CHAPTER 9
Inhibition of Amino Acid Decarboxylases*
William Gilbert Clark
I . Introduction 316 I I . Pyridoxal Kinase 318 I I I . Apoenzyme-Coenzyme Dissociation 319
I V . Transport, Uptake, Binding, Release 320
V. Inanition 322 V I . Tissue Damage, Growth, Neoplasms, Organectomy 322
V I I . Apoenzyme Synthesis 323 V I I I . Inhibition by Apoenzyme Antibody 326
I X . Metals, Chelators, and Metal Complexers 327
X . Cyanide 330 X I . Pyridoxal-5-Phosphate (Codecarboxylase), Vitamin B6, Be-Deficien-
cy, Be-Antagonists 331 A . Ββ-Deficiency and Antagonists 331
B. B6- P 04 Hydrazones 334
C. Toxopyrimidine 335 D . Steroids 337 X I I . Activators, Stabilizers, and Cofactors Other Than Be-P04 337
A . Metals 337 B . Surfactants 337 C. Solvents 337 D . Phosphates and Arsenate 338
E. Miscellaneous 338 X I I I . Carbonyl Reagents and Inhibitors Acting on Coenzyme Other Than
Cyanide, Substrate Analogues, and Pyridoxine Antagonists 339 A . Hydroxylamine, Hydrazides, Semicarbazide, Sulfite, Hydrazine,
Oximes, etc 339 B. Cycloserine (4-Amino-3-isoxazolidone) 343
C. Penicillamine 345 D . Sulfonylureas 345 E. Cysteine, 2,3-Dimercapto-l-propanol ( B A L ) , Glutathione 346
F. Ascorbic Acid 346
* Supported by Grants from the National Mental Health Association, U.S. Public Health Service, American Heart Association, Los Angeles County Heart Association, U.S. Army Chemical Center, American Cancer Society, Office of Naval Research, Helen Hay Whitney, Jr. Foundation, Life Insurance Medical Research Fund.
315
I. INTRODUCTIO N
Although enzymi c decarboxylatio n play s a mino r rol e quantitativel y i n metabolism o f amin o acids , i t i s a n importan t on e becaus e o f th e critica l nature, marke d pharmacologica l activity , an d functio n o f th e en d products . The physiologica l significanc e o f most , i f no t all , o f thes e en d product s stil l remains t o b e clarified , bu t man y o f the m probabl y ar e essentia l fo r a num - ber o f homeostati c regulator y an d adaptiv e mechanism s i n al l livin g organ - isms. I t wa s no t unti l 1936-193 7 tha t amin o aci d decarboxylase s wer e de - scribed b y Okunuk i (1937 ) i n plant s (glutami c aci d decarboxylase ) an d b y Werle (1936 ) an d Holt z an d Heis e (1937 , 1938 ) [histidin e an d 3,4-dihy - droxyphenylanine (dopa ) decarboxylase] . Thi s an d othe r work , includin g data o n inhibitor s o f thes e decarboxylases , hav e bee n reviewe d b y Gal e (1940b, 1946 , 1953) , Holt z (1941) , Storc k (1951) , Jank e (1951) , Karre r (1947), Mardashe w (1949) , Schale s (1951) , Mèiste r (1955 , 1957) , Werl e (1943a,b, 1947 , 1951) , an d Rui z an d Zaragoz a (1959) .
It i s th e purpos e o f thi s revie w t o discus s inhibitio n o f th e amin o aci d decarboxylases i n general , wit h som e emphasi s o n mor e recen t contribu - tions (u p t o earl y 196 1 i n mos t cases) . Inhibitio n b y genera l enzym e
"poisons" o r dénaturant s in vitro, suc h a s trichloroacetat e an d alkylat -
X I V . Hormone s 34 6 A . Insuli n 34 7 B. Pituitar y 34 7 C. Adrena l Corte x 34 7 D . Thyroi d 34 9 E. Se x Hormone s 35 2 X V . Miscellaneou s 35 4
A . Antibiotic s 35 4 B. Antihistamine s 35 4 C. Reserpin e 35 4 D . Tranquilizer s 35 4 E. Tetrahydroisoquinoline s 35 4
F. Foli c Aci d Antagonist s 35 5 X V I . Substrat e Analogue s 35 6
A . Bacteria l an d Plan t Decarboxylase s 35 6 B . Mammalia n (an d Fowl ) Decarboxylase s in vitro 35 7
C. Mammalia n Dop a Decarboxylas e in vitro 35 8 D . Mammalia n 5-HT P Decarboxylas e in vitro 35 8 E. Glutami c Aci d Decarboxylas e in vivo 35 9
F. Dop a Decarboxylas e in vivo 35 9 G. Quinone s an d Potentia l Quinoid s 36 0
H . Ketonuri a 36 1 I . α-Alkyl Substrate Analogues 363
References 366
9. I N H I B I T I O N OF A M I N O ACID DECARBOXYLASES 317 ing agents, will not be included. Clark (1959) and Clark and Pogrund
(1961) recently reviewed the subject of dopa decarboxylase inhibition in vitro and in vivo, and Sourkes and D'lorio discuss the subject in Volume I I of this treatise. Since many inhibitors of these enzyme systems exert their effects directly or indirectly through pyridoxal-5-phosphate (B6-P04), the coenzyme of amino acid decarboxylases, some aspects of the vitamin B6-dependent enzymes in general must be considered. Ref
erence should be made to the extensive reviews available of the pyri- doxine-dependent enzymes by Blaschko (1945a), Gunsalus (1951), Wil
liams et al. (1950), Sherman (1954), Tower (1956, 1959, 1960), Umbreit (1954), Meister (1957), Mathews (1958), Snell (1958), Snell and Jenkins (1959), Siliprandi (1960), Roberts and Eidelberg (1959), Roberts et al (1960), Braunstein (I960), and Axelrod and Martin (1961). Braunstein's review, "Pyridoxal Phosphate" (1960), and that of Snell (1959), "Chemical Structure in Relation to Biological Activities of Vitamin B6, " are particu
larly exhaustive.
The amino acid decarboxylases described so far catalyze the following reactions:
(1 (2;
(3:
(4;
(5;
(β:
(7;
(s:
(9:
do:
(11 (12:
(13 (14 (15 (16 (17 (is:
(i9:
(20 (21 (22 (23 (24 (25:
(26 (27
Glycine —> methylamine L-Alanine —> ethylamine L-Serine —> ethanolamine
α-Aminobutyric acid —• propylamine L-Methionine —> 3-methylthiopropylamine L-Valine —> isobutylamine
L-Norvaline —• butylamine L-Leucine —> isoamylamine
L-Isoleucine —* 3-methylbutylamine L-Aspartic acid —> L-alanine L-Arginine —> agmatine L-Histidine —> histamine L-Aspartic acid —• /3-alanine L-Ornithine —> putrescine L-Lysine —> cadaverine
meso-δ-6-Diaminopimelic acid —> L-lysine α-Aminomalonic acid —* glycine
δ-Hydroxy-L-lysine —» hydroxycadaverine L-Glutamic acid —> 7-aminobutyric acid
Allo-jS-hydroxy-L-glutamic acid —• 7-amino-/3-hydroxybutyric acid 7-Hydroxy-L-glutamic acid —• a-hydroxy-7-aminobutyric acid 7-Methylene-L-glutamic acid —> 7-amino-a-methylene butyric acid L-Cysteic acid —> taurine
L-Cysteinesulfinic acid —> hypotaurine L-Tryptophan - > tryptamine
4-Hydroxy-L-tryptophan —* 4-hydroxytryptamine
5-Hydroxy-L-tryptophan ( ^ δ - Η Τ Ρ " ) - > 5-hydroxytryptamine (serotonin,
"5HT")
(28) 6-Hydroxy-L-tryptophan —» 6-hydroxytryptamine (29) α-Methyl-L-tryptophan —» a-methyltryptamine
(30) a-Methyl-5-hydroxy-L-tryptophan —» a-methyl-5-hydroxytryptamine (31) a. L-Phenylalanine —> phenylethylamine
Ring-substituted hydroxy-L-phenylalanines to corresponding amines, e.g. : b. L-Tyrosine —» tyramine
c. Diiodo-L-tyrosine —* diiodotyramine (Werle, 1947; not confirmed)
d. 2-Hydroxy-L-phenylalanine (o-tyrosine) —> 2-hydroxyphenylethylamine (o-tyramine)
e. 3-Hydroxy-L-phenylalanine (ra-tyrosine) —• ra-tyramine
f. a-Methyl-3-hydroxy-L-phenylalanine (α-methyl-m-tyrosine) —> a-methyl- 3-hydroxyphenylethylamine (a-methyl-ra-tyramine)
g. 3,4-Dihydroxy-L-phenylalanine (dopa) —• 3,4-dihydroxytyramine (dopa
mine)
(32) a. L-Phenylserine —• phenylethanolamine
b. 2-Hydroxy-L-phenylserine —• 2-hydroxyphenylethanolamine c. 3-Hydroxy-L-phenylserine —> 3-hydroxyphenylethanolamine
d. 4-Hydroxy-L-phenylserine —• 4-hydroxyphenylethanolamine (octopamine, norsynephrine)
e. 3,4-Dihydroxy-L-phenylserine (dops) —> 3,4-dihydroxyphenylethanol- amine (arterenol, norepinephrine)
Some of these reactions have been shown to be catalyzed by one and the same enzyme, and possibilities of this kind should be borne in mind in considering this section. Further discussion of this point, classifying and documenting the individual enzymes by their distribution in micro
organisms, plants, and animals, the structures and stereospecificity of their substrates, their kinetics, apoenzyme-coenzyme affinity and dissoci
ation must be sought in the reviews cited. Since the introduction of tracer techniques and amine catabolic enzyme inhibitors, many decarboxylations formerly thought to be absent in animals are being announced.
II. PYRIDOXAL KINASE
Snell (1959) and McCormick et al (1961; McCormick and Snell, 1961) have discussed codecarboxylase kinase (pyridoxal phosphokinase) and have reviewed the literature. Chevillard and Thoai (1951) and Thoai and Chevillard (1951a,b) showed that Mg+ + and Mn+ + activate it, A T P or ADP are necessary, and thiamine inhibits it. Hurwitz (1952, 1953, 1955) showed that some pyridoxine analogues inhibit tyrosine decarboxylase in bacteria, while others inhibit the phosphorylation of pyridoxal in the presence of ATP. Some adenine and purine derivatives inhibit competi
tively and several metal cations activate (cf. McCormick et al., 1961;
McCormick and Snell, 1961). Hurwitz suggested that the adenines and purines act through the activating metallic ions. McCormick and Snell
9. I N H I B I T I O N OF A M I N O ACID DECARBOXYLASES 319 (1959, 1961) and McCormick et al. (1960, 1961) showed that purified pyridoxal phosphokinase from brain is markedly inhibited by a variety of condensation products formed from pyridoxal and hydroxylamine, O-sub- stituted hydroxylamines, hydrazine, and substituted hydrazines; they questioned former explanations of the convulsive effects of such carbonyl reagents which causally related the seizures to lowered brain 7-amino
butyric acid (gaba) through interaction with B6-PO4. Dubnick and co
workers (1960c) postulated that pyridoxal hydrazones are formed from hydrazines and B6-P04 in vivo and exert toxic effects by inhibiting the phosphokinase (cf. Balzerei al., 1960a,b,c; and Baxter and Roberts, 1960).
Recently, Wada and Snell (1961), Turner and Happold (1961), and Wada et al. (1961) described an enzyme which oxidizes pyridoxine and pyri- doxamine phosphates to B6-PO4. Evidently, the primary pathway of the formation of B6-P04 is phosphorylation by the kinase, followed by the action of the oxidase. The oxidase, a flavoprotein with riboflavin-5'- phosphate as a cofactor, is sensitive to thiol reagents, heavy metals, and some phosphorylated analogues of Be, especially pyridoxal phosphate oxime. Greenberg et al. (1959) observed that chlorpromazine had little or no effect on glutamic acid decarboxylase of rat brain, but that the kinase in brains of B6-deficient rats has a greater susceptibility to inhibition by chlorpromazine than that in- normal rats. It is anticipated that many compounds acting on B6-enzyme systems, including the decarboxylases, will be shown to do so by effects on the kinase and the oxidase.
III. APOENZYME-COENZYME DISSOCIATION
Mason (1957) showed that phosphate ions may, at somewhat higher concentrations, inhibit transamination of kynurenine by rat kidney and that this can be reversed by B6-PO4. Certain keto acids in low concentra
tions also prevented the inhibition by decreasing dissociation and main
taining the holoenzyme in a more stable apoenzyme-coenzyme form.
Hartman et al. (1955) showed that inorganic phosphate and arsenate re
activated dialyzed dopa decarboxylase of hog kidney when added to
gether with B6-P04, but that none of these three substances reactivated alone. It is possible that this effect, like that in Mason's preparation, can also be explained by an effect of phosphate or arsenate on the apoenzyme- coenzyme dissociation. A systematic study of this should be undertaken on the effect on decarboxylations.
Werle and Aures (1959) showed that the reaction rates of purified dopa, 3,4-dihydroxyphenylserine (dops), and 5-HTP decarboxylases are differ
ent, but that when the differences in requirement, sensitivity, and affinity
of coenzyme-apoenzyme in the presence of the different substrates are taken into account, the activity ratios of the decarboxylations of these three substrates remain constant through all purification steps, thus indicating that the same enzyme decarboxylates all three substrates. Other evidence supporting this contention now is fairly conclusive and has been summarized elsewhere by this reviewer (1959; Clark and Pogrund, 1961);
and it is discussed later in this chapter.
Ekladius and co-workers (1957) showed that leucine competitively inhibits valine decarboxylase in bacteria, depending on the amount of coenzyme present. Leucine decarboxylation is less dependent on this factor. When coenzyme is not present in excess, the decarboxylation of either amino acid is completely inhibited by the other.
IV. TRANSPORT, UPTAKE, BINDING, RELEASE
Werle (1947) was the first to include among decarboxylase inhibitors substances which might inhibit uptake by cells of the amino acid sub
strates. The subject of the active absorption of amino acids has been reviewed and investigated by Christensen (1955, 1960), Meister (1957), Guroff and Udenfriend (1960), Neame (1961), Edelman (1961), and Gagnon (1961). The uptake can be blocked in some instances by inhibitors of respiratory metabolism or associated phosphorylative enzyme systems.
Schanberg and Giarman (1960a, b) found that the active uptake of labeled 5-HTP by brain and other tissue slices is inhibited by 02 lack, low tem
perature, and 2,4-dinitrophenol (DNP). The uptake was most active in those tissues where 5-HTP decarboxylation is greatest. Schanberg et al.
(1961) also showed that the active uptake of 5-HTP by brain slices is markedly inhibited by tryptophan, tyrosine, and dopa, but not by a-methyl- dopa, glutamic acid, and gaba. This correlated with a decrease of brain 5-HT in rats fed large doses of tryptophan. Active uptake of dopa by mitochondrial particles of guinea pig brain is inhibited by Ca
+ + , Mg+
+ , and nonionic detergents, but not by inhibitors of oxidative phosphorylation
(Iwamoto and Nukada, 1961). Neame (1961) reported active uptake of six amino acids by brain slices, inhibited by 02 lack and cyanide. Weiss- bach et al. (1960b) arid Weissbach and Redfield (1960) reported that tryptophan, but not tyrosine or 5-HTP, is actively absorbed by blood platelets. The process is not inhibited by fluoride as is the active uptake of 5-HT, a glycolytic enzyme-dependent system (vide infra). Christensen and others (cf. Moldave, 1958) found that pyridoxine is involved. Jacobs et. al. (1960) found that the inhibitory effect of D N P on active uptake of L-methionine from the intestine is partially counteracted by Be-P04
9. INHIBITION OF AMINO ACID DECARBOXYLASES 321 but not B6. They also showed the active intestinal absorption of L-tyrosine is B6-P04-dependent [but cf. Guroff and Udenfriend (I960)]. They reviewed work which supports the B6-dependent concept and mechanisms proposed by Pal and Christensen (1958), Riggs and Walker (1958), Christensen et al.
(1960), Christensen (1960), Shishova (1956) and others (cf. Riggs and Walker, 1960, and related studies mentioned in Section X I V , B ) . Akedo et al.
(1960) and Ueda et al. (1960) showed impaired active intestinal absorption of L-amino acids in B6-deficient rats and found that B6-PC>4 prevents the in
hibitory effect of DNP. Tsukada et al. (1960) showed that the active uptake of gaba by brain slices is augmented by pyridoxal. One mechanism proposed for the participation of pyridoxal in, for example, the transport of glycine is that it increases potassium efflux, which in turn increases glycine uptake.
Hempling (1961) in a preliminary abstract tested this theory by measur
ing K 42
fluxes in the presence and absence of pyridoxal during glycine uptake, and found that pyridoxine decreased the influx and efflux, but not if glycine was present, since glycine stimulates potassium fluxes.
Oxender and Royer (1961) found that the stimulation of active amino acid uptake by Ehrlich ascites cells by pyridoxal and B6-P04 was not inhibited by carbonyl reagents; hence, the B6 substances may not be directly involved as carriers.
Christensen (1955) showed that α-methylamino acids were often par
ticularly active in the competitive inhibition of active uptake of amino acids by cells. If a methyl group was introduced on the α-carbon, the degree of accumulation was greater, due perhaps to the formation of a Schiff base with Be-derivatives. It was suggested that the absence of an α-methyl group would prevent electromeric shifts of the double bond which characterized Be-derivatives. Thus, the α-methyl group might pre
vent the diversion of portions of the hypothetical carrier amino acid complex into forms which would not redissociate readily. The active up
take of amino acids is competitively inhibited by other amino acids (Chris
tensen, 1955, 1960; Wiseman and Ghadially, 1955; Hagihira et al., 1960, 1961; Lin and Hastings, 1960; Tenenhouse and Quastel, 1960; Jacquez, 1961), and Lin et al. (1961) showed that a free carboxyl group is essential for the transport, that substitutions on the α-amino group which prevent Schiff base formation block the transport, and that the α-hydrogen seems important, since replacement by a methyl group decreases uptake, for example, a-methyl-DL-tyrosine.
a-Methylphenylalanine analogues have been shown to inhibit the
"nonspecific" aromatic amino acid decarboxylase of mammals (see Sec
tion X V I , I ) . Goldberg et al. (1960), Porter et al. (1961), Hess et al. (1961), and Sourkes et al. (1961) have shown, however, that the effects of these analogues on tissue amine levels in vivo are due much more to a defect
in an uptake-binding-release mechanism than to inhibition of the enzyme.
In connection with uptake, Oates and associates (1960b) found no effect of α-methyldopa on the active absorption of tyrosine by the intestine, brain, or muscle.
V. INANITION
Since starvation can affect synthesis of proteins, including apoenzymes and coenzymes, in vivo, this factor must be controlled by paired feeding and similar measures in experiments involving B6- and other vitamin de
ficiencies, febrile diseases, thyrotoxicosis, neoplasms, and other debilitating states possibly affecting amino acid decarboxylation in vivo (cf. Blaschko, 1945b, Armstrong et al, 1950; Holtz et al, 1956; Blaschko et al, 1948; and Weil-Malherbe, 1956). Failure to consider such nutritional factors by some biochemists casts doubt on many reports in the literature.
VI. TISSUE DAMAGE, GROWTH, NEOPLASMS, ORGANECTOMY Hawkins and Walker (1952), in studying the effect of rapid nuclear and cellular division on soluble and insoluble enzyme activity, examined dopa decarboxylation by rat liver and adrenal medulla after partial hepatectomy.
Before compensatory hypertrophy was initiated, the enzyme activity de
creased, but was restored nearly to normal with compensatory hyper
trophy. This increase, however, was not initiated at the same time as growth. Kahlson et al, using histidine-C
14
on the other hand, found histi
dine decarboxylase (presumably the "specific" L-histidine decarboxylase discussed in Section V I I ) markedly increased in such livers as well as in wound healing (Kahlson, 1960; Kahlson et al, 1960a) and in embryos
(Kahlson et al, 1958, 1959, 1960b; Kahlson and Rosengren, 1959b), and believe that histamine plays a role in growth (Kahlson, 1960) and neo
plasms. Kizer and Chan (1961) found a correlation between impairment of 5-HTP decarboxylation and carcinogenesis in rat liver. Aures, in this laboratory (unpublished), could not confirm Kahlson's report of increased histidine decarboxylase in the rapidly growing liver after partial hepatec
tomy in rats.
Clark (1959; Clark and Pogrund, 1961) showed that if cats and rats are kept in good physiological condition after acute nephrectomy, gastrectomy, enterectomy, or indeed total evisceration, blood pressure responses to intravenously administered dopa remain nearly normal, indicating an
9. I N H I B I T I O N OF A M I N O ACID D E C A R B O X Y L A S E S 323 ubiquitous distribution of dopa decarboxylase. This does not mean, of course, that effects are not seen some time after castration, ovariectomy, pancreatectomy, thyroidectomy, or adrenalectomy (see Section X I V , E ) .
Of importance is the observation of Gey and Pletscher (1960) that 5-HTP decarboxylase activity of brains removed from rats and left at 37°C shows a decrease during the first half-hour. The enzyme activity continues for some time, however. This observation subsequently led the authors (Pletscher and Gey, 1961) to develop a useful assay for 5-HTP decarboxylase activity and its inhibitors by analyzing for brain 5-HT in incubated heads of rats decapitated after injection of 5-HTP and inhibitor.
VII. APOENZYME SYNTHESIS
When discussing inhibitors of enzyme systems, consideration must be made of the effects of agents and conditions on the synthesis of apoenzymes, some of which can be through adaptive or induced mechanisms, both in microorganisms and in higher animals. The former has been reviewed by Pollock (1959) and the latter by Knox et at. (1956).
Gale (1940a) and others have shown that many bacterial enzymes, including amino acid decarboxylases, are deficient or absent if their sub
strates are absent and are adaptively induced when these are added. The story is complicated by cross induction and many other factors; for ex
ample, an uncharacterized inhibitor of several amino acid decarboxylases is produced by Pseudomonas reptilivora, when assays are carried out with heavy cell suspensions (Seaman, 1960). A single inducer may affect more than one enzyme (Ando, 1959b). Further, substrate analogues may com
petitively inhibit induced synthesis. Mandelstam (1956) showed that in
duced syntheses of ornithine and lysine decarboxylases in Bacillus cadaveris are reversibly inhibited by β-phenylserine and 5-methyltryptophan. Pre
sumably many reports of bacterial growth inhibition by substrate analogues, if re-examined experimentally with methods now available, might reveal such effects on induced syntheses of other amino acid decarboxylases, as well as other enzymes.
D-Chloramphenicol (but not the L-isomer), a phenylserine analogue, exerts part, if not all, of its antibiotic action on the ATP-dependent syn
thesis of proteins, including certain bacterial apoenzymes (cf. Section X V , A ) . Thus, Grunberger and Sorm (1954), Grunberger et al. (1948, 1955), and Sorm and Grunberger (1953) showed that this and other anti
biotics (chlortetracycline, oxytetracycline) in subgrowth-inhibitory con
centrations inhibit the production of glutamic acid, lysine, and arginine decarboxylases and aspartic acid oxidase in Escherichia coli, but not that
of tyrosin e decarboxylas e b y Streptococcus faecalis. Sor m et al. (1955 ) als o showed tha t strain s o f E. coli mad e slightl y resistan t t o chloramphenicol , synthesize les s glutami c aci d decarboxylas e tha n thos e mad e highl y re - sistant, whic h hav e norma l levels . Further , Grunberge r an d Sor m (1954 ) showed tha t th e dru g ha s n o inhibitor y effec t o n fou r transaminases , pyridoxal kinase , o r adaptiv e enzyme s whic h participat e i n carbohydrat e metabolism; thu s i t i s fairl y specific . I t i s interestin g tha t Lagerbor g an d Clapper (1952 ) foun d arginine , glutami c acid , tyrosine , an d histidin e de - carboxylases presen t i n som e bu t no t al l o f thirty-thre e strain s o f lacto - bacilli examined . Fou r strain s o f Streptococcus mitis an d tw o o f S. faecalis which lacke d thes e enzyme s produce d arginin e an d tyrosin e decarboxylase s when mad e resistan t t o sulfathiazole . N o explanatio n i s given .
Inhibitors o f nuclei c aci d metabolis m ca n preven t th e adaptiv e synthesi s of apoenzymes . Thus , And o (1959a ) showe d tha t 8-azaguanine , i n concen - trations insufficien t t o affec t cellula r growth , strongl y inhibit s th e induce d synthesis o f histidin e decarboxylas e i n Proteus morganii. Excessiv e amount s of purine s an d pyrimidine s ma y als o inhibi t induce d apoenzym e synthesis , as show n b y Bellam y an d Gunsalu s (1946 ) wit h uraci l an d guanin e o n tyrosine decarboxylas e o f S. faecalis.
Melnykovych an d Johansso n (1955 ) an d Melnykovyc h et al. (Melnyko - vych an d Johansson , 1958 , 1959 ; Melnykovyc h an d Snell , 1958) , i n studie s of th e mechanis m o f growth-stimulatin g effect s o f antibiotic s i n animals , found tha t subgrowth-inhibitor y level s o f severa l bu t no t al l antibiotic s studied, i n additio n t o inhibitin g enzymi c decarboxylatio n o f severa l amin o acid decarboxylase s b y E. coli, stimulat e thei r inductive synthesis . The y speculate o n th e possibilit y tha t thes e antibiotic s ma y favorabl y influenc e animal growt h b y preventin g th e productio n o f toxi c amine s b y intestina l bacterial flora bu t that , insofa r a s th e inductive synthese s ar e concerned , they ma y i n som e wa y favorabl y affec t th e integrit y o f th e holodecar - boxylases. Ehrisman n an d Werl e (1948 ) foun d tha t th e nonadaptiv e synthesis o f histidin e decarboxylas e b y strain s o f E. coli, B. parasarco- physematos, an d a Clostridium i s inhibite d b y semicarbazide . She r an d Mallette (1954 ) foun d tha t bacteriophag e infectio n o f E. coli block s th e adaptive synthesi s o f lysin e decarboxylase . Th e phag e ghost s als o ar e active. A n explanatio n i s lacking .
Rajewsky et al. (1959 ) an d Bùcke r et al. (1960 ) studie d th e ultraviole t irradiation inactivatio n spectru m o f th e inductive synthesi s o f lysin e decarboxylase i n Bacterium cadaveris. Th e spectru m i s identica l wit h tha t of nuclei c acids , an d th e spectru m o f inactivatio n fo r th e maxima l attain - able enzym e activit y i s simila r t o th e actio n spectru m o f a cysteine-ric h protein. The y foun d a satisfactor y correlatio n wit h th e actio n spectru m o f ribonuclease an d sugges t tha t th e velocit y o f th e induce d enzym e synthesi s
9. INHIBITION OF AMINO ACID DECARBOXYLASES 325 s controlled by a nucleic acid but that the final activity is controlled by mother mechanism.
This reviewer could find no reports in the literature of "feedback con
trol" of amino acid decarboxylase synthesis (nor of decarboxylation rate tself for that matter) similar to those discussed by Doy and Pittard
^1960). These authors found that in Aerobacter aerogenes tryptophan may
;ontrol its own biosynthesis by inhibiting ("negative feedback control") he action of an enzyme necessary for a stage prior to anthranilic acid;
they review nine publications on other feedback controls of enzyme syn
thesis. Such considerations have not been made in the case of animal enzymes, but this should be taken into account in studies of inhibitors in intact cells and animals, especially if specificity is claimed.
Knox and associates (1956) in their review on enzymic adaptations in animals do not list decarboxylases among the enzyme systems induced by treatment with substrates. Shishova and Gorozhankina (1959) found in
creased histidine decarboxylase in the blood of rats fed large amounts of histidine in the diet. They also reported an increase in histidine decar
boxylation by livers of rats injected with cortisone. Several enzymes in animals have been reported to be induced adaptively by adrenal cortical hormones (Knox et al, 1956).
Schayer, who introduced a new era in the biochemistry, physiology, and pharmacology of the catechol amines and histamine by first using radio
isotope techniques for the study of the metabolism of these compounds, has made the remarkable discovery that histidine decarboxylase synthesis can be rapidly and markedly stimulated in intact animals by a number of nonspecific "stressful" treatments. These include injections of histamine releasers, histamine itself, endotoxin, adrenaline, 5-HT, burns, exposure to low temperature, development of the tuberculin reaction, and sensitiza
tion to pertussis vaccine (Schayer et al., 1959, 1960; Schayer and Ganley, 1959a, b, 1960, 1961 ; Schayer, 1960, 1961, 1962). It is possible that the sea
sonal variations of dopa decarboxylase activity of guinea pig kidney ex
tracts, shown by Polonovski and co-workers (1946) to be due to temperature effects on the whole animal, are caused by a nonspecific effect on the adaptive synthesis of apoenzyme. In one paper, Schayer et al. (1959) sug
gests that the effect may be related to the production of new, resistant mast cells, active in forming the "specific" L-histidine decarboxylase (vide infra). It would be interesting to see if the enzyme in basophilic leucocytes is adaptively affected because of their similarity to mast cells and because they are the only source of unreleased histamine in most species except the rabbit and contain the enzyme (Hartman et al., 1961).
The adaptive enzyme-induced production of histamine by other tissues is much greater than in blood, however, and Schayer et al. (1959; Schayer and
Ganley, 1959a, b, 1960, 1961 ; Schayer, 1959, 1960, 1961, 1962) have in
voked it in the homeostasis of the microcirculation. If true, the physio
logical implications will be profound.
Kahlson et al. (1960a; Kahlson, 1960) found that the inhibition of histidine decarboxylase activity in intact rats, caused by B6-deficiency and/or semicarbazide inactivation of the coenzyme, is followed by an overshoot in activity, caused by adaptive biosynthesis of new apoenzyme when the inhibition is terminated. It is possible that the histidine decar
boxylase of mast cells (Weissbach et al.y 1961), rat fundus (Schayer, 1956c, 1957), embryonic rat organs (Ganrot et al., 1961), and probably the adaptive enzyme of Schayer (see previous paragraph) and of Kahlson (Section V I ) is an L-histidine decarboxylase, unrelated to the general aromatic amino acid decarboxylase of most mammalian tissues (Lovenberg et al., 1962). An adaptive increase in the nonspecific decarboxylase has not been reported.
It would be of great interest to find effective inhibitors of the "specific"
histidine decarboxylase, since those which inhibit the "nonspecific" enzyme do not inhibit the "specific" one (see Section X V I , I ) .
VIII. INHIBITION BY APOENZYME ANTIBODY
Happold and Ryden (1952) attempted to obtain immune rabbit serum by using tyrosine apodecarboxylase of S. faecalis prepared by the method of Epps (1944, 1945) as antigen, and although a precipitin test was ob
tained against the original antigen, the enzyme activity remained un
changed or was slightly increased after removing the precipitate. Howe and Treffers (1952) were more successful with lysine decarboxylase par
tially purified from E. coli used as antigen in rabbits, obtaining 90%
inhibition of homologous enzyme and no inhibition of glutamic acid de
carboxylase. All activity was found in the precipitate obtained by the precipitin reaction. Gubarev (1960), using glutamic acid decarboxylase of Bacillus dysenteriae purified by starch column electrophoresis as antigen, obtained an active rabbit antiserum which partially inactivated the original homologous enzyme.
If apoenzymes could be sufficiently purified, it would be of interest for purposes of classification to attempt similar experiments with the other decarboxylases, especially in higher organisms. Whether or not inhibitions could be obtained by such immunological approaches in higher animals remains to be seen, but the physiological implications would be of great interest.
The general subject of possible inhibitions by immunological approaches in higher animals was reviewed some years ago by Marrack (1951). If the
9. I N H I B I T I O N OF A M I N O ACID DECARBOXYLASES 327 antigenic groups of the enzyme molecule are not situated in the active centers, presentation of antibody might cause the immune reaction, such as precipitation, with no loss of activity. If the groups are at active centers, then reaction with antibody will inhibit in competition with substrate.
IX. METALS, CHELATORS, AND METAL COMPLEXERS Heavy metal inhibition of amino acid decarboxylases (they will be listed below by amino acid only) has been studied extensively.
A. Ferric Ion
With the exception of one report which omits methods (Martin et al.} 1942), most amino acid decarboxylases studied are inhibited by relatively high concentrations of F e
+ +
+: bacterial (Epps, 1944, 1945; Gale and Epps, 1944; Taylor and Gale, 1945; Krebs, 1950; Saito, 1957); plant glutamic acid (Eggleston, 1958; Bottger and Steinmetzer, 1960); mam
malian dopa only slightly (Fellman, 1959).
B. Ferrous Ion
Bacterial (Taylor and Gale, 1945; Arjona et al, 1950; Eggleston, 1958);
plant glutamic (Eggleston, 1958) ; mammalian glutamic (Eggleston, 1958) but not 5-HTP (Buzard and Nytch, 1957a, b).
C. Cupric Ion
All studied. Bacterial (Epps, 1944, 1945; Gale and Epps, 1944; Taylor and Gale, 1945; Ishikawa and Obata, 1955; Eggleston, 1958; Koizumi et al., 1958; Yamagami, 1958); plant glutamic (Okunuki, 1943; Bôttger and Steinmetzer, 1960) ; mammalian, cysteic, cysteinesulfinic, and glutamic, which may be identical (Davison, 1956b), and 5-HTP (Buzard and Nytch 1957a, b).
D. Silver Ion
Bacterial (Epps, 1944, 1945; Gale and Epps, 1944; Taylor and Gale, 1945; Ekladius et al, 1957; Saito, 1957; Eggleston, 1958; Yamagami, 1958; Sutton and King, I960); plant glutamic (Okunuki, 1943).
E. Mercury Ion
AU studied. Bacterial (Epps, 1944; Gale and Epps, 1944; Taylor and Gale, 1945; Oliver, 1952; Saito, 1957; Ekladius et al, 1957; Koizumi et al, 1958; Yamagami, 1958; Sutton and King, 1960); but not plant glutamic (Okunuki, 1943); mammalian dopa (Fellman, 1959).
F. Lead (Plumbic) Ion
Bacterial histidine and tyrosine (Epps, 1944, 1945).
G. Nickel Ion
Mammalian glutamic (Eggleston, 1958).
H. Cobaltic Ion
No effect on bacterial (Krebs, 1950); plant glutamic (Okunuki, 1943);
and mammalian dopa (Fellman, 1959), but inhibits mammalian glutamic (Eggleston, 1958).
I. Cobaltous Ion
No effect on mammalian dopa (Perry et al, 1955).
J. Aluminum Ion
Slight (Krebs, 1950) or no effect (Arjona et al, 1950) on bacterial (Eg
gleston, 1958); inhibits plant glutamic (Eggleston, 1958).
K. Stannic Ion
Bacterial in some only (Eggleston, 1958).
L Zinc Ion
Slight or no inhibition of mammalian dopa (Fellman, 1959).
M. Cadmium Ion
Bacterial glutamic (Koizumi et al, 1958); no effect on mammalian dopa (Fellman, 1959).
9. I N H I B I T I O N OF A M I N O ACID DECARBOXYLASES 329 N. Manganic Ion
Slightly inhibits bacterial at relatively high concentration (Krebs, 1950;
Nash, 1952); no effect on mammalian dopa (Buzard and Nytch, 1957a, b) nor on histidine (Werle and Krautzun, 1938) except at neutrality, where it does inhibit.
O. Metal Chelating Agents
Little or no effect on bacterial (Gale and Epps, 1944; Krebs, 1950; Dewey et al., 1954; Eggleston, 1958); inhibits bacterial (Ekladius et al., 1957;
Koizumi et al., 1958); no effect on plant glutamic (Eggleston, 1958); no effect on mammalian dopa (Hartman et al., 1955; Perry et al., 1955; Fell- man, 1959); cysteic, cysteinesulfinic, glutamic (Davison, 1956b; Sorbo and Heyman, 1957), inhibits 5-HTP (Beiler and Martin, 1954).
P. Carbon Monoxide
No effect on bacterial enzymes tested (Gale and Epps, 1944), plant glutamic (Beevers, 1951); nor on mammalian histidine (Werle and Heitzer, 1938) and dopa (Blaschko, 1942a), suggesting an absence of copper and iron requirement.
Q. Azide
Little effect on bacterial, except lysine (Gale and Epps, 1944; Taylor and Gale, 1945; Krebs, 1950); nor on mammalian dopa (Blaschko, 1942a;
Perry et al, 1955).
R. Metaphosphate and Pyrophosphate
May inhibit bacterial tyrosine at relatively high concentrations (Sloane- Stanley, 1949a; Krebs, 1950). No effect on mammalian histidine (Werle, 1943b) or dopa (Perry et al, 1955).
S. Sulfides
Inhibit bacterial glutamic (Krebs, 1950) and tyrosine, but histidine only slightly, and lysine not at all (Epps, 1944, 1945; Gale and Epps, 1944);
inhibit mammalian histidine (Werle, 1947), and dopa slightly (Blaschko, 1942a).
T. Dimercaprol (2,3-Dimercapto-l-Propanol; BAL)
Wenzel and Beckloff (1958) present data suggesting that dimercaprol inhibits kidney dopa decarboxylase in renal hypertensive rats more than in normal rats and speculate on a possible role of a metal activator (cf.
also Section X I I ) . In some of these studies intact cells were used and in others, cell-free extracts and semipurified preparations. This could be critical, as could dialysis.
In general, it may be concluded that the amino acid decarboxylases are sensitive to heavy metals but that most, but not all, undialyzed prepara
tions are not metal-dependent.
Thus, the general inhibitions seen with heavy metals probably are due to inactivation of essential thiol groups, and the lack of effect of heavy metal binding and chelating agents suggests that most, but not all (see Section X I , A ) , amino acid decarboxylases, especially mammalian, are not metal-dependent (cf. Section X I I ) .
X. CYANIDE (cf. Sections XI, Β; XIII)
All amino acid decarboxylases studied are cyanide-sensitive but in widely different degrees, including bacterial (Gale, 1941; Gale and Epps, 1944; Epps, 1944, 1945; Taylor and Gale, 1945; Krebs, 1950; Dewey et al, 1954; M0ller, 1954, 1956; and Kauffmann and M0ller, 1955, who used growth and decarboxylation inhibition by cyanide for bacterial classification; Ekladius et al, 1957; Ekladius and King, 1956; Gupta et al, 1960, who used it for classification); plant (Okunuki, 1937, 1942, 1943;
Werle and Raub, 1948; Werle and Peschel, 1949; Morrison, 1950; Cheng et al, 1960); animal, including insect (venom gland of bee) histidine de
carboxylase (Werle and Gleissner, 1951); and aminomalonic acid decar
boxylase of silk gland in silkworms (Shimura et al, 1956); and mammalian histidine (Werle and Hermann, 1937; Werle and Krautzun, 1938; Werle, 1940-1941); dopa (Holtz et al, 1939; Holtz and Heise, 1938; Imiya, 1941;
Blaschko, 1942a; Clegg and Sealock, 1949; Schayer and Kobayashi, 1956);
dops (Werle and Peschel, 1949); aminomalonic (Shimura et al, 1956); and cysteic, cysteinesulfinic, and glutamic acids [Blaschko, 1942b; Wingo and Awapara, 1950; Davison, 1956b; Tursky, 1960 (in vivol); Simmonet et al, I960]. Werle and Peschel (1949) reported that 10~
2
M C N inhibits and 10~
3
M stimulates the decarboxylation of p-hydroxyphenylserine ; 10~
4 M has no effect. No explanation was offered, but the very low rates of reac
tion, measured manometrically, make the data questionable. The effect should be re-examined by spectrofluorimetric or radioisotopic methods.
9. INHIBITION OF AMINO ACID DECARBOXYLASES 331 The inhibition does not occur by removing an essential metal and is reversible in some cases by dialysis. Cyanide is generally classified as a carbonyl group inhibitor (see Braunstein, 1960). Bona vita (1959, 1960a, b), Bonavita and Scardi (1958a, b, 1959b), and Scardi and Bonavita (1957a, b, 1958) demonstrated that cyanide reacts with the 5-formyl group of B6- P 0 4 to form a cyanhydrin. The addition of cyanide to apotransaminase or apodecarboxylase preincubated with B6- P 0 4 caused no inhibition, and Β6-Ρθ4 cyanhydrin is ineffective in activating the apoenzymes. Bonavita
(1960b) reviews the enzymatic implications. Their work provides an ex
planation of apoenzyme-coenzyme site reaction in B6-dependent enzymes by claiming that the 4-formyl group of B6-PO4 is involved in binding to the apoenzyme, rather than forming a Schiff base with the amino acid substrate, as previously suggested by Schlenk and Fischer (1947) and elaborated upon by Metzler, Ikawa and Snell (1954), Braunstein (1960), and others (cf. Sections X I and X I I I ) .
XI. PYRIDOXAL-5-PHOSPHATE (CODECARBOXYLASE), VITAMIN B6/ B6-DEFICIENCY, B6-ANTAGONISTS
(cf. Section XIII)
Since all amino acid decarboxylases are B6-P04-dependent, all agents and conditions which inhibit by inactivating, limiting, or removing the coenzyme must be considered, including B6-deficiency.
A. B6-Deficiency and Antagonists
Most amino acid decarboxylations carried out by bacteria are inhibited in B6-deficient media, the individual differences being due to different apoenzyme-coenzyme affinities.
The exquisite sensitivity of tyrosine apodecarboxylase of bacteria grown in a Be-deficient medium has been used by most investigators for estimating B6-P04 (Sloane-Stanley, 1949a).
Inhibition of amino acid decarboxylations by phosphorylated pyridoxine analogues in microorganisms and animals in vitro and in vivo has been demonstrated repeatedly. Beiler and Martin (1947) showed that phos
phorylated 4-deoxypyridoxine inhibits tyrosine decarboxylase in S. faecalis partially purified by the method of Epps (1944, 1945), but because Martin and Beiler (1947) found that 4-deoxypyridoxine phosphate did not inhibit dopa decarboxylation by rat kidney in vitro, while certain folic acid ana
logues did (cf. Section X V , E ) , they suggested that a folic acid derivative rather than B6- P 04 might be the coenzyme. Umbreit and Waddell (1949)
showed that 4-deoxypyridoxine inhibits tyrosine decarboxylase of bacteria by first being converted to its phosphorylated analogue, which then com
petes with B6-PO4 for the apoenzyme.
Meadow and Work (1958) showed that ω-methylpyridoxamine phos
phate partially inhibits decarboxylation of diaminopimelic acid by Bacillus sphaericus asporogenes in the presence of B6-PO4, but activates it in the absence of the latter.
Snell (1958) has reviewed chemical structure in relation to the biological activities of B6 and its analogues and the mechanism of action of Ββ-Ρ04 in B6-dependent enzymes. Original data are given, which show that of several derivatives tested on S. faecalis tyrosine apodecarboxylase, those which inhibited include pyridoxine phosphate, 4-deoxypyridoxine phos
phate, 3-amino-4,5-dihydroxymethyl-2-methylpyridine phosphate, and pyridoxamine phosphate. Olivard and Snell (1955) showed that ω-methyl- pyridoxal phosphate can replace B6-PO4 in alanine-glutamic acid trans
aminase of bacteria but not cysteine desulfhydrase. Recently, Matsuda and Makino (1961b) showed that pyridoxal-5-si^/ate competitively in
hibits the B6-P04 activation of glutamic acid decarboxylase of brain homogenates of B6-deficient mice. Present evidence suggests that the affinity of tyrosine-glutamic transaminase for its substrate, like that of alanine race- mase, is decreased when the ω-analogue replaces B6-PO4 as coenzyme.
Blaschko et al (1948, 1951, 1953; Blaschko, 1950) were the first to show decreased amino acid decarboxylase activity in B6-deficient ani
mals (rats). Cysteic acid decarboxylase is more sensitive than dopa, and activity is not completely restored by the addition of B6-PO4 in vitro, implying an impairment of apcenzyme synthesis (cf. Sections V and V I I ) . Although Be-deficiency itself had no effect on adrenal medullary cate
chol amine content, resynthesis was slower than normal after depletion by insulin hypoglycemia. Dietrich et al. (Dietrich and Shapiro, 1953;
Dietrich and Borries, 1956) confirmed this and found that dopa decar
boxylase in mouse liver is less sensitive to B6-deficiency than glutamic- aspartic transaminase and cysteine desulfurase. West (1953) also found that Be-deficiency decreases dopa decarboxylase in rat liver and kidney.
In confirmation of Blaschko et al. (1951), this causes a slower resynthesis of adrenal medullary catechol amines after depletion by insulin hypo
glycemia. Sourkes et al. (1960) confirmed Blaschko et al. and West that Be-deficiency does not alter catechol amine levels in the adrenals and extended this to show the same lack of effect in other organs. Concomitant B2-deficiency caused a marked decrease, especially in brain and liver.
Sourkes et al. (1960) also showed less excretion of dopamine after injecting dopa in Ββ-deficient rats.
Pogrund and associates (1955, 1961; cf. Clark, 1959 and Clark and Pogrund, 1961) showed that B6-deficient rats exhibit subnormal responses
9. I N H I B I T I O N OF A M I N O ACID D E C A R B O X Y L A S E S 333 to dopa and to 3-hydroxy- and 3,4-dihydroxyphenylpyruvic acids, which are transaminated in vivo to m-tyrosine and dopa, respectively. Dopamine responses remained normal. Kidney apodopadecarboxylase remained normal, liver apoenzyme decreased, and coenzyme was low in kidney and liver, the former more so, in confirmation of Beaton and McHenry (1953) and Wachstein and Moore (1958).
Roberts et al. (1951) found a 50% decrease in saturation of glutamic acid apodecarboxylase with coenzyme in B6-deficient rat brains. This could be completely reversed by feeding B6. This has been confirmed by Bergeret et al. (1955), Rosen et al. (1959), and others, and the literature is reviewed by Elliott and Jasper (1959), Roberts (1960), Roberts and Eidelberg (1959), Roberts et al. (1960), and Burns and Shore (1961).
Chatagner and co-workers (1954), Bergeret et al. (1955), and Fromageot (1956) confirmed Blaschko et al. that B6-deficiency decreases cysteic acid and cysteinesulfinic acid decarboxylation in rats and rabbits, and both Fromageot (1953-1954) and Chatagner (1959) cover this work in their re
views on sulfur metabolism. Hope (1955, 1957) also confirmed this and reviews the field (1959). Hope presented evidence that glutamic, cysteic, and cysteinesulfinic acids are decarboxylated by the same enzyme in mammals.
Marco (1957) observed that in comparison with in vitro observations, the normal rat brain B6-PC>4 content is lower than that which will allow maximum decarboxylation of glutamic, cysteic and cysteinesulfinic acids.
Massive doses of B6 given in vivo increase it somewhat. Holtz (1959, 1960a) noted a linear activation of dopa decarboxylase activity by B6-P04 in brain homogenates, which does not occur in homogenates and extracts of other organs. Addition of brain extracts to dopa decarboxylase prepara
tions from liver or kidney reactivate the decarboxylation after it has leveled off, in the presence of B6-P04. He postulated that the ethanolamine moiety of brain cephalin, through its amine group, forms a Schiff base with B6-P04, which acts as a coenzyme, much as the B6-PO4 hydrazones studied by Gonnard (see Section X I , B, below).
Buxton and Sinclair (1956) found low 5-HTP decarboxylase activity in Be-deficient rats, restored in vitro by B6-P04. This was confirmed indirectly by Weissbach et al. (1957), who found lower levels of 5-HT in tissues of Be-deficient chicks. Exogenously administered 5-HTP was decarboxylated at a much lower rate in B6-deficiency (cf. Udenfriend et al, 1957). Buzard and Nytch (1957a, b) showed that supplementary B6 in the diet increases 5-HTP decarboxylation by rat kidney, and B6-deficiency reversibly de
creases both apoenzyme and coenzyme, especially the latter.
Schrodt et al. (1960) attempted to induce 5-HTP decarboxylase in
sufficiency with 4-deoxypyridoxine in two patients with a carcinoid syn
drome. There was no change of urinary 5-HIAA excretion, nor in the
symptoms in one case, while in the other the excretion was less. It does not follow, however, that 5-HTP decarboxylation was necessarily decreased in this patient.
Kizer and Chan (1961) found that 5-HTP decarboxylase activity is absent in transplanted and primary rat hepatoma tissue and in pre
cancerous liver tissue of rats treated with carcinogens, the rate of its loss in the latter being correlated with carcinogenesis.
Schayer (1959) showed that tissues of B6-deficient rats have a lowered histidine decarboxylase activity which is partially restored by feeding B6 or by adding B6-P04 in vitro. These B6-deficient rats excreted less his- tamine-C
14
after the injection of histidine-C 14
. Nadkarni and Sreenivasan (1957), using rat liver homogenate, could obtain no decrease in decar
boxylation of serine to ethanolamine in B6-deficiency in comparison with controls. The results would have been more significant if data on the capacity of these livers to decarboxylate other amino acids had been examined. Indirect reports of possible impairment of decarboxylation in Be-deficiency by measurements of tissue amine levels and/or excretion are fraught with uncertainty because of possible effects on apoenzyme syn
thesis, pyridoxal kinase, and on uptake, transport and binding of amines (e.g., Yeh et al., 1959; Ferrari et al., 1957). Also, claims of decreased de
carboxylation in Be-deficiency by even more indirect methods are open to criticism. Thus, Martin (1946) postulated that tyrosine is less toxic in Be-deficiency because of less decarboxylation to tyramine, a toxic amine, yet no direct measurements were made. Similar doubt is cast upon at
tempts to correlate decreased tissue amines with decreased amino acid decarboxylase activity induced by other types of treatment, such as lethal X-irradiation (Anderson et al., 1951).
The reviewer wishes to mention the elegant work of Shukuya and Schwert (1960a, b, c), who have purified bacterial glutamic acid decar
boxylase to the extent that its molecular weight could be estimated. Their studies of its characteristics and kinetics have thrown much light on the mechanism of apoenzyme-coenzyme dissociation and the nature and reactivity of the apoenzyme molecule.
B. B6- P 04 Hydrazones
Certain hydrazides may inactivate B6-P04 by carbonyl group reaction (cf. Section X I I I ) . Gonnard et al. (Gonnard, 1958; Gonnard and Boigné, 1961; Gonnard and Chi, 1958, 1959a, b; Gonnard and Nguyen-Philippon, 1959, 1961) have found that synthetic hydrazones, formed by reacting B6-P04 with the hydrazides of isonicotinic acid (for methods, cf. Curry and Balen, 1960; Testa et al., 1961), benzoic acid, nicotinic acid, picolic
9. I N H I B I T I O N OF A M I N O ACID DECARBOXYLASES 335 acid, p-aminosalicylic acid, and thienic acid, all activate mammalian dopa decarboxylase, glutamic-aspartic transaminase, and kynureninase, all of which are B6-P04-dependent enzymes. Unlike B6-P04 alone, which permits only incomplete decarboxylation of dopa, these hydrazones, also acting as true cofactors (without dissociating), allow the reaction to go to com
pletion, though at a somewhat lower rate. Palm (1958), working with Β6-Ρθ4 isonicotinoylhydrazone, believed that the product is slowly hy
drolyzed to give B6-P04, which allows the reaction to go to completion, although more slowly. Since the aldehyde moiety of the B6-P04 is blocked in the reaction product, Gonnard believes this may be additional evidence against the concept of a mechanism of transamination and decarboxylation involving the formation of Schiff bases between coenzyme and substrate.
The phenomenon also was confirmed by Bonavita and Scardi (1959a, b), who found that synthetic B6-PO4 isonicotinoylhydrazone activated glu- tamic-oxalacetic transaminase of pig heart, although longer incubations were required than with B6-P04 alone. Biehl and Vil ter (1954) and Davison (1956c) had shown that isonicotinic acid hydrazide ( I N H ) inhibits B6-de- pendent enzymes due to hydrazone formation with B6-P04. Hence, Bona
vita and Scardi believed the hydrazone should be inactive as a cofactor, like the B6-P04 cyanhydrin they had studied previously (cf. Section X ) , and were surprised when activation occurred. In seeking an explanation, they considered that (1) B6-PO4 is liberated from the hydrazone, but pointed out that this is unlikely because of the stability of the complex;
and although Youatt (1958) found some release of unphosphorylated pyridoxal from its isonicotinoylhydrazone by tubercle bacilli, there is no evidence that the hydrazone is enzymatically split; (2) the B6-P04 hydra
zone itself is bound as such to the apoenzyme with a subsequent displace
ment reaction with substrate to yield holoenzyme and isonicotinic acid.
This does not rule out the possibility of its reactivity with amino acids. To explain the differences between the B6-P04 hydrazone and B6-P04 cyan
hydrin, it was pointed out that although their fluorescent spectra are similar, the pif</ values of the phenolic hydroxyl groups and pyridine nitrogen are higher for the cyanhydrin. Bonavita (1960b) reviews the mechanism of interaction of B6-PO4 and its isonicotinoylhydrazone and cyanhydrin with several apoenzymes, and questions Snell's criticism (1958) of his previous work. The reader is referred to their polemic.
C. Toxopyrimidine
Toxopy rimidine (2-methyl-4-amino-5-hy droxymethylpy rimidine), the pyrimidine component of thiamine, was first described as a convulsant in animals by Abderhalden (1939a, b, 1940, 1954). This compound and
several of its analogues compete with B6-PO4, and Snell (1958) classifies it with Be-analogues which are antimetabolites after being phosphorylated by pyridoxal kinase. This classification is undoubtedly correct, since its phosphate has been synthesized and found to inhibit semipurified B6-de- pendent apoenzyme systems whereas the unphosphorylated form does not
(Koike, 1954; Makino and Koike, 1954a, b; Haughton and King, 1957).
The toxopyrimidine analogues also act as competitive B6-antagonists in general, both in the growth of microorganisms and on their amino acid decarboxylases (Sakuragi and Kummerow, 1957; Scheunert et al, 1957;
Shintani, 1956). A considerable literature has grown on the essential chemi
cal structure for activity of these compounds because of their potency (Abderhalden, 1954; Kawashima, 1957; Shintani, 1957a, c; Miyake, 1957;
Hayashi, 1957; Nishizawa et al, 1958a; Rindi et al, 1959b). A methyl group appears necessary in the 2-position, an amino in the 4-position, and a radical easily changed to a hydroxymethyl in the 5-position. Abderhalden (1954) originally found antitoxopyrimidine activity in extracts of grains, yeast, etc., and the active factor later was shown to be pyridoxamine.
Papers on "atoxopyrimidine" activity of various B6-derivatives have ap
peared (Morii, 1941; Abderhalden, 1954; Makino et al, 1954; Makino and Koike, 1954a, b; Makino and Kinoshita, 1955; Sakuragi and Kummerow, 1957; Scheunert et al, 1957; Konishi, 1957; Hayashi, 1957; Nozaki, 1958;
Rindi et al, 1959a, b; Rindi and Ferrari, 1959). In addition to pyridoxine, pyridoxal, pyridoxamine, and B6-P04, antagonists of toxopyrimidine effects in vivo include 2-methyl-4-hydroxy-4-hydroxy(or amino) me thy lpyrimidine, and 2-methyl-4-amino-5-f ormylaminomethyl (or 5-aminomethyl) pyrimidine, which Hayashi (1957) designated the "atoxopyrimidine" group of ana
logues. A clue to the metabolism of such compounds was afforded by Shintani (1957b), who found that hydroxymethylpyrimidine administered to rabbits is metabolized and excreted as the 5-carboxy analogue. In animals, these substances are highly potent agents, producing convul
sions ("running fits") and death in small doses. Yamanaka (1957) claims that the site of the convulsive effect is centered in the nuclei lenticu- laris and caudatus of the brain stem. Liver pathology is caused in doses as small as 4 Mg/gm in mice (Kooka, 1957; Nishizawa et al, 1957), which is correlated with decreased pyridoxine, pyridoxamine, and pyridoxal in the liver. It is prevented by treatment with these compounds, and also by such agents as methionine and iV-carbobenzoxyglutamylcholine.
Among the various B6-enzymes which are competitively inhibited by toxopyrimidine in vivo are glutamic acid decarboxylase of liver and brain, and several transaminases (Mandokoro, 1957; Namba, 1957; Yamanaka, 1957; Konishi, 1957; Moriya, 1958; Kobayashi, 1958; Sagawara, 1958;
Nozaki, 1958; Nishizawa, 1958; Nishizawa et al, 1958b, c, d, 1959a, b, c, d, 1960; Rindi et al, 1959a, b, c), and the convulsive symptoms correlate
9. I N H I B I T I O N OF A M I N O ACID DECARBOXYLASES 337 better with decreased decarboxylation to gaba than with other metabolic pathways examined. Purified pyridoxal kinase of yeast is activated by toxopyrimidine (Moriya, 1958) and thus does not compete with pyridoxal, although it also is a substrate.
D. Steroids
Steroids may be powerful inhibitors of Ββ-dependent enzymes through interaction with B6- P 04 (cf. Section X I V ) .
XII. ACTIVATORS, STABILIZERS, AND COFACTORS OTHER THAN B6-P04
A. Metals
Heavy metal ions have been reported to activate some amino acid decarboxylases in plants and animals (Guirard and Snell, 1954; Happold, 1956; Eggleston, 1958; Mazelis, 1959), but there are conflicting reports on their activation of amino acid decarboxylases in mammals (Sorbo and Heyman, 1957; Steensholt et al., 1956). The subject has been reviewed by Dixon and Webb (1958, p. 447) and Braunstein (1960), the latter con
cluding that the experimental evidence for an involvement of metals in activating Ββ-enzymes in general is highly contradictory.
B. Surfactants
Ionic detergents have been reported to inhibit some amino acid de
carboxylases, to stimulate some, and to have no effect on others. In some, but not all cases, these effects may be through permeability effects when intact cells are involved (Krebs, 1948; Nossal, 1952; Oliver, 1952; Hughes 1949, 1950; Storck, 1951; Cosin, 1955, 1956; Yabe et al, 1957; Eggleston, 1957; Crawford, 1958; Baker et al, 1941), and Hughes postulated that surfactants increase the affinity of apoenzyme for substrate or remove an inhibitor by complex formation.
C. Solvents
The effect of solvents in general have been studied on various amino acid decarboxylases (cf. Krebs, 1948, 1950; Mardashew, 1949; Mardashew and Etinghof, 1948; Holtz and Heise, 1938; Shimada et al, 1954; Blaschko, 1942b, c).
Waton (1956a, b) made the interesting observation that many organic solvents activate nonspecific mammalian histidine decarboxylation by rabbit kidney mince and extracts. Pyridine, toluene, benzene, chloro- benzene, and petroleum ether accelerated 6-9 times; chloroform, cyclo- hexane, nitrobenzene, aniline, ether, and carbon tetrachloride 2-3 times;
butanol, tetrachlorethane, pentan-l-ol, cyclohexanone, and acetone slightly; amyl alcohol not at all. The amount used was one drop in the incubation mixture. The mechanism is postulated to be independent of antibacterial action since the incubations were usually an hour or less.
Most of the active solvents have aromatic rings, and the action is inde
pendent of the formation of two phases. Waton speculated that perhaps there is a physical effect on the enzyme, such as a breaking up of micelles and making previously unaccessible sites available for activity. Hughes' (1949, 1950) suggestion of the micellar mode of action of surfactants on enzymes is not acknowledged, and a full explanation of the phenomenon remains to be established. It is possible that the effect is due in part to exposing available enzyme sites by removal of lipoprotein barriers.
Schayer (1957) confirmed Waton that benzene markedly enhances histidine decarboxylase by rat kidney, in this case cell-free, but strongly inhibits that of rat stomach. Rothschild and Schayer (1959) also found that histidine decarboxylase of rat peritoneal mast cells was similar to that of stomach and was different from that of kidney in that benzene markedly inhibits its activity. This, in addition to finding different pH optima, led Schayer to postulate that there are two different enzymes, as discussed elsewhere in this chapter.
D Phosphates and Arsenate
Sloan-Stanley (1949a) and Sorm and Tursky (1954) showed that phos
phate and pyrophosphate, among other anions, affect the reaction between bacterial tyrosine decarboxylase apoenzyme and coenzyme. Several workers have shown that inorganic phosphate and arsenate may accelerate some bacterial, plant, and mammalian amino acid decarboxylases (Gon
nard, 1951; Krebs et al., 1955; Hartman et al., 1955; Eggleston, 1957), and the possibility of an intracellular organic phosphorylated regulator, other than B6-P04, of the B6-dependent decarboxylases should be considered.
E. Miscellaneous
Sankar and Bender (1960) found that D-lysergic acid diethylamide (LSD) stimulates the decarboxylation of glutamic acid by cerebral cortex homogenates, and suggest that LSD may help channel the metabolism of glutamate via the gaba pathway.
9. I N H I B I T I O N OF A M I N O ACID DECARBOXYLASES 339 XIII. CARBONYL REAGENTS AND INHIBITORS ACTING ON CO
ENZYME OTHER THAN CYANIDE, SUBSTRATE ANALOGUES, AND PYRIDOXINE ANTAGONISTS.
A. Hydroxylamine, Hydrazides, Semicarbazide, Sulfite, Hydrazine, Oximes, Etc.
As is usual with Independent enzymes, all amino acid decarboxylases studied so far (fourteen) in bacteria and other microorganisms including yeast are inhibited, usually reversibly, through B6-P04 inactivation, by the usual carbonyl reagents, such as bisulfites, thiosulfate, hydroxylamine, hydrazides, oximes, hydrazines, carbazides, and their derivatives, including the hydrazide amine oxidase inhibitors. The degree of inhibition varies widely, depending on the apoenzyme-coenzyme affinity (Cattanéo-Lacombe and Senez, 1956; Crawford, 1958; Dewey et al., 1954; Ehrismann and Werle, 1948; Ekladius and King, 1956; Ekladius el al, 1957; Epps, 1945;
Gale, 1941; Gale and Epps, 1944; Gangadharam and Sirsi, 1956; Hicks and Clarke, 1959; Hoare, 1956; Kating, 1954; Knivett, 1954; Krebs, 1950;
Krishnaswamy and Giri, 1956; Aoki, 1957a; Mauron and Bujard, 1960;
Miyaki et al, 1959; Nash, 1952; Roberts, 1952b; Saito, 1957; Schormuller and Leichter, 1955a, b; Senez et al., 1959; Shukuya and Schwert, 1960a;
Sutton and King, 1960; Taylor and Gale, 1945; Wachi et al, 1959; Willett, 1958; Yamagami, 1958; Yoneda and Asano, 1953).
Plant glutamic acid decarboxylase also is inhibited by hydrazines, hydrazides, hydroxylamine, and bisulfite (Okunuki, 1942, 1943; Werle and Bruninghaus, 1951; Schales and Schales, 1946a; Beevers, 1951; Matsuda et al, 1955; Rohrlich and Rasmus, 1956; Rohrlich, 1957; Cheng et al, 1960), as are γ-methyleneglutamic acid decarboxylase (Fowden, 1954) and histidine decarboxylase (Werle and Raub, 1948).
Similarly, hydroxylamine and semicarbazide inhibit insect aminomalonic (Shimura et al, 1956) and histidine (Werle and Gleissner, 1951) decar
boxylases; they and hydrazine also inhibit bird (embryo) (Simonnet et al, 1960) and mammalian cysteic and cysteinesulfinic acid decarboxylation in vitro and in vivo (Werle and Bruninghaus, 1951; Canal and Garattini, 1957; Davison, 1956a, c). Davison (1956c) also showed the latter to be inhibited by isonicotinoyl hydrazide ( I N H ) . Nadkarni and Sreenivasan (1957) reported that hydroxylamine inhibits serine decarboxylation by rat liver homogenate. Werle and Heitzer (1938) and Werle (1940), who were the first to describe the effect on such enzymes by carbonyl reagents, re
ported the inhibition of histidine decarboxylase of mammalian tissues (which Werle discovered) by hydroxylamine, semicarbazide, bisulfite, a series of hydrazides including Girard's reagents, hydrazine, guanyl- hydrazone derivatives including the antihypertensive drug 1-hydrazinoph-
thalazine (Apresoline) and some of its analogues (cf. dopa decarboxylase below). Schayer and Kobayashi (1956), Schayer (1957), and Rothschild and Schayer (1959) found that histidine decarboxylase of rabbit platelets, rat mast cells, and other tissues is reversibly inhibited by such agents as hydroxylamine and semicarbazide; Kahlson et al. (1958, 1960a, b; Kahlson and Rosengren, 1959a, b; Kahlson, 1960) found that hydrazine and semi
carbazide inhibited the enzyme in vitro and in vivo in rats, including embryos. Pyridoxine deficiency increases the inhibition. They causally invoke histamine in embryonic, regenerative, and malignant growth, using such inhibition as a tool in their thesis (cf. Section V I I ) .
Mammalian dopa decarboxylase has been known to be inhibited by carbonyl reagents since Imiya (1941) demonstrated the effect with bi
sulfite in tissues of three species in vitro. Semicarbazide inhibits in vitro (Werle, 1943b; Clegg and Sealock, 1949; Langemann, 1951; Fellman, 1959), as do hydroxylamine (Schales and Schales, 1949; Davison, 1956c), hydrazine (Gonnard, 1951), and I N H (Davison, 1956c; Palm, 1958), although Canal and Garattini (1957) could obtain no inhibition with the latter on guinea pig and rat kidney dopa decarboxylase in vivo, in contrast with decarboxylation of cysteic acid by liver, and glutamic acid by brain.
Hydrazine derivatives known to inhibit monoamine oxidase also have been shown to inhibit dopa decarboxylation in vitro, including iproniazid by chicken adrenal gland (Hagen and Welch, 1956) and l-phenyl-2- hydrazinopropane ("JB-516," Catron) by rabbit brain (Brodie et al, 1958,
1959). Levy and Michel-Ber (1960a, b), from indirect pharmacological evidence based on the reversal by B6-P04 of the potentiation of hypnotics by iproniazid and hydrazino-2-octane when 5-HTP or dopa are given, conclude that these monoamine oxidase inhibitors block dopa (and 5-HTP) decarboxylase in vivo. Direct measurements in vivo, however, by Brodie et al. (1958, 1959) and others have shown that iproniazid and l-phenyl-2- hydrazinopropane do not inhibit these enzymes in vivo (but cf. Canal and Maffei-Faccioli, 1959, 1960). In the reviewer's opinion, however, this sub
ject deserves a systematic study, since inhibition by Catron and iproniazid of aromatic amino acid decarboxylation in vitro and in vivo has been observed repeatedly in this laboratory.
Perry et al. (1955) reported the inhibition of mammalian dopa decar
boxylase in vitro by low concentrations of the. antihypertensive agents 1-hydrazinophthalazine (Apresoline), 1,4-dihydrazinophthalazine (Nep- resol), and 2-hydrazino-5-phenylpyridizine. This was confirmed and ex
tended by Werle et al. (1955), who examined the effects of twelve hydrazine and guanylhydrazone derivatives, including aminoguanidine, on mono
amine and diamine oxidases and on histidine and dopa decarboxylases by mammalian preparations in vitro. Schayer and associates (1955) found no
