 0.1 mM iLascorbate reduced only 10% of the metmyoglobin with human methemoglobin led to similar results; thus, about under the conditions used in the experiments described in Table 3.

.E C'J0

E 6d Lii-J

x Lii0 U 00c

C1Ld

2 3 4 5 METHEMOGLOBIN OR METMYOGLOBIN 500 550 600 (mg/ml ) WAVELENGTH, nm FIG. 4. Effect of methemoglobin and metmyoglobin concen- FIG. 2. Effect of L,-DOPA on cytochrome c. The solutions trations on the decarboxylation of DOPA. The assay solutions contained phosphate buffer (pH 7.2) and 0.48 mg per ml of ferri- contained (final volume, 1 ml), phosphate buffer, 1 mM [1- cytochrome c (Curve 1). Curve 2 gives the spectrum after cyto- 14C] LDOPA, and human methemoglobin or sperm-whale chrome c was incubated for 10 min at 37° with 1 mM iDOPA. metmyoglobin. Downloaded by guest on October 1, 2021 2508 Biochemistry: Tate et al. Proc. Nat. Acad. Sci. USA 69 (1972)

TABLE 3. Effect of ascorbate on the decarboxylation of DOPA companied by oxidation of DOPA to an intermediate, which by metmyoglobin might be a quinone or semiquinone that can undergo decarboxylation in a subsequent step. Such an intermediate would Decarboxylation of DOPA be expected to be reduced by ascorbate; a reaction of this L-Ascorbate (mM) (nmol/mg per 10 min) type may explain the observed inhibition of decarboxylation 0 9.10 by low concentrations of ascorbate.* 0.01 6.91 We reported (1) a substantial decrease in DOPA-decar- 0.02 2.60 boxylase activity of erythrocytes obtained from patients 0.05 0.25 with Parkinson's disease who received oral DOPA therapy. 0.10* 0.31 As stated previously, none of the patients studied was anemic; therefore, it seems improbable that these observations reflect Reaction mixtures (1 ml) contained 0.05 M potassium phos- a decrease in the total concentration of blood hemoglobin. phate buffer (pH 7.2), 1 mM EDTA, 1 mM [1-'4C]-DOPA, In the light of the present observations, the decreased DOPA 1.5 mg of sperm-whale metmyoglobin, and iascorbate as indi- decarboxylase activity of the erythrocytes of treated patients cated, incubated for 10 min at 37°. previously observed may probably be ascribed to a decrease * Under these conditions, but in the absence of DOPA, about normally present metmyoglobin was reduced. in the small amount of methemoglobin 10% of the (10) by reduction by the administered DOPA. While the concentrations of methemoglobin and of metmyoglobin are prob- Products of the Reaction. Dopamine has been found in ably normally quite low in vivo, the presence of even small normal adult human erythrocytes (8) and also in the erythro- amounts of these ferriheme proteins might be of considerable cytes of patients with Parkinson's disease that receive significance in relation to the production of dopamine in therapy with L-DOPA (2). As noted earlier, human-blood patients receiving very large doses of DOPA for the therapy serum does not exhibit DOPA decarboxylase activity (1). of the Parkinson's disease. The finding that ferricytochrome In the present work, by procedures similar to those used by c can be readily reduced by DOPA must also be considered in Hinterberger (2), we have confirmed the observation of relation to the effects of therapy with massive doses of L- Hinterberger (2) that [3H]dopamine is formed when [3H]- DOPA; such patients may have abnormalities of oxidative DOPA is incubated with human erythrocyte hemolysates. metabolism. These considerations, which require additional When commercial preparations of methemoglobin and met- study, may be of significance in relation to some of the side myoglobin were incubated with [2-14C]DL-DOPA, only traces effects of therapy with large oral doses of DOPA (11). of dopamine could be detected by chromatography. Studies This work was supported in part by the American Parkinson's on experiments incubated for 1-4 hr at 370 showed the Disease Association. formation of melanin-like pigments that remained at the 1. Tate, S. S., Sweet, R., McDowell, F. H. & Meister, A. origin of the paper chromatograms; incubation of dopamine (1971) Proc. Nat. Acad. Sci. USA 68, 2121-2123. 2. Hinterberger, H. (1971) Biochem. Med. 5, 412-424. itself with metmyoglobin also led to the formation of such 3. Awapara, J., Sandman, R. P. & Hanly, C. (1962) Arch. pigments. The findings indicate that while some dopamine Biochem. Biophys. 98, 520-525. is formed in this reaction, additional chemical transformation 4. Buzard, J. A. & Nytch, P. (1959) J. Biol. Chem. 234, 884- occurs that is associated with polymer formation. These 888. observations are reminiscent of those of Green et al (9), 5. Fellman, J. H. (1959) Enzymologia 20, 366-376. 6. Christenson, J. G., Dairman, W. & Udenfriend, S. (1970) who showed that the reduction of methemoglobin and of Arch. Biochem. Biophys. 141, 356-367. ferricytochrome c by epinephrine is associated with the forma- 7. Porter, C. C. (1971) Fed. Proc. 30, 871-876. tion of adrenochrome, which undergoes additional trans- 8. Bryson, G. & Bischoff, F. (1970) Clin. Chem. 16, 312-317. formation to yield melanin-like pigments. The present findings 9. Green, S., Mazur, A. & Shorr, E. (1956) J Biol. Chem. 220, 237-255. on the interaction of ferriheme proteins with DOPA do not 10. Van Slyke, D. D., Hiller, A., Weisiger, J. R. & Gruz, W. 0. exclude the possible occurrence of additional oxidative reac- (1946) J. Biol. Chemn. 166, 121-148. tions of DOPA and dopamine. 11. Barbeau, A. & McDowell, F. H. (1970) L-DOPA and Parkinsonism (F. A. Davis Co., Philadelphia, Pa.). DISCUSSION 12. Dairman, W. & Christenson, J. G. (1972) Fed. Proc. 31, Abstr. No. 2114, p. 590. The studies reported here, which confirm the earlier finding 13. Yamabe, H. & Lovenberg, W. (1972) Biochem. Biophys. in this laboratory that hemolysates of human erythrocytes Res. Commun. 47, 733-739. decarboxylate DOPA (1), indicate that this decarboxylation is unusual in that it is not catalyzed by a pyridoxal 5'-phos- * Dairman and Christenson (12), and Yamabe and Lovenberg phate-dependent enzyme. The DOPA-decarboxylating ac- (13) have recently confirmed the observation (1) that erythrocyte tivity of erythrocytes is atypical with respect to heat sta- preparations catalyze the decarboxylation of DOPA. The latter bility, activation by pyridoxal 5'-phosphate, inhibition by studies, which led to the "inescapable conclusion that oxyhemo- certain inhibitors, and optical specificity. The findings indi- globin reacts with DOPA, resulting in the release of the carboxyl cate that the decarboxylation of DOPA is due to methe- group" and to the conclusion that methemoglobin is inactive in decarboxylating DOPA, were performed in the presence of moglobin present. Thus, treatment with ferricyanide led to 0.117 mM ascorbic acid (13) which, as shown in the present work, a marked increase in decarboxylase activity, and reduction produces substantial inhibition of DOPA decarboxylation. with 2-mercaptoethanol or dithionite led to marked decreases. (Table 3). In the cited studies (13), the rate of DOPA decarboxyl- The spectral studies reported here indicate that the decar- ation by oxyhemoglobin was 0.12 nmol per mg per 15 min, or boxylation of DOPA is associated with reduction of the ferri- about 2 orders of magnitude lower than we observed in- the heme protein to ferroheme protein. Presumably, this is ac- present studies with methemoglobin. Downloaded by guest on October 1, 2021