- Email: [email protected]
Resistance of Aryl-Substituted Phenylalanines to Decarboxylation in the Rat
ISSN 0306-3623/98 $19.00 1 .00 PII S0306-3623(98)00017-2 All rights reserved
Gen. Pharmac. Vol. 31, No. 3, pp. 437–440, 1998 Copyright 1998 Elsevier Science Inc. Printed in the USA.
Resistance of Aryl-Substituted Phenylalanines to Decarboxylation in the Rat Brian L. Goodwin, Colin R. J. Ruthven and Merton Sandler* Department of Chemical Pathology, Queen Charlotte’s and Chelsea Hospital, Goldhawk Road, London, W6 0XG, United Kingdom ABSTRACT. 1. The in vivo decarboxylation of analogues of DL-phenylalanine with methoxy, ethoxy or methylenedioxy substituents on the aromatic nucleus was assessed in the rat by measuring urinary excretion of the corresponding phenylacetic acids and b-phenylethylamines. Only trace amounts of these amines and acids were excreted. 2. o-Tyrosine was readily decarboxylated, a property that may be attributed to the position and nature of the substituent on the aromatic nucleus. 3. Some phenylpyruvic acids derived from these amino acids exhibited a certain degree of decarboxylation to phenylacetic acids or reduction to phenyllactic acids in the rat. gen pharmac 31;3:437–440, 1998. 1998 Elsevier Science Inc. KEY WORDS. Rat, aryl-substituted phenylalanines, phenylpyruvic acids, b-phenylethylamines, phenyllactic acids, phenylacetic acids
INTRODUCTION Because of their central role in mammalian biochemistry and physiology, the metabolic pathways for the l-isomers of phenylalanine, tyrosine and dopa have received extensive study. Despite their structural similarities, the main metabolic pathway for each is different, largely because of the initial metabolic step. Phenylalanine is rapidly hydroxylated in the para position, but it is not so readily decarboxylated [for a review, see Blau (1979)]. Transamination readily occurs only at high substrate concentration, as found in untreated phenylketonuria (Fo¨lling, 1934). In contrast, tyrosine undergoes almost quantitative transamination and dopa is extensively decarboxylated. Some analogues of tyrosine are readily metabolized, and phenolic acetic acids are prominent among the urinary metabolites. Goodwin et al. (1978) investigated the metabolism of dopa, m-tyrosine and 3-hydroxy-4-methoxy- and 4-hydroxy-3-methoxytyrosine in the rat and, with the exception of the last compound, a preponderance of decarboxylated metabolites was detected. These investigations were therefore extended to an examination of the relation between metabolic pathway and structure in a further range of phenylalanine analogues, in particular to compounds with alkoxy substituents on the aryl nucleus without a phenolic group. Evidence was sought in particular for the formation of products of decarboxylation relative to transamination-type metabolites (formed by transamination of l-amino acids or by oxidative deamination of d-amino acids), because results with phenylalanine, which also lacks a phenolic hydroxyl group, might suggest that the decarboxylation pathway is unlikely to be of major significance without maintained high concentrations of substrate (Blau et al., 1976). MATERIALS AND METHODS
Chemicals Amino acids (dl mixtures) were synthesized from commercially available benzaldehydes by the standard azlactone synthesis (Sealock et *To whom correspondence should be sent, at Department of Pharmacy, King’s College, Manresa Road, London, SW3 6LX, UK. Received 8 October 1997; accepted 15 December 1997.
al., 1951), whereby benzaldehydes and acetylglycine are condensed (in acetic anhydride), with the use of anhydrous sodium acetate as catalyst. The azlactones so formed were reduced in water with sodium amalgam, the azlactone ring being hydrolyzed under the alkaline conditions used, and the amino acids were then released by acid hydrolysis. No attempt was made to resolve these dl mixtures. Amines were prepared by the method of Brossi et al. (1966). The corresponding benzaldehydes were condensed with nitromethane, followed by reduction of the resultant nitrostyrenes with lithium aluminum hydride. Substituted phenylacetic acids were synthesized from the corresponding acetophenones by the modified Willgerodt reaction (Schwenk, 1942; Vogel, 1961) in which thiomorpholides of phenylacetic acids were prepared by refluxing the corresponding acetophenones with sulfur and morpholine. The free acids were released by acid hydrolysis. Pyruvic acids were prepared from azlactones by acid hydrolysis and the lactic acids by the reduction of the pyruvic acids with sodium amalgam in water (Shaw et al., 1956). Structures were confirmed by mass spectrometry.
Animal experiments Amino acids were administered intraperitoneally to female white Wistar rats (150–200 g) at a dose of 100 mg/kg. Where solubility was a problem, they were administered as hydrochloride salts. Pyruvic acids were administered in a similar manner, dissolved in sodium bicarbonate solution. The animals were kept in metabolism cages for 24 hr without food but with water supplied ad libitum. Urine was collected into 0.5 ml of 6 M HCl and stored at 2208 C.
Assay methods Phenolic acid metabolites and lactic acids were extracted from urine and converted into their silyl derivatives prior to gas chromatography (GC) (Goodwin and Sandler, 1975; Goodwin et al., 1974). Nonphenolic phenylacetic acids were measured by the GC method
438
B. L. Goodwin et al.
TABLE 1. Excretion of metabolites after administration of substituted phenylalanines to rats Substituent on benzene ring o-OH o-MeO m-MeO p-MeO 3-EtO-4-OH 2,3-diMeO 2,4-diMeO 2,5-diMeO 3,4-diMeO
Corresponding amine 8.6 (5.4–10.8) 0.025 (0.02–0.025) 0.04 (0.025–0.06) 0.05 (0.04–0.06) — 0.03 (0.03–0.04) 0.22 (0.15–0.33) 0.015 (0.0–0.025) 0.14 (0.08–0.22)
Corresponding phenylacetic acid
Corresponding phenyllactic acid
Corresponding phenylpyruvic acid
13 (8–16)* 0.5 (0.4–0.65) 0.25 (0.2–0.3) 1.3 (0.85–2.0) 0M 0.2 (0.15–0.25) TrM 0.2 (0.0–0.45) 3.0M (2.1–3.9)
0.4 (0.15–0.7) 0 0M 0M M 1.6 (0.95–2.1) — — 0M 7.2M (4.4–9.9)
— 2.4 (2.1–2.8) 0 — — 1.3M (0.9–1.5) — 3.8 (2.4–5.0) 10.0 (6.6–12.1)
3,4-Methylenedioxy
0
0M
0M
3,4,5-triMeO
—
0.25 (0.15–0.45)
TrM
Other metabolites
3,4-Dihydroxyphenylacetic acid 0 4-Hydroxy-3-methoxyphenylacetic acid 0.7 (0.45–1.0) 3,4-Dihydroxyphenylacetic acid 0M 4-Hydroxy-3-methoxyphenylacetic acid 0.7 (0.3–0.9) 3-Methoxytyramine 0.75 (0.6–1.0)
Abbreviation: Tr, trace. Results are expressed as a percentage of the dose excreted in 24 hr. M Confirmed by mass spectrometry. * Corrected for endogenous compound.
developed for phenylacetic acid (Goodwin et al., 1975), after acid hydrolysis to release free acids from their putative conjugates, but chromatographed at a higher oven temperature to compensate for their lower volatility. Putative sulfate conjugates of phenolic amines were hydrolyzed at 1008C and pH 1 for 20 min, and putative acetyl or similar conjugates of all other amine metabolites were hydrolyzed with 6 M HCl at 1008C for 2 hr (an extensive literature for analogous amines suggests that these are their major expected conjugates). Because the amines sought in this study are relatively nonpolar, the pH was adjusted to 10, and free amines were extracted into ethyl acetate. After evaporation of solvent, pentafluoropropionyl derivatives were prepared by warming with pentafluoropropionic anhydride. Phenylpyruvic acids were not amenable to GC (their silyl derivatives tailed badly). They were therefore reduced to the corresponding lactic acids prior to assay. An aliquot of urine (0.5 ml) was mixed with 0.1 ml 10 M NaOH, and 7–10 mg sodium borohydride was added. After about 1 hr, the mixture was acidified and the lactic acid content was assessed as heretofore described. The quantity of the pyruvic acid sought was assessed by difference. All assays were carried out in parallel with urine specimens to which measured quantities of the compounds sought had been added prior to processing; pyruvic acids (rather than lactic acids) were added to urine aliquots and then reduced as heretofore described. RESULTS AND DISCUSSION The most striking finding (Table 1) was the dearth of urinary metabolites found after administration of nonphenolic phenylalanines. The pattern of metabolites obtained with o-tyrosine, however, showed a pattern similar to that observed with m-tyrosine (Goodwin et al., 1978); an appreciable quantity of the decarboxylated metabolite o-tyramine and its further metabolite, o-hydroxyphenylacetic acid,
were found, in confirmation of the work of Blaschko (1949, 1950) and others, whereas only traces of the expected decarboxylated metabolites were detected for most of the remaining amino acids. One further amino acid tested, 3-ethoxy-4-hydroxyphenylalanine, carried a phenolic hydroxyl group; in this compound, the hydroxyl group is masked by an adjacent ethoxy group. The finding that this amino acid yields only small amounts of acidic metabolites mimics the results obtained with 3-methoxytyrosine in which the hydroxyl group also is para to the side chain. Although the presence of a hydroxyl group on the aromatic nucleus may be a prerequisite for efficient decarboxylation in mammals, this reaction is not quantitatively significant if the sole hydroxyl group is para to the side chain. Tyrosine, for instance, is highly resistant to decarboxylation in mammals [e.g., Blaschko (1950)]. This contrasts with the action of gut flora enzymes [e.g., Gale (1940)], which can affect decarboxylation. It is uncertain whether the effects of masking a hydroxyl group by an ortho substituent affects the rate or extent of decarboxylation; the results presented here and previously (Goodwin et al., 1978) suggest that the positional effect is more significant; for instance, 3-hydroxy4-methoxyphenylalanine and o-tyrosine both have a hydroxyl group ortho to a potential masking group, yet are readily decarboxylated, unlike 3-methoxytyrosine. The observation that the nonphenolic phenylalanines in general appear to be poorly metabolized by the expected pathways may be an indirect consequence of the overall biochemical economy of the mammalian organism, in which phenylalanine (itself lacking a phenolic hydroxyl group) is decarboxylated at normal physiological concentrations only as a minor pathway but undergoes extensive p-hydroxylation to tyrosine [for a review, see Goodwin (1979)] to dispose of the excess dietary phenylalanine ingested. The results reported here are in marked contrast with the chemical decarboxylation of tyrosine and its analogues. On heating to about
Phenylalanine Analogues in Rats
439
TABLE 2. Excretion of metabolites after administration of substituted phenylpyruvic acids to rats Substituent on benzene ring
Corresponding phenylacetic acid
Corresponding phenyllactic acid
o-MeO m-MeO p-MeO 3,4-diMeO 3,4,5-triMeO
2.0 (0.45–4.4) 1.9M (1.5–2.4) 29M (21–33) 0M 1.5 (0.8–2.7)
0M 5.0M (2.7–7.4) TrM 11.5M (4.2–22.0) 2.0M (0.5–5.0)
Other metabolites o-Hydroxyphenyllactic acid 0M m-Hydroxyphenylacetic acid 0.55M (0.2–0.95) 4-Hydroxy-3-methoxyphenylacetic acid 0.5M (0.2–0.85)
Abbreviation: Tr, trace. Results are expressed as a percentage of the dose excreted in 24 hr. M Confirmed by mass spectrometry.
3008C, a range of p-hydroxyphenylalanines are decarboxylated (unpublished results) in the same manner as tyrosine (Waser, 1925), providing an efficient method for preparing the amines. Yet m-tyrosine, which is efficiently decarboxylated in vivo, did not yield m-tyramine on heating under these conditions. This suggests that the reaction mechanisms at a molecular level are different. Tyrosine, which carries a hydroxyl group in the para position and is therefore not amenable to the same hydroxylation reaction as phenylalanine and is not appreciably decarboxylated, is catabolized by nearly quantitative transamination (Cammarata and Cohen, 1950) to yield the corresponding pyruvic acid, which is then reduced to p-hydroxyphenyllactic acid as a subsidiary reaction. These acids are not normally excreted to any significant extent, because most of the p-hydroxyphenylpyruvic acid is further metabolized to homogentisic acid prior to oxidation to CO2 and water [for a review, see Goodwin (1972)]. However, pyruvic acids formed from the nonphenolic amino acids would not be expected to undergo further oxidation in that p-hydroxyphenylpyruvate oxidase exhibits a high degree of specificity. Except after administration of 3,4-dimethoxyphenylalanine, the corresponding lactic acids were not detected at more than a small extent. A possible explanation would be that the corresponding pyruvic acids are not readily reduced in vivo. After the administration of several of these pyruvic acids, only a small proportion of the administered dose was excreted as corresponding lactic acids (Table 2). A possible explanation is that the pyruvic acids may have been transaminated back to the amino acids. The urinary excretion of 3,4-dimethoxyphenyllactic and -pyruvic acids after administration of the corresponding amino acid indicates that this amino acid undergoes significant transamination followed by reduction of the pyruvic acid. Decarboxylation of phenylpyruvic acid itself is a well-established metabolic pathway, at least in microorganisms [e.g., see Gopalakrishna (1976)], that may produce phenylacetic acid in the gut, but except in the case of p-methoxyphenylpyruvic acid, little of the corresponding phenylacetic acids were excreted. The relative absence of urinary metabolites of the nonphenolic amino acids does not necessarily imply that they are metabolically inert. This may be illustrated by the fact that tyrosine, under abnormal circumstances, can give rise to massive amounts of urinary metabolites—for instance, in neonatal tyrosinemia (Bloxham et al., 1960), similarly, in untreated phenylketonuria, there is extensive urinary excretion of phenylalanine metabolites (Armstrong et al., 1955; Fo¨lling, 1934; Woolf, 1951) in consequence of defects in the enzyme systems that normally bring about conversion of these amino acids into CO2 and water. However, only small (but readily detectable) amounts of these metabolites are excreted by a normal person. O-Dealkylation of the compounds assessed in this study may occur,
but in general, such metabolites were not sought (many of the expected metabolites arising from dealkylation occur normally in urine, often as plant natural product metabolites). Partial or total removal of the amino acid side chain may occur (Goodwin et al., 1972), but, in mammals, this has been reported only for phenylserines and not for phenylalanines. CONCLUSION Most of the nonphenolic-substituted phenylalanines examined in this study were resistant to decarboxylation. Evidence is presented that transamination of the amino acid l-isomers or oxidation of d-isomers or both to yield phenylpyruvic acids usually accounted for only a small proportion of the administered dose. o-Tyrosine was decarboxylated. Most of the phenylpyruvic acids studied, with the exception of 3,4-dimethoxyphenylpyruvic acid, were poorly reduced to phenyllactic acids. With the exception of p-methoxyphenylpyruvic acid, they were resistant to decarboxylation. SUMMARY Previously published results have demonstrated extensive variations in the preferred metabolic pathways for phenylalanine and arylhydroxylated phenylalanines associated with the normal economy of the mammalian organism. Evidence for the preferred metabolic pathways for a series of synthetic analogues in the rat was therefore sought. Decarboxylation of the nonphenolic amino acids was inefficient, whereas o-tyrosine, in common with l-dopa and m-tyrosine, was readily decarboxylated. This decarboxylation also distinguishes it from p-tyrosine and 3-methoxytyrosine, which are poorly decarboxylated in the rat. Transamination or oxidative deamination of the L- and D-isomers, respectively, or both, as evidenced by excretion of the corresponding pyruvic and lactic acids, was, in general, not demonstrated to be an efficient metabolic process under the experimental conditions used. We thank the Mass Spectrometry Unit at Queen Charlotte’s and Chelsea Hospital for confirming the identity both of pure compounds and urinary metabolites.
References Armstrong M. D., Shaw K. N. F. and Robinson K. S. (1955) Studies on phenylketonuria II: the excretion of o-hydroxyphenylacetic acid in phenylketonuria. J. Biol Chem. 213, 797–804. Blaschko H. (1949) The decarboxylation of o-hydroxyphenylalanine. Biochem. J. 44, 268–274. Blaschko H. (1950) Substrate specificity of amino-acid decarboxylases. Biochim. Biophys. Acta 4, 130–137. Blau K. (1979) Phenylalanine hydroxylase deficiency: biochemical, physiological, and clinical aspects of phenylketonuria and related phenyl-
440 alaninemias. In Aromatic Amino Acid Hydroxylase and Mental Disease (Edited by Youdim M. B. H.), p. 77–139. Wiley, Chichester, UK. Blau K., Goodwin B., Woods H. F. and Youdim M. B. H. (1976) Phenylethylamine formation in perfused rat liver. Br. J. Pharmac. 58, 474P. Bloxham H. R., Day M. G., Gibbs N. K. and Woolf L. I. (1960) An inborn defect in the metabolism of tyrosine in infants on a normal diet. Biochem J. 77, 320–326. Brossi A., Schenker F., Schmidt R., Banziger R. and Leimgruber W. (1966) Synthesen in der Isochinolinreihe. Helv. Chim. Acta 49, 403–411. Cammarata P. S. and Cohen P. P. (1950) The scope of the transamination reaction in animal tissues. J. Biol. Chem. 187, 439–452. ¨ ber Ausscheidung von Phenylbrenztraubensa¨ure in den Fo¨lling A. (1934) U Harn als Stoffwechselanomalie in Verbindung mit Imbezillita¨t. HoppeSeyler’s Z. Physiol. Chem. 227, 169–176. Gale E. F. (1940) The production of amines by bacteria 2: the production of tyramine by Streptococcus faecalis. Biochem. J. 44, 846–852. Goodwin B. L. (1972) Tyrosine Catabolism. Clarendon Press, Oxford. Goodwin B. L. (1979) Phenylalanine hydroxylase. In Aromatic Amino Acid Hydroxylase and Mental Disease (Edited by Youdim M. B. H.), pp. 5–76. Wiley, Chichester, UK. Goodwin B. L. and Sandler M. (1975) Improved resolution using capillary columns. Clin. Chim. Acta 59, 253–254. Goodwin B. L., Johnson R. D., Leask B. G. S., Ruthven C. R. J. and Sandler M. (1972) threo-3,4,-Dihydroxyphenylserine, a poor source of noradrenaline in the rat. Experientia 28, 1298–1299. Goodwin B. L., Ruthven C. R. J. and Sandler M. (1974) Gas chromato-
B. L. Goodwin et al. graphic estimation of urinary homovanillic and 4-hydroxy-3-methoxymandelic acids using capillary columns. Clin. Chim. Acta 55, 111–112. Goodwin B. L., Ruthven C. R. J. and Sandler M. (1975) Gas chromatographic assay of phenylacetic acid in biological fluids. Clin. Chim. Acta 62, 443–446. Goodwin B. L., Ruthven C. R. J., King G. S., Sandler M. and Leask B. G. S. (1978) Metabolism of 3,4-dihydroxyphenylalanine, its metabolites and analogues in vivo in the rat: urinary excretion pattern. Xenobiotica 8, 629–651. Gopalakrishna Y., Narayanan T. K. and Rao G. R. (1976) Synthesis of b-phenethyl alcohol in Candida guilliermondii. Biochem. Biophys. Res. Commun. 69, 417–422. Schwenk E. and Bloch E. (1942) New modification of Willgerodt’s reaction. J. Am. Chem. Soc. 64, 3051–3052. Sealock R. R., Speeter M. E. and Schweet R. S. (1951) The synthesis, resolution, and proof of configuration of isomers of m-tyrosine. J. Am. Chem. Soc. 73, 5386–5388. Shaw K. N. F., Armstrong M. D. and McMillan A. (1956) Preparation of m-hydroxyphenyl-l- and d-lactic acids and other compounds related to m-tyrosine. J. Org. Chem. 21, 1149–1151. Vogel A. I. (1961) A Textbook of Practical Organic Chemistry, 3rd edn, pp. 923–925. Longmans, London. Waser E. (1925) Decarboxylation of tyrosine and leucine. Helv. Chim. Acta 8, 758–773. Woolf L. I. (1951) Excretion of conjugated phenylacetic acid in phenylketonuria. Biochem. J. 49, ix.
Gen. Pharmac. Vol. 31, No. 3, pp. 437–440, 1998 Copyright 1998 Elsevier Science Inc. Printed in the USA.
Resistance of Aryl-Substituted Phenylalanines to Decarboxylation in the Rat Brian L. Goodwin, Colin R. J. Ruthven and Merton Sandler* Department of Chemical Pathology, Queen Charlotte’s and Chelsea Hospital, Goldhawk Road, London, W6 0XG, United Kingdom ABSTRACT. 1. The in vivo decarboxylation of analogues of DL-phenylalanine with methoxy, ethoxy or methylenedioxy substituents on the aromatic nucleus was assessed in the rat by measuring urinary excretion of the corresponding phenylacetic acids and b-phenylethylamines. Only trace amounts of these amines and acids were excreted. 2. o-Tyrosine was readily decarboxylated, a property that may be attributed to the position and nature of the substituent on the aromatic nucleus. 3. Some phenylpyruvic acids derived from these amino acids exhibited a certain degree of decarboxylation to phenylacetic acids or reduction to phenyllactic acids in the rat. gen pharmac 31;3:437–440, 1998. 1998 Elsevier Science Inc. KEY WORDS. Rat, aryl-substituted phenylalanines, phenylpyruvic acids, b-phenylethylamines, phenyllactic acids, phenylacetic acids
INTRODUCTION Because of their central role in mammalian biochemistry and physiology, the metabolic pathways for the l-isomers of phenylalanine, tyrosine and dopa have received extensive study. Despite their structural similarities, the main metabolic pathway for each is different, largely because of the initial metabolic step. Phenylalanine is rapidly hydroxylated in the para position, but it is not so readily decarboxylated [for a review, see Blau (1979)]. Transamination readily occurs only at high substrate concentration, as found in untreated phenylketonuria (Fo¨lling, 1934). In contrast, tyrosine undergoes almost quantitative transamination and dopa is extensively decarboxylated. Some analogues of tyrosine are readily metabolized, and phenolic acetic acids are prominent among the urinary metabolites. Goodwin et al. (1978) investigated the metabolism of dopa, m-tyrosine and 3-hydroxy-4-methoxy- and 4-hydroxy-3-methoxytyrosine in the rat and, with the exception of the last compound, a preponderance of decarboxylated metabolites was detected. These investigations were therefore extended to an examination of the relation between metabolic pathway and structure in a further range of phenylalanine analogues, in particular to compounds with alkoxy substituents on the aryl nucleus without a phenolic group. Evidence was sought in particular for the formation of products of decarboxylation relative to transamination-type metabolites (formed by transamination of l-amino acids or by oxidative deamination of d-amino acids), because results with phenylalanine, which also lacks a phenolic hydroxyl group, might suggest that the decarboxylation pathway is unlikely to be of major significance without maintained high concentrations of substrate (Blau et al., 1976). MATERIALS AND METHODS
Chemicals Amino acids (dl mixtures) were synthesized from commercially available benzaldehydes by the standard azlactone synthesis (Sealock et *To whom correspondence should be sent, at Department of Pharmacy, King’s College, Manresa Road, London, SW3 6LX, UK. Received 8 October 1997; accepted 15 December 1997.
al., 1951), whereby benzaldehydes and acetylglycine are condensed (in acetic anhydride), with the use of anhydrous sodium acetate as catalyst. The azlactones so formed were reduced in water with sodium amalgam, the azlactone ring being hydrolyzed under the alkaline conditions used, and the amino acids were then released by acid hydrolysis. No attempt was made to resolve these dl mixtures. Amines were prepared by the method of Brossi et al. (1966). The corresponding benzaldehydes were condensed with nitromethane, followed by reduction of the resultant nitrostyrenes with lithium aluminum hydride. Substituted phenylacetic acids were synthesized from the corresponding acetophenones by the modified Willgerodt reaction (Schwenk, 1942; Vogel, 1961) in which thiomorpholides of phenylacetic acids were prepared by refluxing the corresponding acetophenones with sulfur and morpholine. The free acids were released by acid hydrolysis. Pyruvic acids were prepared from azlactones by acid hydrolysis and the lactic acids by the reduction of the pyruvic acids with sodium amalgam in water (Shaw et al., 1956). Structures were confirmed by mass spectrometry.
Animal experiments Amino acids were administered intraperitoneally to female white Wistar rats (150–200 g) at a dose of 100 mg/kg. Where solubility was a problem, they were administered as hydrochloride salts. Pyruvic acids were administered in a similar manner, dissolved in sodium bicarbonate solution. The animals were kept in metabolism cages for 24 hr without food but with water supplied ad libitum. Urine was collected into 0.5 ml of 6 M HCl and stored at 2208 C.
Assay methods Phenolic acid metabolites and lactic acids were extracted from urine and converted into their silyl derivatives prior to gas chromatography (GC) (Goodwin and Sandler, 1975; Goodwin et al., 1974). Nonphenolic phenylacetic acids were measured by the GC method
438
B. L. Goodwin et al.
TABLE 1. Excretion of metabolites after administration of substituted phenylalanines to rats Substituent on benzene ring o-OH o-MeO m-MeO p-MeO 3-EtO-4-OH 2,3-diMeO 2,4-diMeO 2,5-diMeO 3,4-diMeO
Corresponding amine 8.6 (5.4–10.8) 0.025 (0.02–0.025) 0.04 (0.025–0.06) 0.05 (0.04–0.06) — 0.03 (0.03–0.04) 0.22 (0.15–0.33) 0.015 (0.0–0.025) 0.14 (0.08–0.22)
Corresponding phenylacetic acid
Corresponding phenyllactic acid
Corresponding phenylpyruvic acid
13 (8–16)* 0.5 (0.4–0.65) 0.25 (0.2–0.3) 1.3 (0.85–2.0) 0M 0.2 (0.15–0.25) TrM 0.2 (0.0–0.45) 3.0M (2.1–3.9)
0.4 (0.15–0.7) 0 0M 0M M 1.6 (0.95–2.1) — — 0M 7.2M (4.4–9.9)
— 2.4 (2.1–2.8) 0 — — 1.3M (0.9–1.5) — 3.8 (2.4–5.0) 10.0 (6.6–12.1)
3,4-Methylenedioxy
0
0M
0M
3,4,5-triMeO
—
0.25 (0.15–0.45)
TrM
Other metabolites
3,4-Dihydroxyphenylacetic acid 0 4-Hydroxy-3-methoxyphenylacetic acid 0.7 (0.45–1.0) 3,4-Dihydroxyphenylacetic acid 0M 4-Hydroxy-3-methoxyphenylacetic acid 0.7 (0.3–0.9) 3-Methoxytyramine 0.75 (0.6–1.0)
Abbreviation: Tr, trace. Results are expressed as a percentage of the dose excreted in 24 hr. M Confirmed by mass spectrometry. * Corrected for endogenous compound.
developed for phenylacetic acid (Goodwin et al., 1975), after acid hydrolysis to release free acids from their putative conjugates, but chromatographed at a higher oven temperature to compensate for their lower volatility. Putative sulfate conjugates of phenolic amines were hydrolyzed at 1008C and pH 1 for 20 min, and putative acetyl or similar conjugates of all other amine metabolites were hydrolyzed with 6 M HCl at 1008C for 2 hr (an extensive literature for analogous amines suggests that these are their major expected conjugates). Because the amines sought in this study are relatively nonpolar, the pH was adjusted to 10, and free amines were extracted into ethyl acetate. After evaporation of solvent, pentafluoropropionyl derivatives were prepared by warming with pentafluoropropionic anhydride. Phenylpyruvic acids were not amenable to GC (their silyl derivatives tailed badly). They were therefore reduced to the corresponding lactic acids prior to assay. An aliquot of urine (0.5 ml) was mixed with 0.1 ml 10 M NaOH, and 7–10 mg sodium borohydride was added. After about 1 hr, the mixture was acidified and the lactic acid content was assessed as heretofore described. The quantity of the pyruvic acid sought was assessed by difference. All assays were carried out in parallel with urine specimens to which measured quantities of the compounds sought had been added prior to processing; pyruvic acids (rather than lactic acids) were added to urine aliquots and then reduced as heretofore described. RESULTS AND DISCUSSION The most striking finding (Table 1) was the dearth of urinary metabolites found after administration of nonphenolic phenylalanines. The pattern of metabolites obtained with o-tyrosine, however, showed a pattern similar to that observed with m-tyrosine (Goodwin et al., 1978); an appreciable quantity of the decarboxylated metabolite o-tyramine and its further metabolite, o-hydroxyphenylacetic acid,
were found, in confirmation of the work of Blaschko (1949, 1950) and others, whereas only traces of the expected decarboxylated metabolites were detected for most of the remaining amino acids. One further amino acid tested, 3-ethoxy-4-hydroxyphenylalanine, carried a phenolic hydroxyl group; in this compound, the hydroxyl group is masked by an adjacent ethoxy group. The finding that this amino acid yields only small amounts of acidic metabolites mimics the results obtained with 3-methoxytyrosine in which the hydroxyl group also is para to the side chain. Although the presence of a hydroxyl group on the aromatic nucleus may be a prerequisite for efficient decarboxylation in mammals, this reaction is not quantitatively significant if the sole hydroxyl group is para to the side chain. Tyrosine, for instance, is highly resistant to decarboxylation in mammals [e.g., Blaschko (1950)]. This contrasts with the action of gut flora enzymes [e.g., Gale (1940)], which can affect decarboxylation. It is uncertain whether the effects of masking a hydroxyl group by an ortho substituent affects the rate or extent of decarboxylation; the results presented here and previously (Goodwin et al., 1978) suggest that the positional effect is more significant; for instance, 3-hydroxy4-methoxyphenylalanine and o-tyrosine both have a hydroxyl group ortho to a potential masking group, yet are readily decarboxylated, unlike 3-methoxytyrosine. The observation that the nonphenolic phenylalanines in general appear to be poorly metabolized by the expected pathways may be an indirect consequence of the overall biochemical economy of the mammalian organism, in which phenylalanine (itself lacking a phenolic hydroxyl group) is decarboxylated at normal physiological concentrations only as a minor pathway but undergoes extensive p-hydroxylation to tyrosine [for a review, see Goodwin (1979)] to dispose of the excess dietary phenylalanine ingested. The results reported here are in marked contrast with the chemical decarboxylation of tyrosine and its analogues. On heating to about
Phenylalanine Analogues in Rats
439
TABLE 2. Excretion of metabolites after administration of substituted phenylpyruvic acids to rats Substituent on benzene ring
Corresponding phenylacetic acid
Corresponding phenyllactic acid
o-MeO m-MeO p-MeO 3,4-diMeO 3,4,5-triMeO
2.0 (0.45–4.4) 1.9M (1.5–2.4) 29M (21–33) 0M 1.5 (0.8–2.7)
0M 5.0M (2.7–7.4) TrM 11.5M (4.2–22.0) 2.0M (0.5–5.0)
Other metabolites o-Hydroxyphenyllactic acid 0M m-Hydroxyphenylacetic acid 0.55M (0.2–0.95) 4-Hydroxy-3-methoxyphenylacetic acid 0.5M (0.2–0.85)
Abbreviation: Tr, trace. Results are expressed as a percentage of the dose excreted in 24 hr. M Confirmed by mass spectrometry.
3008C, a range of p-hydroxyphenylalanines are decarboxylated (unpublished results) in the same manner as tyrosine (Waser, 1925), providing an efficient method for preparing the amines. Yet m-tyrosine, which is efficiently decarboxylated in vivo, did not yield m-tyramine on heating under these conditions. This suggests that the reaction mechanisms at a molecular level are different. Tyrosine, which carries a hydroxyl group in the para position and is therefore not amenable to the same hydroxylation reaction as phenylalanine and is not appreciably decarboxylated, is catabolized by nearly quantitative transamination (Cammarata and Cohen, 1950) to yield the corresponding pyruvic acid, which is then reduced to p-hydroxyphenyllactic acid as a subsidiary reaction. These acids are not normally excreted to any significant extent, because most of the p-hydroxyphenylpyruvic acid is further metabolized to homogentisic acid prior to oxidation to CO2 and water [for a review, see Goodwin (1972)]. However, pyruvic acids formed from the nonphenolic amino acids would not be expected to undergo further oxidation in that p-hydroxyphenylpyruvate oxidase exhibits a high degree of specificity. Except after administration of 3,4-dimethoxyphenylalanine, the corresponding lactic acids were not detected at more than a small extent. A possible explanation would be that the corresponding pyruvic acids are not readily reduced in vivo. After the administration of several of these pyruvic acids, only a small proportion of the administered dose was excreted as corresponding lactic acids (Table 2). A possible explanation is that the pyruvic acids may have been transaminated back to the amino acids. The urinary excretion of 3,4-dimethoxyphenyllactic and -pyruvic acids after administration of the corresponding amino acid indicates that this amino acid undergoes significant transamination followed by reduction of the pyruvic acid. Decarboxylation of phenylpyruvic acid itself is a well-established metabolic pathway, at least in microorganisms [e.g., see Gopalakrishna (1976)], that may produce phenylacetic acid in the gut, but except in the case of p-methoxyphenylpyruvic acid, little of the corresponding phenylacetic acids were excreted. The relative absence of urinary metabolites of the nonphenolic amino acids does not necessarily imply that they are metabolically inert. This may be illustrated by the fact that tyrosine, under abnormal circumstances, can give rise to massive amounts of urinary metabolites—for instance, in neonatal tyrosinemia (Bloxham et al., 1960), similarly, in untreated phenylketonuria, there is extensive urinary excretion of phenylalanine metabolites (Armstrong et al., 1955; Fo¨lling, 1934; Woolf, 1951) in consequence of defects in the enzyme systems that normally bring about conversion of these amino acids into CO2 and water. However, only small (but readily detectable) amounts of these metabolites are excreted by a normal person. O-Dealkylation of the compounds assessed in this study may occur,
but in general, such metabolites were not sought (many of the expected metabolites arising from dealkylation occur normally in urine, often as plant natural product metabolites). Partial or total removal of the amino acid side chain may occur (Goodwin et al., 1972), but, in mammals, this has been reported only for phenylserines and not for phenylalanines. CONCLUSION Most of the nonphenolic-substituted phenylalanines examined in this study were resistant to decarboxylation. Evidence is presented that transamination of the amino acid l-isomers or oxidation of d-isomers or both to yield phenylpyruvic acids usually accounted for only a small proportion of the administered dose. o-Tyrosine was decarboxylated. Most of the phenylpyruvic acids studied, with the exception of 3,4-dimethoxyphenylpyruvic acid, were poorly reduced to phenyllactic acids. With the exception of p-methoxyphenylpyruvic acid, they were resistant to decarboxylation. SUMMARY Previously published results have demonstrated extensive variations in the preferred metabolic pathways for phenylalanine and arylhydroxylated phenylalanines associated with the normal economy of the mammalian organism. Evidence for the preferred metabolic pathways for a series of synthetic analogues in the rat was therefore sought. Decarboxylation of the nonphenolic amino acids was inefficient, whereas o-tyrosine, in common with l-dopa and m-tyrosine, was readily decarboxylated. This decarboxylation also distinguishes it from p-tyrosine and 3-methoxytyrosine, which are poorly decarboxylated in the rat. Transamination or oxidative deamination of the L- and D-isomers, respectively, or both, as evidenced by excretion of the corresponding pyruvic and lactic acids, was, in general, not demonstrated to be an efficient metabolic process under the experimental conditions used. We thank the Mass Spectrometry Unit at Queen Charlotte’s and Chelsea Hospital for confirming the identity both of pure compounds and urinary metabolites.
References Armstrong M. D., Shaw K. N. F. and Robinson K. S. (1955) Studies on phenylketonuria II: the excretion of o-hydroxyphenylacetic acid in phenylketonuria. J. Biol Chem. 213, 797–804. Blaschko H. (1949) The decarboxylation of o-hydroxyphenylalanine. Biochem. J. 44, 268–274. Blaschko H. (1950) Substrate specificity of amino-acid decarboxylases. Biochim. Biophys. Acta 4, 130–137. Blau K. (1979) Phenylalanine hydroxylase deficiency: biochemical, physiological, and clinical aspects of phenylketonuria and related phenyl-
440 alaninemias. In Aromatic Amino Acid Hydroxylase and Mental Disease (Edited by Youdim M. B. H.), p. 77–139. Wiley, Chichester, UK. Blau K., Goodwin B., Woods H. F. and Youdim M. B. H. (1976) Phenylethylamine formation in perfused rat liver. Br. J. Pharmac. 58, 474P. Bloxham H. R., Day M. G., Gibbs N. K. and Woolf L. I. (1960) An inborn defect in the metabolism of tyrosine in infants on a normal diet. Biochem J. 77, 320–326. Brossi A., Schenker F., Schmidt R., Banziger R. and Leimgruber W. (1966) Synthesen in der Isochinolinreihe. Helv. Chim. Acta 49, 403–411. Cammarata P. S. and Cohen P. P. (1950) The scope of the transamination reaction in animal tissues. J. Biol. Chem. 187, 439–452. ¨ ber Ausscheidung von Phenylbrenztraubensa¨ure in den Fo¨lling A. (1934) U Harn als Stoffwechselanomalie in Verbindung mit Imbezillita¨t. HoppeSeyler’s Z. Physiol. Chem. 227, 169–176. Gale E. F. (1940) The production of amines by bacteria 2: the production of tyramine by Streptococcus faecalis. Biochem. J. 44, 846–852. Goodwin B. L. (1972) Tyrosine Catabolism. Clarendon Press, Oxford. Goodwin B. L. (1979) Phenylalanine hydroxylase. In Aromatic Amino Acid Hydroxylase and Mental Disease (Edited by Youdim M. B. H.), pp. 5–76. Wiley, Chichester, UK. Goodwin B. L. and Sandler M. (1975) Improved resolution using capillary columns. Clin. Chim. Acta 59, 253–254. Goodwin B. L., Johnson R. D., Leask B. G. S., Ruthven C. R. J. and Sandler M. (1972) threo-3,4,-Dihydroxyphenylserine, a poor source of noradrenaline in the rat. Experientia 28, 1298–1299. Goodwin B. L., Ruthven C. R. J. and Sandler M. (1974) Gas chromato-
B. L. Goodwin et al. graphic estimation of urinary homovanillic and 4-hydroxy-3-methoxymandelic acids using capillary columns. Clin. Chim. Acta 55, 111–112. Goodwin B. L., Ruthven C. R. J. and Sandler M. (1975) Gas chromatographic assay of phenylacetic acid in biological fluids. Clin. Chim. Acta 62, 443–446. Goodwin B. L., Ruthven C. R. J., King G. S., Sandler M. and Leask B. G. S. (1978) Metabolism of 3,4-dihydroxyphenylalanine, its metabolites and analogues in vivo in the rat: urinary excretion pattern. Xenobiotica 8, 629–651. Gopalakrishna Y., Narayanan T. K. and Rao G. R. (1976) Synthesis of b-phenethyl alcohol in Candida guilliermondii. Biochem. Biophys. Res. Commun. 69, 417–422. Schwenk E. and Bloch E. (1942) New modification of Willgerodt’s reaction. J. Am. Chem. Soc. 64, 3051–3052. Sealock R. R., Speeter M. E. and Schweet R. S. (1951) The synthesis, resolution, and proof of configuration of isomers of m-tyrosine. J. Am. Chem. Soc. 73, 5386–5388. Shaw K. N. F., Armstrong M. D. and McMillan A. (1956) Preparation of m-hydroxyphenyl-l- and d-lactic acids and other compounds related to m-tyrosine. J. Org. Chem. 21, 1149–1151. Vogel A. I. (1961) A Textbook of Practical Organic Chemistry, 3rd edn, pp. 923–925. Longmans, London. Waser E. (1925) Decarboxylation of tyrosine and leucine. Helv. Chim. Acta 8, 758–773. Woolf L. I. (1951) Excretion of conjugated phenylacetic acid in phenylketonuria. Biochem. J. 49, ix.