Inhibitors of phenylalanine ammonialyase: 1-aminobenzylphosphonic acids substituted in the benzene ring

Article abstract of DOI:10.1016/S0031-9422(01)00425-3

Dextrorotatory 1-amino-3′,4′-dichlorobenzylphosphonic acid was found to be a potent inhibitor of the plant enzyme phenylalanine ammonia-lyase both in vitro and in vivo from among the ring-substituted 1-aminobenzylphosphonic acids and other analogues of phenylglycine. A structure activity relationship analysis of the results obtained permits predictions on the geometry of the pocket of the enzyme and is a basis in the strategy of better inhibitor synthesis.

Full text of DOI:10.1016/S0031-9422(01)00425-3

Phytochemistry 59 (2002) 9–21
www.elsevier.com/locate/phytochem
Inhibitors of phenylalanine ammonia-lyase:
1-aminobenzylphosphonic acids substituted in the benzene ring
Jerzy Zona,*, Nikolaus Amrheinb, Roman Gancarza
.
a
´
InstituteofOrganicChemistry,BiochemistryandBiotechnology,Wroc•awUniversityofTechnology,WybrzezeWyspianskiego27,50-370Wroc•aw,Poland
bInstitute of Plant Sciences, Swiss Federal Institute of Technology (ETH), Universitatstraße 2, CH-8092 Zurich, Switzerland
¨
¨
Received 30 July 2001; received in revised form 1 October 2001
Abstract
Dextrorotatory 1-amino-30,40-dichlorobenzylphosphonic acid was found to be a potent inhibitor of the plant enzyme phenylalanine
ammonia-lyase both in vitro and in vivo from among the ring-substituted 1-aminobenzylphosphonic acids and other analogues of
phenylglycine. Astructure activity relationship analysis of the results obtained permits predictions on the geometry of the pocket of
the enzyme and is a basis in the strategy of better inhibitor synthesis. # 2002 Published by Elsevier Science Ltd.
Keywords: Phenylalanine ammonia-lyase; 1-Aminoalkylphosphonic acids; Enzyme inhibitors
1. Introduction
recently been introduced as a potentially useful inhibitor
as well (Hashimoto et al., 2000).
Phenylalanine ammonia-lyase (PAL; EC 4.3.1.5) cat-
alyses the first committed step of phenylpropanoid
metabolism in higher plants, yielding the precursor of a
large number of compounds such as lignins, flavonoids,
coumarins, stilbenes, benzoic acid derivatives, and even
an alkaloid, with diverse biological functions (Hahl-
brock and Scheel, 1989). We have previously identified
2-aminoindan-2-phosphonic acid (AIP) as a potent
inhibitor of phenylalanine ammonia-lyase (Zon and
Amrhein, 1992) both in vitro and in vivo. Earlier, two
other strong inhibitors of PAL had been identified: (Æ)-2-
aminooxy-3-phenylpropanoic acid (Amrhein and God-
eke, 1977) and (R)-1-amino-2-phenylethylphosphonic
acid (Janas et al., 1985; Laber et al., 1986). Of these
three compounds, AIP appears to have the highest
potential as a specific inhibitor in vivo and has been used
as such in numerous studies, the most recent examples
being those of Ruuhola and Julkunen-Tiitto (2000),
Hrubcova et al. (2000), Schaller et al. (2000), Chen and
McClure (2000), Chen and Chen (2000) and Kostenyuk
et al. (2001). Thus, specific PAL inhibitors are valuable
tools in the study of the phenylpropanoid pathway and
its functions in vivo. Acinnamic acid derivative has
Additionally, derivatives of PAL inhibitors carrying,
for example, a photoreactive or luminophoric group,
may assist in the exploration of the enzyme’s active site
(Kotzyba-Hilbert et al., 1995).
In continuing our studies on phenylalanine ammonia-
lyase (Zon and Laber, 1988; Zon and Amrhein, 1992;
Zon, 1996; Lewandowicz et al., 1999; Gloge et al., 2000),
we report here the synthesis of a set of various analogues
of phenylglycine (1–41, Fig. 1), mostly ring-substituted
1-aminobenzylphosphonic acids, and their evaluation as
potential inhibitors of PAL both in vitro and in vivo.
2. Results and discussion
2.1. Synthesis
Our main effort was directed towards the synthesis of
a wide spectrum of phosphonic and phosphinic analo-
gues of a-phenylglycine and testing them as potential
inhibitors of phenylalanine ammonia-lyase. To achieve
this, we relied on various reports in the literature on the
synthesis of ring-substituted 1-aminobenzylphosphonic
and -phosphinic acids.
We have obtained a set of 1-aminobenzylphosphonic
acids (1, 3–25) using three synthetic routes: (i) direct
* Corresponding author. Fax: +48-71-328-40-64.
E-mail address: zon@kch.ch.pwr.wroc.pl (J. Zon).
0031-9422/02/$ - see front matter # 2002 Published by Elsevier Science Ltd.
PII: S0031-9422(01)00425-3

´
J. Zon et al. / Phytochemistry 59 (2002) 9–21
10
Fig. 1. Structures of the compounds tested 1–41.
phosphonylo-amidation of aldehydes (Oleksyszyn,
1987; Soroka, 1990), (ii) phosphonylation of substituted
benzylidenebisacetamides (Soroka et al., 1990), and (iii)
hydrophosphonylation of substituted benzylidene-
benzhydrylamines (Green et al., 1994), starting from
ring-substituted benzaldehydes. The three methods (A–C
as described in the Experimental) differed in their effec-
tiveness, depending on the structure of the respective sub-
stituted benzaldehydes (Table 1 and Fig. 1). However, the
most convenient route for the synthesis of substituted 1-
aminobenzylphosphonic acids was found to be that pro-
ceeding via the benzylidenebenzhydrylamines (method
C). So far, we have found only a single substituted ben-
zaldehyde that did not provide the corresponding ami-
nophosphonic acid via this method, i.e. anthracene-9-
carboxaldehyde (not described in this paper).
Racemic 1-amino-3-phenylpropylphosphonic acid (2)
was synthesised by reduction of the corresponding oxime
(method D). Racemic 1-amino-40-methylbenzylpho-
sphonous acid (26) was obtained as described previously
from p-toluidenebisacetamide and hypophosphorous acid
(method D). Racemic 1-amino-40-methylbenzyl(methyl)-
phosphinic acid (27) was obtained from p-toluidenebis-
acetamide and methyldichlorophosphine in acetic acid
according to the published procedure (method D).
Racemic ring-substituted phenylglycines (29 and 32)
were obtained by the Strecker synthesis (method E) and
were used as reference compounds.
Optically active 1-aminobenzylphosphonic acids (35–
40) and (+)-1-amino-3-phenylpropylphosphonic acid
(41) were obtained by resolution of the diastereomeric
amides of the corresponding diphenyl (Æ)-1-aminoalk-
ylphosphonates and (À)-O,O0-dibenzoyl-l-tartaric acid
according to known procedure (method F) and their
enantiomeric purity (595% ee) was estimated by
31P{1H} NMR of their corresponding diastereomeric
amides (Kafarski et al., 1983).
The yields (Table 1) of the above mentioned amino
acids (1–27, 29, 32 and 35–41) are given for analytically
pure samples, and such samples were employed in the
biological tests described below.
2.2. Inhibitory activity
The compounds 1–41 (Fig. 1) were evaluated as inhi-
bitors of phenylalanine ammonia-lyase both in vitro

´
J. Zon et al. / Phytochemistry 59 (2002) 9–21
11
Table 1
Analytical data of the compounds 1–41
Compound
Method of
synthesis
Yielda
(%)
M.p.
(ꢀC)
IR
ꢀ
NMR (D2O–DClb or D2O–D2SO4c) ꢁ,
(ppm)
max(KBr) (cmÀ1
)
1
C
26
274–277d
3132, 1621,
1539, 1270,
1213, 1080, 915
1H (300 MHz)b: 4.77 (1 H, d, JHP=17.0 Hz,
NCHP) and 7.52 (5 H, bs, C6H5); 31P{1H}
(121 MHz)b: 13.86 (s)
2
D
24
285–287d
3027, 1603,
1533, 1249,
1179, 1033,
1017, 929
1H (100 MHz)b: 2.40–3.00 (2 H, m,
C6CCH2CN), 3.25–3.45 (2 H, m,
NCCCH2C6), 4.00–4.35 (1 H, m, NCHP), 7.8
(5 H, bs, C6H5)
d
d
d
3
4
5
6
A25
A32
A25
B
264–267
286–289
254–256
10
3130, 1201,
1073, 909
1H (100 MHz)b: 2.60 (3 H, s, C6CH3), 5.25
(1 H, d, JHP=17.0 Hz, NCHP) and 7.86 (4 H,
bs, C6H4)
3124, 1256,
1084, 912
1H (100 MHz)c: 2.77 (3 H, s, C6CH3), 5.20 (1
H, d, JHP=18.0 Hz, NCHP), 7.66 (4 H,
bs, C6H4)
3139, 1205,
1081, 914
1H (300 MHz)b: 2.52 (3 H, s, C6CH3), 5.03
(1 H, d, JHP=17.0 Hz, NCHP), 7.86 (4 H, bs,
C6H4)
277–280e
273–277e
3150, 1210,
1170, 1135,
1070, 910
1H (300 MHz)c: 4.79 (1 H, d, JHP=17.2 Hz,
NCHP), 7.68 (2 H, d, JHH=7.9 Hz, C6H2),
7.84 (2 H, d, JHH=8.1 Hz, C6H2); 31P {1H}
(121 MHz)c: 12.31 (s)
7
B
11
3140, 1620,
1533, 1250, 1085
1H (300 MHz)c: 1.19 (3 H, t, JHH=6.5 Hz,
CH3C), 2.68 (2 H, q, JHH=7.5 Hz, CCH2),
4.68 (1 H, d, JHP=18.0 Hz, NCHP), 7.35–
7.45 (4 H, m., C6H4)
8
9
B
B
12
9
247–249d
271–274d
3226, 1260,
1198, 1085, 905
1H (80 MHz)c: 4.9 (1H, d, JHP=17.0 Hz,
NCHP), 7.1–7.7 (4H, m, C6H4)
3421, 1272,
1195, 1087, 914
1H (300 MHz)c: 4.67 (1H, d, JHP=17.0 Hz,
NCHP), 7.35–7.57 (4H, m, C6H4); 31P{1H}
(121 MHz)c: 12.67 (s)
10
11
C
B
27
30
265–268d
246–248d
3660, 3320,
1258, 1177,
1090, 920
1H (300 MHz)c: 4.87 (1H, d, JHP=17.5 Hz,
NCHP), 7.38–7.60 (4H, m, C6H4); 31P{1H}
(121 MHz)c: 11.80 (s)
3606, 3310,
1248, 1172,
1035, 933
1H (100 MHz)c: 5.72 (1H, d, JHP=17 Hz,
NCHP), 7.60–8.21 (4H, m., C6H4)
12
13
14
B
20
274–276e
3113, 1268,
1180, 1071, 900
1H (100 MHz)c: 5.30 (1H, d, JHP=18.0 Hz,
NCHP), 7.70–8.28 (4H, m., C6H4)
d
A28
B
290–296
23
3660, 3340,
1227, 1085, 917
1H (60 MHz)c: 5.32 (1 H, d, JHP=17.0 Hz,
NCHP), 7.67–8.33 (4 H, m., C6H4)
257–260d
3130, 1273,
1200, 1076, 925
1H (300 MHz)c: 4.99 (1H, d, JHP=17.6 Hz,
NCHP), 7.23–7.39 (2H, m, C6H4), 7.48–7.62
(2H, m, C6H4); 31P{1H} (121 MHz)c: 12.33
(d, JFP=3.8 Hz)
15
16
B
B
21
22
275–278d
273–275d
3113, 1260,
1200, 1150,
1085, 990
1H (300 MHz)c: 4.83 (1H, d, JHP=17.4 Hz,
NCHP), 7.20–7.40 (3H, m, C6H4), 7.45–7.60
(1H, m, C6H4); 31P{1H} (121 MHz)c: 12.00 (s)
3590, 3440,
1270, 1185,
1105, 1093,
1035, 980
1H (300 MHz)c: 4.84 (1H, d, JHP=17.3 Hz,
NCHP), 7.20–7.32 (3H, m, C6H4), 7.48–7.57
(1H, m, C6H4); 31P{1H} (121 MHz)c: 12.49 (s)
(continued on next page)

´
J. Zon et al. / Phytochemistry 59 (2002) 9–21
12
Table 1 (continued)
Compound
Method of
synthesis
Yielda
(%)
M.p.
(ꢀC)
IR
ꢀ
NMR (D2O–DClb or D2O–D2SO4c) ꢁ,
(ppm)
max(KBr) (cmÀ1
)
17
A10
>350d
3130, 1245,
1170,1030, 953
1H (60 MHz)c: 5.23 (1 H, d, JHP=17.0 Hz,
NCHP), 7.40 (2 H, d, JHH=8.0 Hz, C6H2),
7.83 (2 H, d, JHH=8.0 Hz, C6H2)
d
18
A45
242–245
3570, 3380,
1255, 1180, 972
1H (60 MHz)c: 4.18 (3 H, s, OCH3), 5.20 (1
H, d, JHP=17.0 Hz, NCHP), 7.41 (2 H, d,
JHH=8.0 Hz, C6H2), 7.85 (2 H, d, JHH=8.0
Hz, C6H2)
19
20
21
22
B
B
B
C
31
16
15
34
256–259e
286–289e
275–278e
276–279e
3429, 3169,
1257, 1202,
1078, 933
1H (100 MHz)c: 2.96 (3H, s, CH3), 5.37 (1H,
d, JHP=18.0 Hz, NCHP), 7.82–8.22 (2H, m,
C6H3), 8.63 (1H, s, C6H3)
3420, 3126,
1278, 1204,
1084, 921
1H (300 MHz)c 4.68 (1H, d, JHP=16.9 Hz,
NCHP), 7.25–7.50 (3H, m, C6H3) ); 31P{1H}
(121 MHz)c: 12.48 (d, JFP=3.9 Hz)
3423, 1241,
1210, 1056, 942
1H (300 MHz)c: 4.87 (1H, d, JHP=17.7 Hz,
NCHP), 7.38–747 (1H, m, C6H3), 7.62–7.72
(2H, m, C6H3); 31P{1H} (121 MHz)c: 11.06 (s)
3140, 1623,
1538, 1272,
1203, 1084, 910
1H (300 MHz)c: 2.26 and 2.28 (6H, two
overlapping singlets, 30,40-(CH3)2C6H3), 4.89
(1H, d, JHP=17.4 Hz, NCHP), 7.18–7.30
(3H, m, C6H3); 31P{1H} (121 MHz)c: 12.45 (s)
d
d
23
24
25
A35
A13
B
253–255
235–238
9
3460, 3270,
1255, 1177,
1158, 1087, 913
1H (300 MHz)c: 4.92 (1 H, d, JHP=17.0 Hz,
NCHP), 7.50–7.75 (3 H, m, C10H7), 7.85–8.10
(4 H, m, C10H7)
3425, 3240,
1264, 1197,
1183, 1065, 924
1H (300 MHz)c: 5.80 (1 H, d, JHP=17.6 Hz,
NCHP), 7.60–7.70 (4 H, m, C10H7), 8.00–8.15
(3 H, m, C10H7)
244–248e
238–241d
3432, 1222,
1199, 1087, 943
1H (300 MHz)c: 2.68 (3H, s, CH3), 5.80 (1H,
d, JHP=17.5 Hz, NCHP), 7.38–7.83 (4H, m,
C
10H6), 8.10–8.45 (2H, m, C10H6); 31P{1H}
(121 MHz)c: 12.79 (s)
26
27
D
D
20
33
2920, 2305,
1640, 1235,
1195, 1183,
1065, 952
1H (300 MHz)b: 2.23 (3H, s, CH3), 4.54 (1H,
d, JHP=14.0 Hz, NCHP), 7.25
(4H, bs, C6H4); 31P{1H} (121 MHz)b: 22.43
(t, JPD=88.5 Hz) and small singlet for the
nondeuteriated compounds 22.84 (s)
267–269e
2858, 2305,
1634, 1231,
1208, 1184,
1094, 1062, 885
1H (300 MHz)b: 1.74 (3H,d, JHP=14.4 Hz,
PCH3), 2.48 (3H, s, C6CH3), 5.08 (1H, d,
JHP=12.0 Hz, NCHP), 7.14–7.30 (4H, m,
C6H4); 31P{1H} (121 MHz)b: 45.20 (s)
f
28
29
E
9
8
247–250d
2980, 1585,
1520, 1357
1H (300 MHz)b: 1.59 (3 H, s, C6CH3), 4.46
(1 H, s, NCHCO), 6.58 (2 H, d, JHH=8.3 Hz,
C6H4), 6.63 (2 H, d, JHH=8.3 Hz, C6H4).
f
f
30
31
32
E
206–208d
2934, 1675,
1566, 1512, 1373
1H (300 MHz)b: 5.89 (1H, s, NCHCO),
7.33–7.58 (4H, m, C10H7), 7.58–7.78 (2H, m,
C10H7), 7.99 (1H, d, JHH=8.0 Hz, C10H7)
f
f
33
34
(continued on next page)

´
J. Zon et al. / Phytochemistry 59 (2002) 9–21
13
Table 1 (continued)
Compound
Method of
synthesis
Yielda
(%)
M.p.
(ꢀC)
IR
ꢀ
NMR (D2O–DClb or D2O–D2SO4c) ꢁ,
(ppm)
max(KBr) (cmÀ1
)
35g
36g
37g
38g
39g
40g
41g
F
F
F
F
F
F
F
5
280–285c
275–278
265–267
280–283d
272–276
255–257
273–277
3132, 1621,
1539, 1270,
1213, 1080, 915
1H (300 MHz)b: 4.70 (1 H, d, JHP=17.0 Hz,
NCHP) and 7.50 (5 H, bs, C6H5); 31P{1H}
(121 MHz)b: 13.76 (s)
10
5
3139, 1205,
1081, 914
1H (300 MHz)b: 2.52 (3 H, s, C6CH3), 5.03
(1 H, d, JHP=17.0 Hz, NCHP), 7.86
(4 H, bs, C6H4)
3423, 1241,
1210, 1056, 942
1H (300 MHz)c: 4.87 (1H, d, JHP=17.7 Hz,
NCHP), 7.38–747 (1H, m, C6H3), 7.62–7.72
(2H, m, C6H3); 31P{1H} (121 MHz)c: 11.06 (s)
10
12
4
3132, 1621,
1539, 1270,
1213, 1080, 915
1H (300 MHz)b: 4.70 (1 H, d, JHP=17.0 Hz,
NCHP) and 7.50 (5 H, bs, C6H5); 31P{1H}
(121 MHz)b: 13.76 (s)
3139, 1205,
1081, 914
1H (300 MHz)b: 2.52 (3 H, s, C6CH3), 5.03
(1 H, d, JHP=17.0 Hz, NCHP), 7.86
(4 H, bs, C6H4)
3423, 1241,
1210, 1056, 942
1H (300 MHz)c: 4.87 (1H, d, JHP=17.7 Hz,
NCHP), 7.38–747 (1H, m, C6H3), 7.62–7.72
(2H, m, C6H3); 31P{1H} (121 MHz)c: 11.06 (s)
9
3027, 1603,
1533, 1249,
1179, 1033,
1017, 929
1H (100 MHz)b: 2.40–3.00 (2 H, m, C6CCH2CN),
3.25–3.45
(2 H, m, NCCCH2C6), 4.00–4.35 (1 H, m, NCHP),
7.8 (5 H, bs, C6H5)
a
Yield was calculated for an analytically pure sample; for methods A, B, C, and E based on the aldehyde, for method D based on the carboxylic
acid or the aldehyde, for method F based on the diphenyl (Æ)-1-aminoalkylphosphonate.
b
NMR solvents.
NMR solvents.
c
Literature melting points: 1, 285–286 ꢀC (Green et al., 1994); 2, 286–287 ꢀC (Kafarski et al., 1995); 3, 228–230 ꢀC (Kafarski et al., 1995); 4, 286–
d
289 ꢀC (Gancarz and Wieczorek, 1977); 5, 254–256 ꢀC (—ukszo and Tyka, 1977); 8, 229–231 ꢀC (Kafarski et al., 1995); 9, 235–236 ꢀC (Kafarski et al.,
1995); 10, 279–281 ꢀC (Oleksyszyn and Gruszecka, 1981); 11, 222–223 ꢀC (Kafarski et al., 1995); 13, 276–279 ꢀC (Green et al., 1994); 14, 256–258 ꢀC
(Kafarski et al., 1995); 15, 279–280 ꢀC (Kafarski et al., 1995); 16, >300 ꢀC (Maier and Diel, 1991); 17, decomposition (Soroka et al., 1990); 18, 286–
288 ꢀC (Green et al., 1994); 23, 298–302 ꢀC (—ukszo and Tyka, 1977); 24, 249–251 ꢀC (Soroka et al., 1990); 26, 237–238 ꢀC (Tyka and Hagele, 1989);
29, 260–265 ꢀC (Doyle et al., 1962); 32, 199–201 ꢀC (Bretscheider et al., 1988); 35, 280–282 ꢀC (Kafarski et al., 1983); 38, 278–279 ꢀC (Kafarski et al.,
1983).
e
Elemental analysis for new compounds: 6, found N 5.40, P 12.00, C8H9F3NO3P requires N 5.49, P 12.14; 7, found N 6.43, P 14.25, C9H14NO3P
requires N 6.51, P 14.39; 12, found Br 29.95, N 5.22, P 11.78, C7H9BrNO3P requires Br 30.03, N 5.27, P 11.64; 19, found N 11.23, P 12.97,
C8H11N2O5P requires N 11.38, P 12.58; 20, found N 6.30, P 13.68, C7H8F2NO3P requires N 6.28, P 13.88; 21, found Cl 27.44, N 5.25, P 12.40,
C7H8Cl2NO3P requires Cl 27.70, N 5.47, P 12.10; 22, found N 6.80, P 14.60, C9H14NO3P requires N 6.51, P 14.39; 25, found N 5.40, P 12.20,
C12H14NO3P requires N 5.58, P 12.33; 27, found N 6.83, P 15.70, C9H14NO2P requires N 7.03, P 15.55; 36, found N 6.90, P 15.31, C8H12NO3P
requires N 6.96, P 15.40; 37, found Cl 27.44, N 5.25, P 12.40, C7H8Cl2NO3P requires Cl 27.70, N 5.47, P 12.10; 39, found N 6.85, P 15.27,
C8H12NO3P requires N 6.96, P 15.40; 40, found Cl 27.51, N 5.28, P 12.31, C7H8Cl2NO3P requires Cl 27.70, N 5.47, P 12.10; 41, found N 6.43, P
14.30, C9H14NO3P requires N 6.51, P 14.39.
f
Commercial products.
Determined specific rotation: 35, ½ꢂŠ578+18.0Æ1 (c 1, 1 M NaOH) (literature ½ꢂŠ578+19Æ1 (c 2, 1 M NaOH); Kafarski et al., 1983); 38,
25
20
g
ꢀ
ꢀ
25
20
½ꢂŠ578À21.5Æ1 ꢀ (c 1, 1 M NaOH) (literature ½ꢂŠ578À20Æ1 ꢀ (c 2, 2 M NaOH); Kafarski et al., 1983); for new compounds 36, 37, 39, 40, and 41 see
Experimental.
and in vivo. The inhibition constants (Ki) of buckwheat
PAL, and the respective concentrations of the com-
pounds which inhibited the anthocyanin production in
illuminated buckwheat hypocotyls by 50% (I50 values),
were determined as described previously (Zon and
Amrhein, 1992).
Compounds 1 and 2 are homologues with shorter and
longer side chains, respectively, of 1-amino-2-pheny-
lethylphosphonic acid, a known inhibitor of PAL (Janas
et al., 1985; Laber et al., 1986). Even though compound 2
was a somewhat more potent inhibitor than 1, we pursued
the synthesis of the series of compounds (3–18) related to
1-aminobenzylphosphonic acid (1), because of their lower
conformational flexibility. Racemic monosubstituted 1-
aminobenzylphosphonic acids (3–18) containing the
following groups: chloro, bromo, fluoro, methyl, tri-
fluoromethyl, ethyl, hydroxy, and methoxy in ortho,
meta and para position, respectively, were synthesised

´
J. Zon et al. / Phytochemistry 59 (2002) 9–21
14
(Fig. 1). Compounds 10 and 5 (chloro or methyl in para
position) exhibited the highest inhibitory activities in
this series of racemic monosubstituted 1-aminobenzyl-
phosphonic acids (Table 2). For compound 10 it was
established that, at 100 mM concentration, it causes a
21-fold increase in the endogenous concentration of
soluble phenylalanine over a period of 24 h in illumi-
nated buckwheat hypocoptyls (data not shown). This
provides additional evidence that PAL is, indeed, an in
vivo target of the inhibitor, as was previously shown for
AIP (Zon and Amrhein, 1992). Introducing the chloro
or methyl group at the meta position (compounds 9 and
4, respectively) affected the inhibitory activity only
slightly (Table 2). In contrast, introduction of the same
substituents at the ortho position significantly reduced
the inhibitory activity (compounds 8 and 3, Table 2).
Racemic 1-amino-30,40-dichlorobenzylphosphonic acid
(21) was identified as the most potent inhibitor both in
vitro and in vivo in this subclass consisting of racemic
mono- and disubstituted 1-aminobenzylphosphonic
acids (Table 2). We obtained and evaluated separately
dextrorotatory (37) and laevorotatory (40) 1-amino-
30,40-dichlorobenzylphosphonic acid. The enantiomer 37
was a much more potent inhibitor of PAL in vitro and
in vivo than its laevorotatory counterpart (Table 2). We
propose the R configuration (corresponding to the nat-
ural L-configuration) for 37, based on a stronger inhi-
bitory activity of (R)-(+)-1-aminobenzylphosphonic acid
(35) as compared to the enantiomer 38 (G•owiak et al.,
1977). With the same argument, we propose R-configura-
tion for dextrorotatory 1-amino-40-methylbenzylpho-
sphonic acid (36). Consequently, the S-configuration is
proposed for the compounds 39 and 40.
Incidentally, we did not observe a significant difference
in the inhibitory activities of dextroratatory (41) and
racemic (2) 1-amino-3-phenylpropylphosphonic acids,
respectively (Table 2). The latter compound 2 was
reported to inhibit potato phenylalanine ammonia-lyase
and anthocyanin synthesis in buckwheat hypocotyls
(Janas and Olechowicz, 1994).
Acomprehensive study of the inhibitory activities of
substituted analogues of 1-amino-2-phenylethylpho-
sphonic acid (the phosphonic acid analogue of phenylala-
nine), with a small substituent like fluorine in the benzene
ring at any position, revealed that such a substituent low-
ers the inhibitory activity only insignificantly. This is not
surprising considering the atomic radius of fluorine.
Any other substituents at para position, however, sub-
stantially diminish the inhibitory activity (Maier and
Diel, 1994).
Table 2
Evaluation of compounds 1–41 as inhibitors of phenylalanine ammo-
nia-lyase and anthocyanin synthesis
Compound
Ki PAL
(mM)
I50 anthocyanin
synthesis (mM)
1
6.5
5.0
8.5
1.5
0.25
8.3
34
250
110
4860
23
2
3
4
5
30
6
240
>1000
<1000
10
7
8
a
9
2.0
0.21
5.3
1.0
0.29
a
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
10
1710
15
22
b
67
3.3
1.8
a
25
b
40
>1000
82
16
4.0
1.0
0.21
0.89
3.7
0.30
7.5
a
It is worth mentioning that compounds 23–25 with
the aminomethylphosphonic moiety adjacent to the
naphthalene rather than the benzene scaffold had lower
inhibitory activities (Table 2). This was particularly evi-
dent for the in vivo inhibition (I50 values), possibly
reflecting poor uptake of these compounds. Neither
phenylglycine nor phosphinic analogues of phenylgly-
cine were suitable compounds in the search for PAL
inhibitors (Table 2: 28, 29 and 31, as well as 26 and 27).
5
24
220
140
b
b
25
550
<5000
640
b
9.0
0.24
a
1.3
1.0
2.3
300
3.4
0.17
0.08
a
1500
3000
3300
>5000
53
2.3. Structure–activity relationship studies
15
In this study, two different biological activity factors
have been analysed, i.e. the inhibition constant (Ki) of
buckwheat PAL (‘‘in vitro’’ activity), and the I50 value
for the inhibition of anthocyanin synthesis (‘‘in vivo’’
activity). In the latter assay, anthocyanins in illuminated
buckwheat hypocotyls represent one of the ultimate
products of phenylpropanoid metabolism, and in vivo
inhibition of PAL will, therefore, lead to a reduction of
2.2
b
a
b
3.1
10
965
60
a=Less than 50% inhibition at 1 mM concentration. b=Less than 40%
inhibition of light-induced anthocyanin synthesis in buckwheat hypocotyls
at 1 mM concentration.

´
J. Zon et al. / Phytochemistry 59 (2002) 9–21
15
anthocyanin accumulation. In this test, other factors,
such as differential uptake and metabolism of an inhi-
bitor, will also contribute to its biological activity. Fur-
thermore, targets other than PAL may be affected, and
one cannot a priori expect a correlation between the in
vitro and in vivo activities, respectively, of a set of inhi-
bitors. In fact, the plot shown in Fig. 2 reveals, that Ki
and I50 values are not strongly correlated for all data
points. There are, however, clusters of points, which can
be interpreted as follows. The aminocarboxylic acids
(filled squares), the aminophosphonic acids with sub-
stituents longer than one carbon atom at para position
(open squares), the aminophosphonic acids with sub-
stituents in ortho and meta position other than a fused
benzene ring (asterisks), are very distinct from each
other and also from the rest of the aminophosphonic
acids (filled circles). Analysis of the outliners from the
best correlation line allows to formulate the following
conclusions: (i) the main factor influencing the activity
is just the presence of a aminomethylphosphonic group
attached to benzene ring, (ii) the biological activity
drops when a substituent longer than one carbon atom
is attached at position para in benzene ring, which can
suggest that the PAL active site is not very deep, (iii) the
activity drops when substituent at position ortho is
other than fused benzene ring, which might indicate that
in this region the pocket is relatively flat.
We have also analysed the correlation between each
of the biological responses and electronic, steric and
lipophilic parameters of the substituents (Table 3). The
correlation is very small, and correlation factors did not
exceed a value of 0.27, meaning that there was no cor-
relation at all.
3. Conclusions
We have found that (+)-1-amino-30,40-dichlor-
obenzylphosphonic acid (37) is quite a strong inhibitor
of phenylalanine ammonia-lyase both in vitro and in
vivo. Its absolute configuration is proposed to be R.
This conformationally more flexible compound 37 and
the conformationally locked 2-aminoindan-2-phos-
phonic acid (Ki=0.08 mM; I50=1.5 mM) (Zon and
Amrhein, 1992) are of comparable potency.
Presently, little information is available on the struc-
ture of the active site of PAL, but both site-directed
mutagenesis (Schuster and Retey, 1994; Langer et al.,
1997) and comparison with the recently determined
three-dimensional structure of histidine ammonia-lyase
(Schwede et al., 1999) will advance our understanding of
this important enzyme of higher plants and provide the
basis for an understanding of the action of the PAL
inhibitor.
Fig. 2. Correlation analysis of potency of compounds to inhibit PAL in vitro (Ki) and anthocyanin synthesis in vivo (I50).
Table 3
Correlation coefficientsa between structural parameters and biological activity of evaluated PAL inhibitors
Biological test
Electronic parameters (ꢃ)
Steric parameters (Es)
Lipophilic parameters (ꢄ)
Antocyanin synthesis reduce (I50
Inhibition constant (Ki)
)
À0.0215
0.2684
0.2007
0.1494
0.0981
0.0624
a
Correlation coefficients given in the Table 3 were calculated using statistical package STATGRAPH 6.0 and are significant at the confidence
level P=0.05.

´
J. Zon et al. / Phytochemistry 59 (2002) 9–21
16
ꢀ
On the other hand, our inhibitor studies may allow
dure. Yield: 21.92 g; 56%; bp 132–134 C (2 mm Hg).
GC analysis shows 94% of 1-methyl-4-naphthaldehyde
(retention time 7.9 min) and 6% of 1-methylnaphtha-
lene (retention time 1.1 min on column filled with 2%
the design of inhibitors containing a photoreactive
group, and such compounds may then serve as probes
into the active site of PAL.
ꢀ
1
DEGS on Chromosorb G, 1 m, 180 C). H NMR (60
MHz, CCl4, Me3SiOSiMe3 external): ꢁ 2.67 (3 H, s,
CH3), 7.20–8.08 (5 H, m, C10H6), 9.23 (1 H, d, J=9 Hz,
C10H6), 10.13 (1H, s, CHO). The aldehyde was used to
obtain compound 25 (method B).
4. Experimental
1H NMR spectra were recorded at 100 MHz with a
Tesla BS 497, at 80 MHz with a Tesla BS 587A, and at
300 MHz with a Bruker DRX 300 instrument. 31P{1H}
NMR spectra with broad band proton decoupling were
recorded at 121 MHz with a Bruker DRX 300 instrument.
IR spectra were recorded with a Perkin Elmer 2000 FT–
IR spectrometer. Optical rotations were determined on an
Optical Activity LTD Automatic polarimeter using 5 cm
cell. GC analyses were performed on a Perkin Elmer F-11
chromatograph. GC–MS analyses were performed on a
Hewlett Packard 5890 series II chromatograph with mass
selective detector 5971A. Melting points were determi-
nated using Boetius apparatus and are uncorrected. Ele-
mental analyses were performed by Laboratory of
Elemental Analysis of Institute Organic Chemistry, Bio-
chemistry and Biotechnology, Wroc•aw. TLC was carried
out on commercially available pre-coated plates (Merck
Kieselgel 60). THF was freshly distilled from lithium alu-
minium hydride. The following commercially available
amino acids were tested: (Æ)-phenylglycine (28) (Sigma),
(Æ)-2-amino-20-fluorophenylacetic acid (30) (Fluka),
(Æ)-2-amino-40-fluorophenylacetic acid (31) (Fluka),
(S)-(+)-phenylglycine (33) (Fluka) and (R)-(À)-phe-
nylglycine (34) (Sigma). All other amino acids were
synthesised by one of the method described below
(method A, B, C, D, E, and F). After synthesis, the
respective aminophosphonic acid was recrystallised
from water in the following manner. It was dissolved in
a minimum volume of boiling water, the solution trea-
ted with charcoal, filtered and filtrate was concentrated
under atmospheric pressure to about one thirds of
starting volume. The crystals of aminophosphonic acid
was filtered and dried on air. The yield, melting point,
3,4-Dimethylbenzaldehyde was obtained from com-
mercially available 3,4-dimethylbenzoic acid in a three-
step synthesis. 3,4-Dimethylbenzoic acid was esterified
by methanol-dry hydrochloride yielding corresponding
methyl 3,4-dimethylbenzoate. To the suspension of
lithium aluminium hydride (2.00 g, 0.0527 mol) in dry
THF (45 ml) a solution of methyl 3,4-dimethylbenzoate
(6.86 g, 0.0418 mol) in dry THF (15 ml) was added
dropwise under stirring and under nitrogen. The mix-
ture was refluxed for 3 h. Then, second portion of
lithium aluminium hydride (0.543 g, 0.01431 mol) was
added and the mixture was refluxed for an additional 6
h. After that time water (1.8 ml) was carefully added
under stirring to the cold reaction mixture (0–5 ꢀC), and
then 10% sulfuric acid (19.5 ml) was added dropwise.
After filtration, the solid residue was extracted with
methylene chloride (50 ml). The organic extract and the
filtrate were combined, dried with anhydrous potassium
carbonate and the solvents were evaporated under
reduced pressure. The residue (5.319 g) was dissolved in
boiling hexane (20 ml) and let to stand to crystallise,
yielding 3,4-dimethylbenzyl alcohol: 4.515 g; 79%; mp
ꢀ
58–61 C. Elemental analysis: found C 79.40, H 8.90,
C9H12O requires C 79.37, H 8.88. In the subsequent step
3,4-dimethylbenzyl alcohol (1.000 g, 0.00734 mol) was
stirred at room temperature for 90 h in dry methylene
chloride (30 ml) with manganese (IV) oxide (5.00 g).
The reaction was monitored by TLC: hexane-diethyl
ether (4:1 v/v), iodine, Rf for the aldehyde 0.35 and for
the alcohol 0.09. After filtration out manganese (IV)
oxide the methylene chloride was evaporated providing
crude 3,4-dimethylbenzaldehyde (0.8774 g, 89%) suffi-
1
1
elemental analysis, IR selected band, H NMR spectra
ciently pure for the next reaction. H NMR (300 MHz,
and inhibitory activity were determined (Table 1). The
inhibition constants (Ki) of buckwheat PAL, and the
respective concentrations of the compounds which
inhibited the anthocyanin production in illuminated
buckwheat hypocotyls by 50% (I50 values), were deter-
mined as described previously (Zon and Amrhein,
1992). Standard deviations for Ki values were in the
range of 5–8%, for I50 values in the range of 10–15%.
Assays for compounds with I50 >1000 mM were per-
formed only once.
CDCl3): ꢁ 2.28 and 2.25 (6H, two overlapping singlets,
CH3 and CH3), 7.18–7.25 (1H, m, C6H3), 7.51–7.58 (2H,
m, C6H3), 9.87 (1H, s, CHO). IR ꢀfmilamx cmÀ1: 1694
(C¼O). 3,4-Dimethylbenzaldehyde spontaneously oxi-
dised to the corresponding acid even in a freezer. So
fresh 3,4-dimethylbenzaldehyde (0.7603 g, 0.00566 mol)
and benzhydrylamine (1.0386 g, 0.00566 mol) were dis-
solved in methylene chloride (20 ml) and left for 2 h.
Then the solution was dried with anhydrous potassium
carbonate (0.6 g), and was evaporated to get solid 3,4-
(dimethylbenzylidene)bezhydrylamine: 1.4039 g; 83%;
1-Methyl-4-naphthaldehyde was obtained from com-
mercially available 1-methylnaphthalene and a,a-
dichloromethyl methyl ether in presence of titanium
(IV) chloride according to Rieche et al. (1960) proce-
ꢀ
mp 72–73 C. The last compound was used to obtain
(Æ)-1-amino-30,40-dimethylbenzylphosphonic acid (22)
(method C).

´
J. Zon et al. / Phytochemistry 59 (2002) 9–21
17
4.1. Synthesis of racemic 1-aminobenzylphosphonic acids
from the corresponding substituted benzaldehyde,
acetamide, acetyl chloride and phosphorous(III) chloride
according to Oleksyszyn (1987) and Soroka (1990)
procedures
(25 ml) was left for 3 h at room temperature, then was
dried with anhydrous potassium carbonate (5 g) and a
residue was treated with hexane. Aresulted crude (ben-
zylidene)diphenylmethylamine (0.02 mol) and diethyl
phosphite (2.58 ml, 0.02 mol) were heated for 4 h at
ꢀ
100Æ3 C to obtain a crude diethyl 1-diphenylmethyla-
Asolution of substituted benzaldehyde (0.02 mol) and
acetamide (2.36 g, 0.04 mol) in glacial acetic acid (10 ml)
was cooled on a water-ice bath. At a temperature below
15 ꢀC acetyl chloride (1.42 ml, 0.02 mol) was added
dropwise while stirring. The reaction mixture was stirred
for 3 h on the bath, then phosphorus (III) chloride (1.74
ml, 0.02 mol) was added dropwise while keeping the
minobenzylphosphonate. The crude product was treated
with hexane. Asolid diethyl 1-diphenylmethylamino-
benzylphosphonate was filtered, and dried in open air.
The solid diethyl 1-diphenylmethylaminobenzylpho-
sphonate, were refluxed for 4 h with concentrated
hydrochloric acid (20 ml) and water (20 ml). The
hydrolysate was extracted with methylene chloride
(2ꢁ15 ml), and the aqueous fraction was concentrated
under reduced pressure. The solution was evaporated
under reduced pressure, and residue was treated with
methanol (20 ml), evaporated under reduced pressure
and again dissolved in methanol (20 ml). The solution
was treated with propylene oxide (1.40 ml, 0.02 mol)
yielding a crude crystalline product. The last one was
recrystallised one or more time from water until analy-
tically pure aminophosphonic acids was obtained
(Table 1).
ꢀ
temperature below 15 C. Then mixture was gradually
heated to 60 ꢀC during 1 h and then to 90 ꢀC during the
next 2 h. Volatile by-products were evaporated under
reduced pressure. The residue was refluxed by 10 h with
concentrated hydrochloric acid (20 ml) and water (20 ml).
Tarry semisolid or oily impurities were separated by
decantation or extraction with chloroform. An acidic
solution was decolourised by a charcoal. Then evaporated
under reduced pressure, and the residue was treated with
methanol (10 ml). The insoluble ammonia chloride was
filtered off and the filtrate was treated with propylene
oxide (1.40 ml, 0.02 mol) which yielded a crude crystalline
product. The last one was recrystallised one or more time
from water until analytically pure aminophosphonic acids
was obtained (Table 1).
4.4. Synthesis of racemic 1-aminoalkylphosphonic acids
by miscellaneous procedures
(Æ)-1-Amino-3-phenylpropylphosphonic acid (2)
This was obtained from 3-phenylpropanoic acid in
four steps. Thus, 3-phenylpropanoic acid was converted
to 3-phenylpropanoyl chloride by phosphorous (III)
chloride using standard method. Then, diisopropyl 1-
hydroxyimino-3-phenylpropylphosphonate was obtained
according to Zon (1984) procedure starting from 3-
phenylpropanoyl chloride (0.02 mol) yielding corre-
4.2. Synthesis of racemic 1-aminobenzylphosphonic acids
from N,N0-benzylidenebisacetamides and
phosphorous(III) chloride in acetic acid followed by
acidic hydrolysis as described by Soroka et al. (1990)
Asubstituted benzaldehyde (0.06 mol), acetamide (7.01
g, 0.12 mol), glacial acetic acid (7 ml) and acetic anhydride
(7 ml) were refluxed for 3 h. Then, after cooling the mix-
ture to room temperature, a crude bisacetamide was crys-
tallised. The crystals were filtered off, washed twice with
methanol and dried in open air to yield sufficiently pure
derivatives of N,N0-(benzylidene)bisacetamides for the
next reaction. To a solution of the crude derivatives of
N,N0-(benzylidene)bisacetamide (0.02 mol) in glacial
acetic acid (10 ml), phosphorus (III) chloride (1.74 ml,
0.02 mol) was added dropwise keeping temperature
below 15 ꢀC while stirring. Then the mixture was gradu-
ally heated up to 60 ꢀC during 1 h and finally up to 90 ꢀC
within additional 2 h. Then the residue was worked up
analogously as in the method A.
ꢀ
1
sponding oxime: 3.0 g; 48%; mp 97–99 C (hexane); H
NMR (60 MHz, CDCl3, Me4Si): ꢁ 1.4 (12 H, dd, J=2.5
Hz, J=7 Hz, OCCH3), 2.6–3.1 (4 H, m, CCH2
CH2C(¼NOH)P), 4.83 (2 H, dq, J=14 Hz, J=14 Hz,
OCHC), 7.2 (5 H, bs, C6H5) and 11.5 ppm (1 H, s,
C¼NOH). The oxime (3.0 g) was subjected to reduction
with zinc in formic acid according to Kowalik et al.
(1981) procedure to give the oxalate of diisopropyl (Æ)-
1-amino-ꢀ3-phenylpropylphosphonate: 2.2 g; 60%; mp
160–165 C. The oxalate was hydrolysed by a refluxing
mixture of conc HCl (20 ml) and water (20 ml) for 10 h.
After standard work-up crude 2 (1.3 g) was isolated.
The crude aminophosphonic acid was dissolved in water
(800 ml), treated with charcoal and the solution was
concentrated by evaporation under atmospheric pres-
sure to a volume of about 150 ml. After crystallization
(Æ)-1-amino-3-phenylpropylphosphonic acid (2) was
collected by filtration and dried (Table 1).
4.3. Synthesis of racemic 1-aminobenzylphosphonic acid
by addition of diethyl phosphite to N-benzylidenebenz-
hydrylamine and subsequent acidic hydrolysis according
to Green et al. (1994) procedure
(Æ)-1-Amino-40-methylbenzylphosphonous acid (26)
was obtained from p-tolualdehyde by Tyka and Hagele
procedure (Tyka and Hagele, 1989).
Asolution of benzaldehyde (0.05 mol) and diphe-
nylmethylamine (8.61 ml, 0.05 mol) in methylene chloride

´
J. Zon et al. / Phytochemistry 59 (2002) 9–21
18
(Æ)-1-Amino-40-methylbenzyl(methyl)phosphinic acid
mol), benzyl carbamate (0.10 mol) and triphenyl phos-
phite (0.10 mol) in glacial acetic acid (15 ml), followed
by bromohydrogenolysis and treatment of the resulting
hydrobromide esters with gaseous ammonia in ethereal
suspension or neutralisation with aqueous 2 M sodium
hydroxide and extracting the free ester with chloroform.
Diphenyl (Æ)-1-amino-40-methylbenzylphosphonate
was obtained from p-tolualdehyde: 8.10 g; 21%; mp 74–
79 ꢀC (diethyl ether); 1H NMR (80 MHz, CCl4, Me4Si):
2.09 (2 H, bs, NH2), 2.35 (3 H, s, C6CH3), 4.60 (1 H, d,
J=16 Hz, NCHP), 6.6–7.6 (14 H, m, NC(C6H5)2 and
C6H4).
(27) was obtained as follow. Methyldichlorophosphine
(2.7 ml, 0.03 mol) was added dropwise to glacial acetic
acid (20 ml) at room temperature while stirring. Stirring
was continued for 30 min, then N,N0-(4-methylbenzyli-
dene)bisacetamide (6.60 g, 0.03 mol) was added in a one
portion. Within a few minutes the bisamide had dissolved.
ꢀ
The reaction mixture was heated at 60–70 C for 2 h.
Volatile components including acetic acid were pumped
off with a water respirator keeping the reaction mixture in
hot water bath under to give a light-yellow residue. The
water (30 ml) and concentrated hydrochloric acid (30 ml)
were added to the residue and the mixture was refluxed for
10 h. Tarry semisolid or oily impurities were separated by
decantation or extraction with chloroform. An acidic
solution was decolourised by a charcoal. The solution
was evaporated under reduced pressure, and residue
was treated with methanol (20 ml) and concentrated
hydrochloric acid (0.5 ml). The insoluble ammonia
chloride was filtered off. The filtrate was treated with
propylene oxide (2.1 ml, 0.03 mol) which yielded a crude
crystalline aminophosphinic acid 27 which was isolated
by filtration, 4.50 g (75%). The recrystallisation from
water yielded the analytically pure product (Table 1).
Diphenyl (Æ)-1-amino-30,40 -dichlorobenzylphosph-
onate was obtained from 3,4-dichloro-benzaldehyde:
ꢀ
1
20.4 g; 50%; mp 104–107 C (diethyl ether); H NMR
(300 MHz, CD3Cl): 2.1 (2H, bs, NH2), 4.63 (1H, d,
J=16.7 Hz, NCHP), 7.00–7.70 (13 H, m, NC(C6H5)2
and C6H3Cl2).
Diphenyl (Æ)-1-amino-3-phenylpropylphosphonate
was obtained from hydrocinnamaldehyde: 5.91 g; 23%;
mp 83–85 ꢀC (diethyl ether); 1H NMR (300 MHz,
CD3Cl): 1.59 (2H, bs, NH2), 1.83–2.02 (1H, m, CCHC),
2.18–2.37 (1H, m, CCHC), 2.69–2.83 (1H, m, CCH2C6),
2.89–3.02 (1H, m, CCH2C6), 3.18–3.32 (1H, m, NCHP),
6.92–7.25 (15H, m, P(OC6H5)2 and CC6H5); 31P{1H}
NMR(121 MHz, CD3Cl): 23.45 (s).
4.5. Synthesis of racemic phenylglycines according to
Greenstein and Winitz (1961) procedure
4.7. Resolution of the amides from diphenyl 1-
aminobenzylphosphonates and O,O0-dibenzoyl-l-tartaric
anhydride according to Kafarski et al. (1983) procedure
Potassium cyanide (3.0 g, 0.046 mol) and ammonium
chloride (2.5 g, 0.047 mol) were dissolved in a pressure
flask, in a minimum volume of water (15 ml). Then, p-
tolualdehyde or 1-naphthaldehyde (0.045 mol) and
methanol (20 ml) was added until reaction mixture
became homogenous. The flask was corked up and
shacked from time to time, and left for 8 h at room tem-
perature. Then the flask was opened (caution: hydrogen
cyanide), and the mixture was extracted with ether (20
ml). Ethereal fraction was washed with water (5 ml).
The ether layer was evaporated and the residue treated
with 10% hydrochloric acid (10 ml) and finally evapo-
rated. The rest was hydrolysed with 6 M hydrochloric
acid (45 ml) for 8 h. The aqueous solution was deco-
lourised with a charcoal, and evaporated to dryness.
The residue was dissolved in methanol (15 ml), filtered,
and treated with propylene oxide (2.8 ml, 0.04 mol). A
crude amino acid (29 or 32) was separated by filtration.
Then, the amino acid was recrystallised from acetic
acid–water mixture to get the analytically pure sample.
The racemic mixture of diphenyl 1-aminoalkylpho-
sphonates were treated by (À)-O,O0-dibenzoyl-l-tartaric
acid anhydride yielding corresponding diastereoizomers
according to Kafarski procedure (Kafarski et al., 1983).
The diastereoizomers were additionally recrystalised
from a solvent (for details see below) until no further
improvement of separation were achieved as seen by
their 31P{1H} NMR.
(+)-1-Amino-40-methylbenzylphosphonic acid (36)
was obtained as follows. The one from two diphenyl 1-
[(10,20-dibenzoyloxy-20-carboxyethanocarbonyl)amino]-
400-methyl-benzylphosphonate diastereoisomers having
lower melting point was obtained from the benzene
solution, after separation of the crystalline product and
evaporation of the solvent. The residue (13.3 g) was
crystallized from diethyl ether (30 ml) providing the
amide with lower melting point: 7.65 g; 37%; mp 96–
102 ꢀC. The amide was extracted with boiling tert-
butylmethyl ether (4ꢁ100 ml), filtered and the combined
extracts were concentrated to a final volume of about
130 ml. After crystallization, the amide was filtered giv-
ing crystals of the amide: 6.18 g; 30%; mp 102.5–
4.6. Synthesis of optically active 1-aminobenzyl-
phosphonic acids. Synthesis of racemic diphenyl 1-amino-
alkylphosphonates according to Oleksyszyn et al. (1979)
procedure
20
107.5 ꢀC; ½ꢂŠ578À41.6Æ1 ꢀ (c 1, acetone); 31P{1H} NMR
Racemic diphenyl 1-aminoalkylphosphonates were
obtained by heating the corresponding aldehydes (0.15
(121 MHz, CDCl3): ꢁ 16.04 (s) and 16.17 (s) with a ratio
of >20:1.

´
J. Zon et al. / Phytochemistry 59 (2002) 9–21
19
(+)-1-Amino-40-methylbenzylphosphonic acid (36)
was obtained from the amide (4.47 g, 0.00644 mol) by
hydrolysis in a mixture of acetic acid and 40% hydro-
bromic acid, followed by ion exchange chromatography
purification and recrystallisation from water yielding:
(À)-1-Amino-30,40-dichlorobenzylphosphonic acid (40)
was obtained as follows. The corresponding diphenyl 1-
[(10,20 -dibenzoyloxy-20 -carboxyethanocarbonyl)amino]-
300,400-dichloro-benzylphosphonate diastereoisomer with
lower melting point was obtained from the benzene
solution left after separation of the crystalline product
and evaporation of the solvent. The residue (12.8 g) was
crystallised from diethyl ether (23 ml) providing the
amide with lower melting point: 3.20 g; mp 167–175 ꢀC.
The amide was extracted with boiling tert-butylmethyl
ether (150 ml), filtered and the solution was con-
centrated to a final volume of about 10 ml. After
recrystallisation from tert-butylmethyl ether, the amide
was filtered, dried in open air providing the amide: 1.20
25
0.43 g; 33%; mp 275–278 ꢀC; ½ꢂŠ578+20.0Æ1 ꢀ (c 1, 1 M
NaOH).
(À)-1-Amino-40-methylbenzylphosphonic acid (39)
was obtained as follows. The corresponding diphenyl 1-
[(10,20 -dibenzoyloxy-20 -carboxyethanocarbonyl)amino]-
400-methylbenzylphosphonate diastereoisomer with
higher melting point was isolated as crystalline product
from the benzene (60 ml) solution of the reaction mixture
of diphenyl (Æ)-1-amino-40-methylbenzylphosphonates
(10.41 g, 0.0295 mol) and O,O0-dibenzoyl-l-tartaric
anhydride (Butler and Cretcher, 1933) (10.03 g, 0.02995
25
g, 7%, mp 172–175.5 ꢀC; ½ꢂŠ578À57.0Æ1 ꢀ (c 1, acetone).
(À)-1-Amino-30,40-dichlorobenzylphosphonic acid (40)
was obtained from the amide (1.15 g, 0.00154 mol) by
hydrolysis in a mixture of acetic acid and 40% hydro-
bromic acid, followed by ion exchange chromatography
ꢀ
mol) yielding: 9.62 g, 47%, mp 171–178 C. The amide
was recrystallised from benzene (150ꢀ ml) to get the
amide: 9.20 g, 45%, mp 173.5–177.7 C. The product
was dissolved in boiling diisopropyl ether (320 ml), and
the solution was concentrated to a final volume of about
purification and recrystallisation from water yielding:
25
578
ꢀ
ꢀ
0.10 g, 25%, mp 255.5–257 C, ½ꢂŠ À23.0Æ1 (c 1, 1
180 ml. After crystallisation, crystals were filtered yielding
20
M NaOH).
the amideꢀ: 6.50 g; 32%; mp 173.5–175.5 ꢀC; ½ꢂŠ
(+)-1-Amino-3-phenylpropylphosphonic acid (41)
was obtained as follows. The reaction of diphenyl (Æ)-
1-amino-3-phenylpropylphosphonate (5.12 g, 0.0139
mol) with O,O0-dibenzoyl-l-tartaric anhydride (Butler
and Cretcher, 1933) (4.74 g, 0.0139 mol) in benzene
solution.(28 ml) resulted in a mixture of both amide
diastereoisomers of diphenyl 1-[(10,20-dibenzoyloxy-20-
carboxyethanocarbonyl)amino] - 3 - phenylpropylphos-
578
À47.5Æ1 (c 1, acetone); 31P{1H} NMR (121 MHz,
CDCl3): ꢁ 16.16 (s) and 16.04 (s) with a ratio of >25:1.
(À)-1-Amino-40-methylbenzylphosphonic acid (39)
was obtained from the amide (3.81 g, 0.00549 mol) by
hydrolysis in a mixture of acetic acid and 40% hydro-
bromic acid, followed by ion exchange chromatography
purification and crystallisation from water yielding: 0.42 g,
25
38%, mp 272–276 ꢀC, ½ꢂŠ578À22.0Æ1 ꢀ (c 1, 1 M NaOH).
phonate as crystalline product: 9.70 g, 98%, mp 130–
25
(+)-1-Amino-30,40-dichlorobenzylphosphonic acid (37)
was obtained as follows. The corresponding diphenyl 1-
[(10,20 -dibenzoyloxy-20 -carboxyethanocarbonyl)amino]-
300,400-dichloro-benzylphosphonate diastereoisomer with
higher melting point was isolated as crystalline product
from the benzene (46 ml) solution of the reaction mixture
181 C, ½ꢂŠ578À79.0Æ1 (c 1, acetone). The amide (8.49
g) was treated with boiling benzene (50 and 25 ml) to
obtain the insoluble crystals: 4.53 g, 46%, mp 89–
ꢀ
ꢀ
25
177 ꢀC, ½ꢂŠ578À57.0Æ1 ꢀ (c 1, acetone) and the solution.
The crystalline product was treated again with boiling
diisopropyl ether (500 ml). The insoluble crystals were
mainly one diastereoisomer of diphenyl 1-[(10,20-diben-
zoyloxy - 20 - carboxyethanocarbonyl)amino] - 3 - phenyl-
propylphosphonate which was used in preparation of
(+)-1-amino-3-phenylpropylphosphonic acid: 2.74 g,
of
diphenyl
(Æ)-1-amino-30,40-dichlorobenzylpho-
sphonate (9.50 g, 0.023 mol) and O,O0-dibenzoyl-l-tar-
taric anhydride (Butler and Cretcher, 1933) (8.0 g, 0.023
mol) yielding: 4.60 g, 26%, mp 102–198 ꢀC. The product
was dissolved in boiling diisopropyl ether (1000 ml), and
the solution was concentrated to a final volume of about
400 ml. After crystallization, crystals were filtered pro-
25
578
28%, mp 98–165 ꢀC, ½ꢂŠ À78.6Æ1 (c 1, acetone).
ꢀ
(+)-1-Amino-3-phenylpropylphosphonic acid (41)
was obtained from the amide (2.68 g, 0.00379 mol) by
hydrolysis in a mixture of acetic acid and 40% hydro-
bromic acid, followed by ion exchange chromatography
20
578
ꢀ
viding the amide: 1.91 g, 11%, mp 199–207 C, ½ꢂŠ
ꢀ
À25.5Æ1 (c 1, acetone).
(+)-1-Amino-30,40-dichlorobenzylphosphonic
acid
purification and recrystallisation from water yielding:
25
(37) was obtained from the amide (1.87 g, 0.00251 mol)
by hydrolysis in a mixture of acetic acid and 40% hydro-
bromic acid, followed by ion exchange chromatography
(0.2 M hydrochloric acid was used as eluent). The eluate
was evaporated to dryness, the residue was dissolved in
methanol (5 ml) and treated with propylene oxide (0.17
ml, 0.0017 mol). The crystalline product was finally pur-
ified by crystallization from water yielding: 0.25 g, 39%,
0.10 g, 31%, mp 273–277 ꢀC, ½ꢂŠ À26.2Æ1 ꢀ (c 1, 1 M
578
NaOH). Enantiomeric purity of the compound 41 was
determined by capillary electrophoresis with a-cyclo-
dextrin (Dzygiel et al., 2000) and was estimated to be at
.
least 99% (only one pick on chromatogram, whereas
two of them were observed for the racemate in the same
conditions).
The diisopropyl ether (500 ml) solution contained a
mixture of diastereoisomers as determined by capillary
25
mp 265–267 ꢀC, ½ꢂŠ578+15.5Æ1 ꢀ (c 1, 1 M NaOH).

´
J. Zon et al. / Phytochemistry 59 (2002) 9–21
20
electrophoresis with a-cyclodextrin of the product after
hydrolysis.
Hashimoto, M., Hatanaka, Y., Nabeta, K., 2000. Novel photoreactive
cinnamic acid analogues to elucidate phenylalanine ammonia-lyase.
Bioorg. Med. Chem. Lett. 10, 2481–2483.
Hrubcova, M., Cvikrova, M., Eder, J., Zon, J., Machackova, I., 2000.
ˇ
Effect of inhibition of phenylpropanoid biosynthesis on peroxidase
and IAA-oxidase activities and auxin content in alfalfa suspension
cultures. Plant Physiol. Biochem. 38, 949–956.
Acknowledgements
Janas, K.M., Filipiak, A., Kowalik, J., Mastalerz, P., Knypl, J.S.,
1985. 1-Amino-2-phenylphosphonic acid: an inhibitor of l-phe-
nylalanine ammonia-lyase in vitro. Acta Biochim. Pol. 32, 131–
143.
This work was partially supported by the Polish
Science Foundation Council (project number 3 T09A158
08 to J.Z.). The NMR and IR spectra were recorded by
.
Mrs. Urszula Walkowiak and Mrs. Elzbieta Mroz,
Janas, K.M., Olechowicz, D., 1994. 1-Amino-3-phenylpropylpho-
sphonic acid, the inhibitor of l-phenylalanine ammonia-lyase activ-
ity of higher plants. Acta Biochim. Pol. 41, 191–193.
respectively. The elemental and GC analyses were per-
formed by Mrs. Czes•awa Andrzejewska and Dr. Andrzej
Nosal, respectively (all at The Wroc•aw University of
Technology). At The University of Opole, Ms. Ewa
Rudzinska determined optical purity of (+)-1-amino-3-
phenylpropylphosphonic acid. At ETH Zurich, Mr.
Philippe Roy provided excellent technical assistance in
the biochemical experiments.
Kafarski, P., Lejczak, B., Szewczyk, J., 1983. Optically active 1-ami-
noalkanephosphonic acids. Dibenzoyl-l-tartaric anhydride as an
effective agent for the resolution of racemic diphenyl 1-aminoalk-
anephosphonates. Can. J. Chem. 61, 2425–2430.
Kafarski, P., Lejczak, B., Tyka, R., Koba, Pliszczak, E., Wieczorek,
P., 1995. Herbicidal activity of phosphonic, phosphinic, and phos-
phonous analogues of phenylglycine and phenylalanine. J. Plant
Growth Regul. 14, 199–203.
Kostenyuk, I.A., Zon, J., Burns, J.K., 2001. Phenylalanine ammonia
lyase gene expression during abscission in citrus. Physiologia Plan-
tarum, in press.
References
Kotzyba-Hibert, F., Kapfer, I., Goeldner, M., 1995. Recent trends in
photoaffinity labeling. Angew. Chem., Int. Ed. Engl. 34, 1296–1312.
Kowalik, J., Kupczyk-Subotkowska, L., Mastalerz, P., 1981. Pre-
paration of dialkyl 1-aminoalkanephosphonates by reduction of
dialkyl 1-hydroxyiminoalkanephosphonates with zinc in formic
acid. Synthesis 57–58.
Amrhein, N., Godeke, K.H., 1977. a-Aminooxy-b-phenylpropionic
acid—a potent inhibitor of phenylalanine ammonia-lyase in vitro
and in vivo. Plant Sci.Lett. 8, 313–317.
Bretscheider, T., Miltz, W., Munster, P., Steglich, W., 1988. New
syntheses of a-amino acids based on N-acylimino acetates. Tetra-
hedron 44, 5403–5414.
Laber, B., Kiltz, H.-H., Amrhein, N., 1986. Inhibition of phenylala-
nine ammonia-lyase in vitro and in vivo by 1-amino-2-phenylethyl-
phosphonic acid, the phosphonic analogue of phenylalanine. Z.
Naturforsch. 41c, 49–55.
Butler, C.L., Cretcher, L.H., 1933. Note on the preparation of diben-
zoyl-d-tartaric acid. J. Am. Chem. Soc. 55, 2605–2606.
Chen, H., Chen, F., 2000. Effect of yeast elicitor on the secondary
metabolism of Ti-transformed Silvia miltiorrhiza cell suspension
cultures. Plant Cell Reports 19, 710–717.
Langer, B., Rother, D., Retey, J., 1997. Identification of essential
amino acids in phenylalanine ammonia-lyase by site-directed muta-
genesis. Biochemistry 36, 10867–10871.
Chen, M., McClure, J.W., 2000. Altered lignin composition in pheny-
lalanine ammonia-lyase—inhibited radish seedlings: implication for
seed-derived sinapoyl ester as lignin precursors. Phytochemistry 53,
365–370.
Lewandowicz, A., Jemielity, J., Kanska, M., Zon, J., Paneth, P.,
1999. Tritium secondary kinetic isotope effect on phenylalanine
ammonia-lyase catalyzed reaction. Arch. Biochem. Biophys. 370,
216–221.
Doyle, F.P., Fosker, G.R., Nayler, J.H.C., Smith, H., . Derivatives of
6-aminopenicillanic acid. Part I. a-Aminobenzylpenicillin and some
related compounds. J. Chem. Soc.1962 1440–1444.
—ukszo, J., Tyka, R., 1977. New protective groups in the synthesis of
1-aminoalkanephosphonic acids and esters. Synthesis 239–240.
Maier, L., Diel, P.J., 1991. Organic phosphorous compounds. Part 94.
Preparation, physical and biological properties of aminoaryl-
methylphosphonic and -phosphonous acids. Phosphorus Sulfur and
Silicon 57, 57–64.
.
Dzygiel, P., Rudzinska, E., Wieczorek, P., Kafarski, P., 2000. Deter-
mination of opitical purity of phosphonic acid analogues of aro-
matic amino acids by capillary electrophoresis with a-cyclodextrin.
J. Chromatogr. A895, 301–307.
Gancarz, R., Wieczorek, J.S., 1977. Auseful method for the prepara-
tion of 1-aminoalkanephosphonic acids. Synthesis 625.
Gloge, A., Zon, J., Kovari, A., Poppe, L., Retey, J., 2000. Phenylala-
nine ammonia-lyase: the use of its broad substrate specificity for
mechanistic investigations and biocatalysis-synthesis of l-arylala-
nines. Chem. Eur. J. 6, 3386–3390.
Maier, L., Diel, P.L., 1994. Synthesis, physical and biological proper-
ties of the phosphorus analogues of phenylalanine and related
compounds. Phosphorus Sulfur and Silicon 90, 259–279.
Oleksyszyn, J., 1987. An amidoalkylation of trivalent phosphorus
compounds with P(O)H functions including acetic acid solutions of
PCl3, RPCl2 or R2PCl, diesters of phosphorous acid and phospho-
rus-III-acids. J. Prakt. Chem. 329, 19–28.
G•owiak, T., Sawka-Dobrowolska, W., Kowalik, J., Mastalerz, P.,
Soroka, M., Zon, J., 1977. Absolute configuration of optically
active aminophosphonic acids. Tetrahedron Lett. 18, 3965–3968.
Green, D., Patel, G., Elgendy, S., Baban, J.A., Claeson, G., Kakkar,
V.V., Deadman, J., 1994. The synthesis of 1-aminobenzylpho-
sphonic acids from benzylidenediphenylmethylamines, for use as
structural units in antithrombotic tripeptides. Tetrahedron 50,
5099–5108.
Oleksyszyn, J., Gruszecka, E., 1981. Amidoalkylation of phosphorous
acid. Tetrahedron Lett. 22, 3537–3540.
Oleksyszyn, J., Subotkowska, L., Mastalerz, P., 1979. Diphenyl 1-
aminoalkanephosphonates. Synthesis 985–986.
Rieche, A., Gross, H., Hoft, E., 1960. Synthesen aromatischer Alde-
hyde mit Dichlormethyl-alkylathern. Chem. Ber. 93, 88–94.
Ruuhola, T.M., Julkunen-Tiitto, M.-R.K., 2000. Salicylates of intact
Salix myrsinifolia plantlets do not undergo rapid metabolic turn-
over. Plant Physiol. 122, 895–905.
Greenstein, J.P., Winitz, M., 1961. Chemistry of the Amino Acids,
Vol. 3. John Wiley & Sons, New York. (pp. 2694–2694)..
Hahlbrock, K., Scheel, D., 1989. Physiology and molecular biology of
phenylpropanoid metabolism. Annu. Rev. Plant Physiol. Plant Mol.
Biol. 40, 347–369.
Schaller, A., Roy, P., Amrhein, N., 2000. Salicylic acid-independent
induction of pathogenesis-related gene expression by fusicoccin.
Planta 211, 599–606.

´
J. Zon et al. / Phytochemistry 59 (2002) 9–21
21
Schuster, B., Retey, J., 1994. Serine-202 is the putative precursor of the
active site dehydroalanine of phenylalanine ammonia lyase. FEBS
Lett. 349, 252–254.
Tyka, R., Hagele, G., 1989. Synthesis of aminobenzylphosphonous
acids via direct amidoalkylation of hypophosphorous acid. Phos-
phorus Sulfur and Silicon 44, 103–107.
Schwede, T.F., Retey, J., Schulz, G.E., 1999. Crystal structure of his-
tidine ammonia-lyase revealing a novel polypeptide modification as
the catalytic electrophile. Biochemistry 38, 5355–5361.
Soroka, M., 1990. The synthesis of 1-aminoalkylphosphonic acids. A
revised mechanism of the reaction of phosphorus trichloride, amides
and aldehydes or ketones in acetic acid (Oleksyszyn reaction). Lie-
bigs Ann. Chem. 331–334.
Zon, J., 1984. Synthesis of diisopropyl 1-nitroalkanephosphonates
from diisopropyl 1-oxoalkanephosphonates. Synthesis 661–663.
Zon, J., 1996. Phosphonic analogues of phenylalanine and histidine as
strong inhibitors of phenylalanine and histidine ammonia-lyases.
Phosphorus Sulfur and Silicon 109–110, 349-352.
Zon, J., Amrhein, N., 1992. Inhibitors of phenylalanine ammonia-
lyase: 2-aminoindan-2-phosphonic acid and related compounds.
Liebigs Ann. Chem. 625–628.
Soroka, M., Jaworska, D., Szczce˛sny, Z., 1990. Synthesis of 1-ami-
noalkylphosphonic acids via amidoalkylation of phosphorous
acid by N,N0-alkylidenebisamides. Liebigs Ann. Chem. 1153–
1155.
Zon, J., Laber, B., 1988. Novel phenylalanine analogues as putative inhi-
bitors of enzymes acting on phenylalanine. Phytochemistry 27, 711–
714.