Structure-activity studies of L-canaline-mediated inhibition of porcine alanine aminotransferase

Article abstract of DOI:10.1021/tx9600199

L-Canaline [L-2-amino-4-(aminooxy)butanoic acid] (L-CAN) and a family of eleven structurally related analogs were synthesized and evaluated for their inhibitory effect on PLP-dependent alanine aminotransferase (AlaAT) (EC 2.6.1.2) obtained from porcine heart. These congeners were selected to determine the stereochemical, aliphatic chain length, and aminooxy substitutional effects on L-CAN-mediated inhibition of AlaAT activity. L-CAN was the most effective inhibitor of the tested compounds; 10^-7 M L-CAN elicited a 55% reduction in AlaAT activity after a 5 min exposure. This deleterious effect results from the ability of L-CAN to react avidly with the PLP moiety of the enzyme to form a stable, L-CAN-PLP oxime. In contrast, the methyl and ethyl esters of L-CAN reduced AlaAT activity by only 8% and 6%, respectively. While all of the L-enantiomeric forms of the tested compound were more potent AlaAT inhibitors than their corresponding D-stereoisomers, the D-enantiomers, particularly D-canaline, were active. Chain shortening or lengthening dramatically curtailed L-CAN-mediated loss in AlaAT activity, but the replacement of the α-amino group with a hydrogen was of little consequence in this regard. AlaAT was treated with L-CAN in the presence of free PLP to assess PLP capacity to protect AlaAT against 10^-7 M L-CAN- dependent inactivation. L-CAN retained approximately two-thirds of its inhibitory ability in the presence of equimolar PLP, but AlaAT inhibition was reduced 90% by a 10-fold excess of PLP over L-CAN.

Full text of DOI:10.1021/tx9600199

Chem. Res. Toxicol. 1996, 9, 1293-1297
1293
Str u ctu r e-Activity Stu d ies of L-Ca n a lin e-Med ia ted
In h ibition of P or cin e Ala n in e Am in otr a n sfer a se
David R. Worthen,† Darian K. Ratliff,‡ Gerald A. Rosenthal,*,‡ L. Trifonov,§ and
Peter A. Crooks†,§
Graduate Center for Toxicology, Laboratory of Biochemical Ecology, and College of Pharmacy,
University of Kentucky, Lexington, Kentucky 40506-0054
Received February 5, 1996X
L-Canaline [L-2-amino-4-(aminooxy)butanoic acid] (L-CAN) and a family of eleven structurally
related analogs were synthesized and evaluated for their inhibitory effect on PLP-dependent
alanine aminotransferase (AlaAT) (EC 2.6.1.2) obtained from porcine heart. These congeners
were selected to determine the stereochemical, aliphatic chain length, and aminooxy
substitutional effects on L-CAN-mediated inhibition of AlaAT activity. L-CAN was the most
effective inhibitor of the tested compounds; 10-7 M L-CAN elicited a 55% reduction in AlaAT
activity after a 5 min exposure. This deleterious effect results from the ability of L-CAN to
react avidly with the PLP moiety of the enzyme to form a stable, L-CAN-PLP oxime. In contrast,
the methyl and ethyl esters of L-CAN reduced AlaAT activity by only 8% and 6%, respectively.
While all of the L-enantiomeric forms of the tested compound were more potent AlaAT inhibitors
than their corresponding D-stereoisomers, the D-enantiomers, particularly D-canaline, were
active. Chain shortening or lengthening dramatically curtailed L-CAN-mediated loss in AlaAT
activity, but the replacement of the R-amino group with a hydrogen was of little consequence
in this regard. AlaAT was treated with L-CAN in the presence of free PLP to assess PLP
capacity to protect AlaAT against 10-7 M L-CAN-dependent inactivation. L-CAN retained
approximately two-thirds of its inhibitory ability in the presence of equimolar PLP, but AlaAT
inhibition was reduced 90% by a 10-fold excess of PLP over L-CAN.
Sch em e 1. Ar gin a se-Med ia ted Hyd r olytic
Degu a n id in a tion of L-Ca n a va n in e to L-CAN a n d
Ur ea
In tr od u ction
L-Canavanine, L-2-amino-4-(guanidinooxy)butanoic acid
(L-CAV),1 is a naturally occurring L-arginine analog
stored by a variety of leguminous plants (1). L-CAV is a
potent L-arginine antagonist that exhibits antimetabolic
activity in many living systems (2). CAV’s biological
effects include nitric oxide synthase inhibition (3), anti-
microbial activity (4), leukocyte chemotaxis modulation
(5, 6), autoimmune reactions (7), and insecticidal activity
(8). L-CAV may also give rise to other toxic compounds,
including vinylglyoxylate, produced after L-CAV under-
goes oxidation and subsequent â,γ-elimination (9). The
chemotherapeutic potential of L-CAV and its derivatives
has also been demonstrated in several in vitro and in vivo
models; they possess great promise as novel antineoplas-
tic agents (10-12).
L-CAV undergoes arginase (EC 3.5.3.1)-catalyzed hy-
drolytic deguanidination, producing L-canaline (L-CAN)
and urea (Scheme 1) (13). L-CAN, L-2-amino-4-(amino-
oxy)butanoic acid, a structural analog of L-ornithine,
reacts avidly with the PLP moiety of vitamin B6-contain-
ing enzymes to form a stable oxime that rapidly inacti-
vates the enzyme (14, 15). L-CAN is a potent inhibitor
of ornithine aminotransferase (EC 2.6.1.13) by virtue of
its reaction with the PLP moiety of the enzyme to form
an oxime that destroys the catalytic activity of the
enzyme (16, 17). L-CAN may also act as an antimetabo-
lite by forming stable oximes with 2-oxo acid-containing
substrates of PLP-dependent enzymes (18).
Since the utility of L-CAV as a chemotherapeutic agent
may be compromised by the inherent toxicity of its L-CAN
derivative, a more detailed understanding of the toxicity
of L-CAN and its congeners may be useful in furthering
the design of less toxic L-CAV analogs. To better under-
stand the structure-activity relationship of L-CAN in the
inactivation of PLP-containing enzymes, we synthesized
a series of L-CAN analogs and evaluated their inhibitory
effect on the PLP-dependent activity of alanine ami-
notransferase (EC 2.6.1.2) (AlaAT) isolated from porcine
heart. These compounds were designed to permit an
examination of the effects of changing the stereochem-
istry about the chiral center, varying aliphatic chain
length, removing the R-amino group, and modification
of the aminooxy moiety of L-CAN and related compounds
on PLP-dependent AlaAT activity. The capacity of free
PLP to counteract L-CAN-mediated inactivation of AlaAT
was also evaluated.
* To whom correspondence should be addressed. Phone, (606) 257-
2652; fax, (606) 257-8944; e-mail, garose@ukcc.uky.edu.
† Graduate Center for Toxicology.
‡ Laboratory of Biochemical Ecology.
§ College of Pharmacy.
Ma ter ia ls a n d Meth od s
X Abstract published in Advance ACS Abstracts, October 15, 1996.
1
Abbreviations: L-CAN, L-canaline; L-CAV, L-canavanine; PLP,
Ch em ica ls. L-CAV was isolated from an acidic methanolic
extract of J ack bean (Canavalia ensiformis) seed and purified
pyridoxal phosphate; AlaAT, alanine aminotransferase; D-CAN, D-
canaline; Tris, tris(hydroxymethyl)aminomethane; mU, milliunits.
S0893-228x(96)00019-7 CCC: $12.00 © 1996 American Chemical Society

1294 Chem. Res. Toxicol., Vol. 9, No. 8, 1996
Worthen et al.
by recrystallization (19). L-CAN, prepared by arginase-mediated
hydrolysis of L-CAV, was isolated as the dipicrate salt and
converted to the free base via L-CAN sulfate (19). All of the
remaining L-CAN derivatives were synthesized as described
elsewhere (19).2 All biochemicals used in this study, including
a suspension of porcine heart alanine aminotransferase in 1.8
M ammonium sulfate [120 units (mg of protein)-1] and rabbit
muscle lactic dehydrogenase [Type II, 960 units (mg of protein)-1
]
were purchased from Aldrich/Sigma Chemical Co. (St. Louis,
MO).
En zym e Assa y. Alanine aminotransferase activity was
assayed by reacting 100 µmol of L-alanine and 5 µmol of
2-oxoglutarate in 0.3 mL of 125 mM Tris buffer (pH 7.6) with
0.4 mL of 125 mM Tris buffer (pH 7.6) containing 26.25 mU of
AlaAT/0.4 mL for 10 min at 37 °C. After 10 min, the reaction
was terminated by immersing the assay tubes in a boiling water
bath for 1 min. After cooling on ice to 23 °C, 0.7 mL of a 440
µM solution of reduced NAD (NADH) in 250 mM Tris buffer
(pH 7.6) containing 52.8 units of rabbit muscle lactate dehy-
drogenase per 0.7 mL was added, and the mixture allowed to
stand at 23 °C for 5 min. The oxidation of NADH was monitored
at 340 nm using a Gilford Response spectrophotometer. This
coupled enzyme assay was employed to quantify AlaAT activity
by comparing the reduced coenzyme content of the samples
containing AlaAT to comparable samples containing all reagents
except AlaAT. All assays were conducted in quadruplicate
employing two independent experiments.
F igu r e 1. L-Canaline-mediated inhibition of alanine ami-
notransferase activity. AlaAT, at a concentration of 26.25 mU/
0.4 mL, was incubated in 125 mM Tris buffer (pH 7.6) containing
L-CAN at the specified concentration at 37 °C. After the specified
incubation time, a 0.4 mL aliquot of the enzyme-inhibitor
mixture, containing 26.25 mU of AlaAT, was added to 0.3 mL
of the substrate solution and assayed for enzymatic activity as
described in the text. L-CAN: 10-9 ([), 10-8 (1), 10-7 (2), 10-6
(9), and 10-5 M (b).
In h ibition Stu d ies. AlaAT inhibition studies were con-
ducted in an analogous manner. Here, the enzyme was incu-
bated in 125 mM Tris buffer (pH 7.6, 26.25 mU of AlaAT/0.4
mL) which contained the tested inhibitor at the specified
concentration. After the indicated AlaAT/inhibitor incubation
time, 0.4 mL of the enzyme-inhibitor mixture, containing 26.25
mU of AlaAT, was added to 0.3 mL of the substrate solution
and assayed for enzymatic activity as described above. Enzy-
matic activity of AlaAT exposed to the various inhibitors at each
concentration and after each exposure time was compared to
that of similarly processed AlaAT preparations that did not
contain an inhibitor.
P yr id oxa l P h osp h a te P r otection . In order to determine
the capacity of free PLP to protect AlaAT from L-CAN-directed
inhibition, free PLP at the indicated concentration was added
to the enzyme and control assay tubes and incubated for 5 min.
The specified concentration of L-CAN was then added, and the
assays were conducted as described above. These assays were
conducted in order to quantify the inhibitory effect of L-CAN
over time in the presence of various ratios of L-CAN:free PLP.
than the naturally occurring stereoisomer in curtailing
AlaAT activity. The estimated IC50 of D-CAN was ap-
proximately one-seventh that of L-CAN.
The lower homologues of D- and L-CAN, namely, D- and
L-2-amino-3-(aminooxy)propanoic acid, show minimal (ca.
20%) inhibitory activity and do not significantly differ
in their respective ability to inhibit AlaAT. The confor-
mationally restrained, five-membered aminooxy hetero-
cycle (2S,4S)-(aminooxy)proline, assayed at 10-7 M,
similarly attenuates AlaAT activity by 15% after 10 min.
At this concentration, the five-carbon linear congener L-2-
amino-5-(aminooxy)pentanoic acid, the longest analog
tested, decreases AlaAT activity by 6% after 10 min and
by 16% after 30 min.
Carboxylic acids bearing a terminal aminooxy moiety
and lacking an R-amino group are among the most active
compounds examined. Both ω-(aminooxy)propanoic acid
and ω-(aminooxy)butanoic acid display time- and con-
centration-dependent AlaAT inhibition that is compa-
rable to L-CAN. At 10-7 M, ω-(aminooxy)propanoic acid
reduces AlaAT activity 42% after 10 min and 52% after
30 min; the comparable values for ω-(aminooxy)butanoic
acid were 49% and 55%, respectively.
Deter m in a tion of th e IC50 Va lu e. The concentration of
drug required to reduce AlaAT activity by 50% after 10 min (IC50
value) was estimated from the x-intercept of a log-linear plot of
% AlaAT activity remaining vs log10 of the inhibitor concentra-
tion.
N-Substituted ω-(aminooxy) R-amino acids have no
effect on AlaAT activity. In contrast to L-2-amino-5-
(aminooxy)pentanoic acid, which contains a free terminal
aminooxy group, the N-methylaminooxy compound, L-2-
amino 5-(N-methylaminooxy)pentanoic acid, is inactive
at all tested concentrations over the entire time course
of the experiment. A similar finding is noted for L-CAV,
Resu lts
Figure 1 illustrates the typical concentration and time-
dependent inhibition of AlaAT by L-CAN. At an L-CAN
concentration greater than 10-7 M, L-CAN-mediated
inhibition of AlaAT is rapid and significant after 5 min
(Figure 1). At lower concentrations, loss of AlaAT activity
is decreased (Figure 1). The potency of L-CAN as an
AlaAT inhibitor is demonstrated by a 55% reduction in
AlaAT after a 5 min exposure to 10-7 M drug. In
contrast, the esterified L-CAN derivatives, namely, 1-
methyl-L-canaline and 1-ethyl-L-canaline, reduce AlaAT
activity only 8% and 6%, respectively. Most interesting
is our finding that D-CAN was significantly less active
the guanidinooxy-containing precursor of L-CAN.
A
summary of the chemical structures of the tested drugs
and their AlaAT inhibition activity constitutes Table 1.
Table 2 provides a comparison of AlaAT inhibition by all
of the tested drugs used in our study.
When L-CAN is added to analogous enzyme solutions
containing various concentrations of free PLP, L-CAN-
mediated inactivation of AlaAT is reduced in a PLP
concentration-dependent manner (Figure 2). Equimolar
PLP, in the presence of 10-7 M L-CAN for 10 min, reduces
AlaAT inhibition from 57% to 38%. At 10:1 and higher
ratios of PLP to L-CAN, AlaAT inhibition is reduced by
2
Na Phuket, S. R., Trifonov, L. S., Crooks, P. A., Rosenthal, G. A.,
Freeman, J ., and Strodel, W. E. (1996) Synthesis and structure-
activity studies of some antitumor congeners of L-canavanine. Drug
Dev. Res. (submitted).

L-Canaline Inhibition of Alanine Aminotransferase
Chem. Res. Toxicol., Vol. 9, No. 8, 1996 1295
Ta ble 1. Su m m a r y of Tested Com p ou n d s a n d Th eir
Estim a ted IC50
F igu r e 2. The ability of free PLP to protect against L-CAN-
dependent inhibition of alanine aminotransferase. AlaAT was
dissolved in 125 mM Tris buffer (pH 7.6) containing various
concentrations of free PLP and incubated for 5 min at 37 °C.
L-CAN was then added to give a final concentration of 10-7
M
L-CAN, and the mixture was incubated for 10 min at 37 °C. After
a 10 min incubation, a 0.4 mL aliquot of the mixture, containing
26.25 mU of AlaAT, was added to 0.3 mL of the substrate
solution and assayed for activity as described in the text. The
presence of free PLP did not affect AlaAT activity in the control
samples. The PLP:L-CAN ratio: 10 000:1 (b), 1000:1 (9), 100:1
(2), 10:1 (1), 1:1 ([), and 0.1:1 (\).
cofactor than its free aminooxy-containing analog. Chain
lengthening of L-CAN to generate L-2-amino-5-(ami-
nooxy)pentanoic acid significantly curtails inhibition of
AlaAT activity.
The presence of a carboxyl group attached to the chiral
center is important for inhibitory activity while an
attached R-amino group is not. Inhibition of AlaAT
activity is reduced when the aliphatic chain length,
relative to L-CAN, is shortened or lengthened. A bulky
cyclic congener such as (2S,4S)-(aminooxy)proline is a
poor inhibitor.
Strecker and Eliasson (21) had reported that L-CAV
is an effective aminotransferase inhibitor. Our study
reveals that L-CAV is devoid of inhibitory activity after
exposing AlaAT to a 10-5 M concentration of this com-
pound for 30 min. These researchers (21) also reported
that terminal amino carboxylates analogous to native
enzyme substrates are effective aminotransferase inhibi-
tors. Though hydroxylamine itself is an effective ami-
notransferase inhibitor (16, 17), our experimental find-
ings suggest that an unsubstituted or terminal carboxyl
group may be a key structural feature of this series of
substituted hydroxylamine AlaAT inhibitors. The ef-
fectiveness of (2S,4S)-(aminooxy)proline at reducing AlaAT
activity at 10-7 M and lower drug concentrations supports
this belief. Despite its bulky, rigid five-membered het-
erocyclic structure, which bears little structural resem-
blance to the substrates for this enzyme, its free ami-
nooxy and carboxyl groups conferred significant capacity
to inhibit aminotransferase activity, even more than had
been attributed to the parent compound, L-proline, in rat
liver (22). In addition, our intention was to utilize an
analog of (2S,4S)-(aminooxy)proline having defined or
fixed stereochemistry, i.e., transoid arrangement of the
-ONH2 and -COOH groups, in order to obtain informa-
tion on the conformation required for inhibitor binding
to the enzyme.
more than 90%. Interestingly, detectable inhibition of
AlaAT activity by L-CAN was observed when the PLP to
L-CAN ratio was as high as 10 000:1. AlaAT control
assays revealed that free PP in the reaction mixture did
not affect base-line AlaAT activity.
Discu ssion
The nonprotein amino acid L-CAN is a potent inhibitor
of AlaAT activity; it can affect profoundly PLP-containing
enzymes whose substrates bear little structural relation-
ship to L-CAN. Our experimental efforts confirm that,
regardless of substrate, stereochemical, or functional
analogy to L-CAN, a free aminooxy group is essential for
the expression of L-CAN’s antimetabolic effects. When
compared to the activity of their L-stereoisomers, the
significant loss of AlaAT inhibitory activity noted with
D-CAN and D-2-amino-3-(aminooxy)propanoic acid sug-
gests that the intact aminooxy moiety may be more
important than the stereochemistry of the chiral center.
This finding is also supported by the lack of inhibitory
activity for L-CAN derivatives with an N-substituted
aminooxy group. L-2-Amino-5-(aminooxy)pentanoic acid,
which contains a free aminooxy group, is an effective
AlaAT inhibitor. In contrast, its aminooxy N-methylated
derivative is completely inactive, regardless of drug
concentration or exposure time assayed. This methylated
compound may be devoid of inhibitory activity, or it may
simply react much more slowly with the enzyme-bound
Despite their capacity to present a freely-rotating
aminooxy group for nucleophilic attack on the enzyme-

1296 Chem. Res. Toxicol., Vol. 9, No. 8, 1996
Worthen et al.
Ta ble 2. P er cen t of Ala AT Activity Rem a in in g a fter In cu ba tion w ith 10-7 M of th e In d ica ted Com p ou n d a
incubation time (min)
compound
5
10
20
30
Group 1. Canaline and Canaline Esters
L-canaline
D-canaline
1-methyl-L-canaline
1-ethyl-L-canaline
43.5 ( 3.9
77.5 ( 3.0
92.2 ( 2.0
94.7 ( 2.1
40.7 ( 3.7
38.3 ( 2.7
72.1 ( 2.4
89.8 ( 2.0
91.1 ( 2.1
28.5 ( 2.8
67.7 ( 3.8
81.8 ( 2.4
83.7 ( 3.0
73.8 ( 3.4
91.3 ( 2.2
94.0 ( 1.8
Group 2. ω-(Aminooxy) Amino Acids
L-2-amino-3-(aminooxy)propanoic acid
D-2-amino-3-(aminooxy)propanoic acid
L-2-amino-5-(aminooxy)pentanoic acid
(2S,4S)-(aminooxy)proline
91.0 ( 2.8
96.7 ( 3.8
93.0 ( 2.1
89.0 ( 2.9
85.9 ( 1.1
79.5 ( 1.2
86.3 ( 3.6
84.6 ( 3.6
83.4 ( 2.2
73.4 ( 1.5
92.2 ( 3.4
82.2 ( 3.4
82.0 ( 2.1
89.5 ( 3.7
93.6 ( 1.6
84.7 ( 2.2
Group 3. ω-(Aminooxy) Carboxylic Acids
ω-(aminooxy)propanoic acid
ω-(aminooxy)butanoic acid
66.1 ( 4.0
51.6 ( 2.5
58.4 ( 3.2
50.9 ( 3.2
53.3 ( 3.1
50.0 ( 3.1
48.6 ( 3.2
46.5 ( 2.8
Group 4. Substituted ω-(Aminooxy) Amino Acids
L-2-amino-5-(N-methylaminooxy)pentanoic acid
L-canavanine
99.8 ( 2.6
100 ( 3.2
100 ( 2.4
98.2 ( 2.8
100 ( 2.6
98.0 ( 2.1
100 ( 2.0
97.9 ( 3.3
a
Each value represents the AlaAT activity remaining after exposure to 10-7 M drug for the stipulated time period. AlaAT activity
was assayed as described in the text. Value represents the mean ( SEM of 2 independent determinations conducted in quadruplicate.
Sch em e 2. P r op osed Mech a n ism s for L-CAN-Dir ected In h ibition of a P LP -Con ta in in g En zym ea
a
Reaction A represents the direct nucleophilic attack of the terminal aminooxy group on the enzyme-bound imine form of PLP, leading
to the production of an aminooxy aminal, hydrolysis, and formation of the stable oxime without substantial functional or stereochemical
influence on the reaction. Reaction B represents the classical trans-Schiffization reaction, where an initial attack by the R-amino group
of the amino acid on the protonated imine is immediately followed by intramolecular nucleophilic attack by the terminal aminooxy group
on the resulting protonated imine. This cyclic aminooxy aminal is rapidly hydrolyzed, giving the ultimate chemically stable oxime and
preventing further active site catalysis. This oxime does not undergo facile acid-mediated hydrolysis to the parent amino acid and PLP
due to the expected low pKa value of the imino function. Some L-CAN may undergo transamination to give 2-oxo-L-CAN and pyridoxamine
phosphate (path C). The pyridoxamine phosphate form of AlaAT would be resistant to the further action of L-CAN, and 2-oxo-L-CAN has
been shown to undergo cyclization (18) to an inactive product.
bound PLP moiety, L-2-amino-5-(aminooxy)pentanoic
acid, L-2-amino-3-(aminooxy)propanoic acid, and (2S,4S)-
(aminooxy)proline were significantly less active than
L-CAN. These observations raise an important mecha-
nistic question regarding the precise mechanism of
enzyme inhibition by L-CAN. L-CAN and its congeners
may bind enzyme-bound PLP directly with their ω-ami-
nooxy groups, as does (aminooxy)acetate (17). The strong
nucleophilic character of the aminooxy nitrogen, en-
hanced by an R-effect from lone pair electrons on the
adjacent oxygen, may favor this mechanism (Scheme 2,
path A). It is important to note that the -ONH2 group
of L-CAN is strongly nucleophilic, yet has a relatively low
pKa value (ca. pH 3). Thus, L-CAN is not protonated at
physiological pH and therefore is much more nucleophilic
than the corresponding -CH2NH2 moiety of L-ornithine

L-Canaline Inhibition of Alanine Aminotransferase
Chem. Res. Toxicol., Vol. 9, No. 8, 1996 1297
(3) Teale, D. M., and Atkinson, A. M. (1994) L-Canavanine restores
under these conditions. We propose that the oxime
resulting from the reaction of PLP and L-CAN will be very
stable due to resistance to the protonation of the -ONdC
moiety (Scheme 2). It may be less likely that L-CAN and
its congeners first combine via the enzyme-bound PLP
imine with their R-amino group in the first step of the
classical PLP-dependent enzyme trans-Schiffization reac-
tion, followed by intramolecular attack by the terminal
aminooxy group on the imine carbon to form the ultimate
terminal aminooxy-PLP oxime (Scheme 2, path B).
Beeler and Churchich (15) demonstrated that L-CAN
attacks the Schiff base linkage of the enzyme-bound PLP
moiety of cystathionase much faster than it reacts with
free aldehydic PLP, an effect also observed in other
enzyme systems (16). This finding suggests that coop-
erative binding at the active site is an important com-
ponent of the enzyme inhibitory activity of L-CAN. The
carboxyl group may be particularly important for this
cooperative effect. As was observed with (aminooxy)-
acetate (16), we demonstrate that three- and four-carbon
terminal aminooxy carboxylic acids are also highly ef-
ficacious aminotransferase inhibitors. The potent AlaAT
inhibition exerted by ω-(aminooxy)propanoic acid and
ω-(aminooxy)butanoic acid obviates the necessity of an
R-amino group for such activity in the three- and four-
carbon compounds. Indeed, such compounds with an
R-amino group in the D-configuration (L-2-amino-3-(ami-
nooxy)propanoic acid and D-CAN) are less active than
their parent terminal aminooxy carboxylic acids.
blood pressure in
a rat model of endotoxic shock. Eur. J .
Pharmacol. 271, 87-92.
(4) Robertson, A. T., Bates, R. C., and Stout, E. R. (1984) Reversible
inhibition of bovine parovirus DNA replication by aphidicolin and
L-canavanine. J . Gen. Virol. 65, 1497-1505.
(5) Belenky, S. N., Robbins, R. A., Rennard, S. I., and Gossman, G.
L. (1993) Inhibitors of nitric oxide synthase attenuate human
neutrophil chemotaxis in vitro. J . Lab. Clin. Med. 122, 388-394.
(6) Belenky, S. N., Robbins, R. A., and Rubinstein, I. (1993) Nitric
oxide synthase inhibitors attenuate human monocyte chemotaxis
in vitro. J . Leukocyte Biol. 53, 498-503.
(7) Alcocer-Varela, J ., Iglesias, A., Llorente, L., and Alarcon-Segovia,
D. (1985) Effects of L-canavanine on T cells may explain the
induction of systemic lupus erythematosus by alfalfa. Arthritis
Rheum. 28, 52-57.
(8) Rosenthal, G. A. (1991) The biochemical basis for the potent
antimetabolic effects of L-canavanine. Phytochemistry 30, 1055-
1058.
(9) Hollander, M. M., Reiter, A. J ., Horner, W. H., and Cooper, A. J .
(1989) Conversion of canavanine to alpha-keto-gamma-guanidi-
nooxybutyrate and to vinylglyoxylate and 2-hydroxyguanidine.
Arch. Biochem. Biophys. 270, 698-713.
(10) Green, M. H., Brooks, T. L., Mendelsohn, J ., and Howell, S. B.
(1980) Antitumor activity of L-canavanine against L1210 murine
leukemia. Cancer Res. 40, 535-537.
(11) Thomas, D. A., Rosenthal, G. A., Gold, D. V., and Dickey, K. (1986)
Growth inhibition of a rat colon tumor by L-canavanine. Cancer
Res. 46, 2898-2903.
(12) Swaffar, D. S., Ang, C. Y., Desai, P. B., and Rosenthal, G. A. (1994)
Inhibition of the growth of human pancreatic cancer cells by the
arginine antimetabolite, L-canavanine. Cancer Res. 54, 6054-
6058.
(13) Damodaran, M., and Narayanan, K. G. A. (1940) A comparative
study of arginase and canavanase. Biochem. J . 34, 1449-1459.
(14) Rosenthal, G. A (1981) A mechanism of L-canaline toxicity. Eur.
J . Biochem. 114, 301-304.
(15) Beeler, T., and Churchich, J . E. (1976) Reactivity of the phos-
phopyridoxal groups of cystathionase. J . Biol. Chem. 251, 5267-
5271.
(16) Rahiala, E.-L., Kekomaki, M., J anne, J ., Raina, A., and Raiha,
N. C. R. (1971) Inhibition of pyridoxal enzymes by L-canaline. Acta
Chem. Scand. 27, 3861-3867.
(17) Kito, K., Sanada, Y., and Katunuma, N. (1978) Mode of inhibition
of ornithine aminotransferase by L-canaline. J . Biochem. (Tokyo)
83, 201-206.
(18) Cooper, A. J . L. (1984) Oxime formation between R-keto acids and
L-canaline. Arch. Biochem. Biophys. 233, 603-610.
(19) Bass, M., Crooks, P. A., Harper, L., Na Phuket, S., and Rosenthal,
G. A. (1995) Large scale production and chemical characterization
of the protective higher plant allelochemicals: L-canavanine and
L-canaline. Biochem. Ecol. Syst. (in press).
(20) Rosenthal, G. A., Dahlman, D. L., Crooks, P. A., Na Phuket, S.,
and Trifonov, L. S. (1995) Insecticidal properties of some L-
canavanine derivatives. J . Food Agric. Chem. 43, 2728-2734.
(21) Strecker, H. J ., and Eliasson, E. E. (1966) Ornithine δ-transami-
nase activity during the growth cycle of Chang’s liver cells. J .
Biol. Chem. 251, 5267-5271.
An intriguing finding of our investigation is that
ω-(aminooxy)butanoic acid, containing the same number
of carbons as L-CAN, is nearly as inhibitory as L-CAN to
AlaAT activity, while chain shortening to yield L-2-amino-
3-(aminooxy)propanoic acid, the lower homologue of
L-CAN, dramatically curtails AlaAT inactivation.
Ack n ow led gm en t. This work was supported by
National Science Foundation Grant IBN-9302875 and a
Graduate Fellowship from the Graduate School of the
University of Kentucky to D.R.W.
Registr y Nu m ber s Su p p lied by Au th or . L-Cana-
vanine, 543-38-4; L-canaline, 496-93-5; arginase, 9000-
96-8.
Refer en ces
(1) Rosenthal, G. A. (1991) Nonprotein amino acids as protective
allelochemicals. In Herbivores: Their Interaction with Secondary
Plant Metabolites (Rosenthal, G. A., and Berenbaum, M. R., Eds.)
2nd ed., pp 1-34, Academic Press, San Diego, CA.
(2) Rosenthal, G. A. (1977) The biological effects and mode of action
of L-canavanine, a structural analog of L-arginine. Q. Rev. Biol.
52, 155-178.
(22) Segal, H. L., Abraham, G. J ., and Matsuzawa, T. (1968) Interac-
tion of rat liver alanine aminotransferase with L-proline. Biochem.
Biophys. Res. Commun. 30, 63-68.
TX9600199