Purification and characterization of malonyl-coenzyme A: 21-hydroxypregnane 21-O-malonyltransferase (Dp21MaT) from leaves of Digitalis purpurea L.

Article abstract of DOI:10.1016/j.phytochem.2007.08.025

With respect to the cardenolide pathway and the characterization of enzymes involved in the formation of cardenolides, a malonyltransferase, termed malonyl-coenzyme A: 21-hydroxypregnane 21-O-malonyltransferase (Dp21MaT) has been purified. The enzyme catalyses the transfer of the malonyl moiety from malonyl-coenzyme A to 21-hydroxypregnane substrates. Malonyltransferase activity was checked in several potential starting materials including fresh leaves and cell suspension cultures from different plants. Fresh Digitalis purpurea L. leaves turned out to be the best enzyme source. The purification protocol included ammonium sulphate precipitation, hydrophobic interaction chromatography on Phenylsepharose 6 FF, ion exchange chromatography on Source 30 Q, affinity chromatography on Cibacron Blue 3GA and gel filtration on Superdex 75. Gel filtration and native SDS-PAGE analysis showed that Dp21MaT exists as a monomer with a molecular mass of 27 kDa. Its pI, as determined by isoelectric focusing, was 4.66. The enzyme showed maximal activity at pH 6.5 when incubated at 42 °C. The energy of activation was 29.28 kJ mol^-1, whereas that of inactivation was 48.57 kJ mol^-1. Dp21MaT was purified 252-fold with a yield of about 1%. Hanes plots of kinetic data indicated K_m values of 99 μM (V_max 47.57 μkat kg^-1) and 28.44 μM (V_max 39.4 μkat kg^-1 protein) for 3β-benzoyloxy-5β-pregnane-14β,21-dihydroxy-20-one and malonyl-CoA, respectively.

Full text of DOI:10.1016/j.phytochem.2007.08.025

Available online at www.sciencedirect.com
PHYTOCHEMISTRY
Phytochemistry 69 (2008) 619–626
www.elsevier.com/locate/phytochem
Purification and characterization of malonyl-coenzyme
A: 21-hydroxypregnane 21-O-malonyltransferase (Dp21MaT)
from leaves of Digitalis purpurea L.
*
´
Serge Philibert Kuate, Rodrigo M. Padua, Wilhelm F. Eisenbeiss, Wolfgang Kreis
Department of Pharmaceutical Biology, Institute of Biology, Friedrich-Alexander University of Erlangen-Nuremberg, Staudtstr. 5,
D-91058 Erlangen, Germany
Received 19 July 2007; received in revised form 15 August 2007
Available online 22 October 2007
Abstract
With respect to the cardenolide pathway and the characterization of enzymes involved in the formation of cardenolides, a malonyl-
transferase, termed malonyl-coenzyme A: 21-hydroxypregnane 21-O-malonyltransferase (Dp21MaT) has been purified. The enzyme
catalyses the transfer of the malonyl moiety from malonyl-coenzyme A to 21-hydroxypregnane substrates. Malonyltransferase activity
was checked in several potential starting materials including fresh leaves and cell suspension cultures from different plants. Fresh Digitalis
purpurea L. leaves turned out to be the best enzyme source. The purification protocol included ammonium sulphate precipitation, hydro-
phobic interaction chromatography on Phenylsepharose 6 FF, ion exchange chromatography on Source 30 Q, affinity chromatography
on Cibacron Blue 3GA and gel filtration on Superdex 75. Gel filtration and native SDS–PAGE analysis showed that Dp21MaT exists as
a monomer with a molecular mass of 27 kDa. Its pI, as determined by isoelectric focusing, was 4.66. The enzyme showed maximal activ-
ity at pH 6.5 when incubated at 42 ꢀC. The energy of activation was 29.28 kJ molꢀ1, whereas that of inactivation was 48.57 kJ molꢀ1
.
Dp21MaT was purified 252-fold with a yield of about 1%. Hanes plots of kinetic data indicated Km values of 99 lM (Vmax
47.57 lkat kgꢀ1) and 28.44 lM (Vmax 39.4 lkat kgꢀ1 protein) for 3b-benzoyloxy-5b-pregnane-14b,21-dihydroxy-20-one and malonyl-
CoA, respectively.
ꢁ 2007 Elsevier Ltd. All rights reserved.
Keywords: Digitalis purpurea; Plantaginaceae; Enzyme purification; Malonyl-coenzyme A; 21-Hydroxypregnane 21-O-malonyltransferase; 21-O-
Hydroxypregnanes; Cardenolides; Cardiac glycosides
1. Introduction
lides is a result of their ability to inhibit sodium–potassium
ATPase, allowing calcium to accumulate in the myocytes
and thereby leading to an increase in the force of
myocardial contraction. Other studies demonstrated that
in addition to its inotropic effects, cardenolides have neuro-
hormonal and autonomic actions that may prevent heart
failure. Cardenolides proved to be one of the most effective
drugs to manage chronic heart failure (Dec, 2003; Pervaiz
et al., 2006; Rising et al., 2006). Plants still are the sole
source for their acquisition. To improve the production
of these important biologically active metabolites it is nec-
essary to further understand the biosynthetic pathway. A
malonyltransferase has been reported to play a regulatory
role in cardenolide biosynthesis and this step may be
Cardiac insufficiency is still a major concern in devel-
oped countries given its significant influence on morbidity
and mortality (Jessup and Brozena, 2003). The beneficial
effects of foxglove (Digitalis purpurea L.) on this condition
were already reported as early as 1785 (Withering, 1785).
Presently, 23 different cardiac glycosides (Luckner and
Wichtl, 2000) have been identified in extracts prepared
from D. purpurea. The positive inotropic effect of cardeno-
*
Corresponding author. Tel.: +49 9131 852 8241; fax: +49 9131 852
8243.
E-mail address: wkreis@biologie.uni-erlangen.de (W. Kreis).
0031-9422/$ - see front matter ꢁ 2007 Elsevier Ltd. All rights reserved.
doi:10.1016/j.phytochem.2007.08.025

620
S.P. Kuate et al. / Phytochemistry 69 (2008) 619–626
rate-limiting. This enzyme catalyses the transfer of a malo-
nyl moiety to the 21-hydroxy group of 21-hydroxypregn-
anes, most likely initiating butenolide ring formation
(Stuhlemmer and Kreis, 1996; Kreis et al., 1998; Luber,
2002).
O-malonyltransferase (Dp21MaT) from D. purpurea leaves.
Most of the Dp21MaT activity was pelleted between 20%
and 60% (NH4)2SO4. After re-dissolving the protein, the
solution was passed through a phenylsepharose column.
The purification and the characterization of the malo-
nylCoA: 21-Hydroxypregnane 21-O-malonyltransferase
(Dp21MaT) from young fresh leaves from D. purpurea L.
is described here for the first time.
a100
0.3
0.25
0.2
Protein
content (mg/ml)
0.5 M
0.2
0.15
0.1
80
60
40
20
0
(NH4)2SO4
Enzyme
activity (AU)
0.15
0.1
2. Results
0.05
0
2.1. Starting material
0.05
0
Digitalis lanata, D. purpurea, Isoplexis canariensis,
Isoplexis sceptrum (Plantaginaceae), Galphimia glauca
(Malpighiaceae), cell suspension cultures from D. lanata
strains K3OHD and W.1.4 and D. purpurea underwent
an enzymatic screening for 21-O-malonyltransferase as well
as 21-O-malonylesterase activity. The esterase is an inter-
fering enzyme suspected to impair the activity of the malo-
nyltransferase catalysing the reverse reaction of the enzyme
in question. Considerable 21-O-malonyltransferase activity
was observed in young fresh leaves from D. lanata, D. pur-
purea, and in I. canariensis, whereas hardly any activity was
detectable in I. sceptrum (Table 1). Equally, no activity was
recorded in cell suspension cultures regardless of the
source. Strangely, the young leaves from G. glauca chosen
as negative control for malonyltransferase showed the
highest 21-O-malonylesterase activity but no 21-O-malo-
nyltransferase activity. Finally, young D. purpurea leaves
were used as the starting material, since high 21-O-malo-
nyltransferase activity and only low interfering 21-O-
malonlyesterase were detected in this source.
0
50
100
150
Elution volume (ml)
100
80
60
40
20
0
0.3
0.25
0.2
0.5
0.4
0.3
0.2
0.1
0
b
Protein
content (mg/ml)
0.5 M NaCl
Enzyme
0.15
0.1
activity (AU)
0.05
0
0
20
40
60
80
100
120
140
Elution volume (ml)
80
50
40
30
20
10
0
35
30
25
20
15
10
5
c
60
40
20
0
Protein
content (g/ml)
0.5 M NaCl
Enzyme
activity (AU)
2.2. Purification of the Dp21MaT
0
Five purification steps have been employed for the puri-
fication of malonyl-coenzyme A: 21-hydroxypregnane 21-
0
5
10
15
20
Elution volume (ml)
1
12
d
Table 1
Relative activities of malonyltransferase and malonylesterase in potential
starting materials
Protein content
(g/ml)
10
8
0.8
0.6
0.4
0.2
0
Enzyme activity
Plant material
Relative activitya (%)
(AU)
Malonyltransferase Malonylesterase
6
D. purpurea young leaves
D. lanata young leaves
I. canariensis young leaves
I. sceptrum young leaves
D. lanata W.1.4 cell suspension
D. lanata K3OHD cell
suspension
100
107
80
t
t
t
11
13
21
16
39
25
4
2
0
0
5
10
15
20
Elution volume (ml)
D. purpurea cell suspension
G. glauca young leaves
0
0
35
100
Fig. 1. Purification steps of the Dp21MaT. (a) Hydrophobic interaction
chromatography on Phenylsepharose 6 FF; (b) anion exchange chroma-
tography on Source 30 Q, repeat; (c) affinity chromatography on Cibacron
Blue F3G-A; (d) gel filtration on Superdex 75.
t: trace.
a
The activities of D. purpurea young leaves (malonyltransferase) and G.
glauca (malonylesterase) are taken as 100%.

S.P. Kuate et al. / Phytochemistry 69 (2008) 619–626
621
Table 2
Purification of Dp21MaT from fresh leaves of D. purpurea L.
Purification step
Total protein (mg)
Total activity (pkat)
Specific activity (lkat kgꢀ1
)
Purification (-fold)
Recovery (%)
Crude extracta
(NH4)2SO4 precipitation
Phenylsepharose 6 FF
Source 30 Q
Source 30 Q repeat
Cibacron Blue F3G-A
Superdex 75
776
668
105
16
3.14
0.12
0.025
739
692
598
479
108
11
0.95
1.03
5.7
30
32
1
1.1
6
31
34
100
94
81
65
15
92
240
96
252
1.5
0.8
6
a
After passing through Sephadex G-25.
An ammonium sulphate gradient from 0.5 to 0 M was
applied and the enzyme activity eluted with ammonium
sulphate-free elution buffer. This step led to a six-fold puri-
fication with an 81% recovery of the enzyme activity. The
enzyme adsorbed at pH 9.5 to Source 30 Q from which it
could be eluted with 0.2 M NaCl. Repeating this step, the
enzyme was purified 65-fold with a 31% recovery of its
activity (Fig. 1b). Affinity chromatography on Cibacron
Blue (Fig. 1c) yielded a purification factor of 100. The final
step consisting of gel filtration on Superdex 75 (Fig. 1d) led
to a 252-fold purified enzyme with a 0.81% recovery of
activity. The specific activity at this level was 240 lkat kgꢀ1
protein (Table 2).
Native PAGE and SDS–PAGE were used to identify the
band representing the investigated enzyme during purifica-
tion. After gel filtration, a protein of 27 kDa was seen in all
fractions containing Dp21MaT activity. Similarly, in native
PAGE only the fraction containing the 27 kDa protein
showed Dp21MaT activity.
determined by SDS–PAGE (Fig. 2). The isoelectric point
(pI) of Dp21MaT as determined by isoelectric focusing
was 4.66.
2.3. Temperature and pH optima
The highest Dp21MaT activity was observed at 42 ꢀC,
half-maximal activities at 25 and 60 ꢀC. The Arrhenius plot
of the data covering the activation range indicated the
energy of activation to be 29.28 kJ molꢀ1, whereas the
energy of inactivation was 48.57 kJ molꢀ1 (Fig. 3a). Q10
determined between 30 and 40 ꢀC was 0.88. The influence
of the pH on Dp21MaT activity was examined at 42 ꢀC.
The highest activity was seen at pH 6.5 in 50 mM McIlva-
ine and 50 mM Na–Pi buffer with half-maximal activities at
pH 5.8 and 8.3 (Fig. 3b).
2.4. Acyl-acceptor and donor specificities
This band was eluted from the native PAGE gels and its
malonyltransferase activity demonstrated. The purified
Dp21MaT was a monomer showing a single protein band
in native PAGE and a molecular mass of 27 kDa as
The purified enzyme was used to determine acyl-accep-
tor and donor specificities as well as kinetic constants of
3
2.5
2
kDa
1.5
250
1
0.5
0
Thermal activation
150
Thermal denaturation
100
75
0.003
0.0032
1/T (˚K)
0.0034
0. 0036
50
37
14
12
10
8
Dp21MaT
25
6
15
50 mM Na-Borate
50 mM Tris / HCl
50 mM NaPi
4
2
50 mM McIlvaine
10
0
5
5.5
6
6.5
7
7.5
p H
8
8.5
9
9.5 10
Fig. 2. Dp21MaT silver-stained after SDS–PAGE (right lane). Protein
standards in the left lane. The Dp21MaT fraction analysed here was
obtained from a native PAGE gel by gel elution.
Fig. 3. (a) Thermal inactivation and activation of the Dp21MaT. (b)
Effect of pH on Dp21MaT activity.

622
S.P. Kuate et al. / Phytochemistry 69 (2008) 619–626
OH
OH
Table 3
Substrate specificity of Dp21MaT:acyl-acceptor
O
O
CH3
OH
CH3
OH
Substratea
Relative activityb (%)
CH3
H
CH3
H
1
2
3
4
5
6
7
a
100
200
150
20
0
60
O
O
H
H
H3C
O
O
H
H
2
1
OH
OH
OH
50
O
O
CH3
O
CH3
CH 3
OH
Chemical structures of acyl-acceptors listed in this table are presented
in Fig. 4.
CH3
H
CH3
H
CH 3
H
b
The specific activity seen with 3b-benzoyloxy-5b-pregnane-14b,21-
H
H
H
H
H
dihydroxy-20-one and malonyl-CoA are taken to be 100%.
O
HO
HO
5
H
4
3
Kinetic studies using the standard substrate 1 together
with malonyl-CoA as co-substrate were performed keeping
the concentration of one substrate constant while varying
the other. Hyperbolic plots indicating Michaelis–Menten
kinetics were obtained. Hanes plots [S]/v = f[S] (Fig. 5) of
the data indicated Km values of 99 and 28.44 lM for com-
pound 1 and malonyl-CoA, respectively, and Vmax values
of 47.57 and 39.4 lkat kgꢀ1 protein, respectively, for com-
pound 1 and malonyl-CoA. An excess of 1 did not result in
substrate inhibition, whereas concentrations of malonyl-
CoA greater than 0.3 mM inhibited Dp21MaT.
OH
OH
O
O
CH3
OH
CH3
OH
CH3
H
CH3
H
H
H
O
O
H
H
O
O
H3C
H3C
OAc
OH
O
O
OAc
OH
OAc
OH
OAc
OH
O
O
2.5. Effect of divalent metals and ascorbic acid on the activity
of Dp21MaT
O
O
O
O
CH3
CH3
CH3
CH3
6
7
Various divalent cations were tested for their possible
inhibitory or stimulatory effect on the Dp21MaT activity.
Fig. 4. Structures of 21-hydroxylated steroids that were examined as acyl-
acceptors of Dp21MaT. (1) 3b-Benzoyloxy-5b-pregnane-14b,21-dihy-
droxy-20-one; (2) 3b-acetoxy-5b-pregnane-14b,21-dihydroxy-20-one; (3)
5b-pregnane-3b,4b,21-trihydroxy-20-one; (4) 3b,21-dihydroxypregn-5-en-
20-one (21-hydroxypregnenolone); (5) 21-hydroxypregn-4-en-3,20-dion
(21-hydroxyprogesterone); (6) 21-hydroxy ketone from digitoxin tretraac-
etate; (7) 21-hydroxy ketone obtained from digitoxin.
Dp21MaT. Apart from the standard acyl-acceptor 3b-ben-
zoyloxy-5b-pregnane-14b,21-dihydroxy-20-one (1) we also
examined the specificities of Dp21MaT toward several
other substrates for their possible acceptance in standard
assay conditions, namely 3b-acetoxy-5b-pregnane-14b,21-
dihydroxy-20-one (2), 5b-pregnane-3b,4b,21-trihydroxy-
20-one (3), 21-hydroxypregnenolone (4), 21-hydroxypro-
gesterone (5), 21-hydroxyketone prepared from digitoxin
tretraacetate (6) and 21-hydroxyketone prepared from digi-
toxin (7) (Fig. 4). It was found that 5 was not accepted at
all, whereas substrates 1, 2 and 3 were accepted very well.
Substrates 4, 6, 7 were malonylated albeit to a lesser extent
(Table 3).
Apart from malonyl-CoA, other CoA esters like acetyl-
CoA, benzoyl-CoA and acetoacetyl-CoA were tested
as acyl-donors, but only malonyl-CoA served as an
acyl-donor. In another set of experiments all of these com-
pounds, as well as coenzyme A, were tested for their possi-
ble inhibitory effects, but only coenzyme A (0.4 mM) was
found to impair the malonylation reaction, reducing it by
73%.
Fig. 5. Hanes plot used to determined Km and Vmax values for substrate
3b-benzoyloxy-5b-pregnane-14b,21-dihydroxy-20-one [a] and malonyl-
CoA [b].

S.P. Kuate et al. / Phytochemistry 69 (2008) 619–626
623
Ni2+, Fe2+, Cd2+, Ca2+, Zn2+, Co2+, Ba2+, Mg2+, Mn2+
,
Cu2+ were added as their dichloride salts ranging in con-
centration from 0.1 to 5 mM. None of the ions tested influ-
enced the enzyme activity significantly. Ni2+, Ba2+ and
Mg2+ seemed to slightly enhance Dp21MaT activity.
Ascorbic acid also slightly promoted the activity.
OH
O
OH
O
O
O
CH3
OH
O
O
O
CH3
OH
HSCoA
CH3
CoAS
OH
CH3
H
H
Dp21MaT
H
H
R
R
O
O
H
H
2.6. Stability of the enzyme
Scheme 1. Reaction catalysed by the malonyl-coenzyme A: 21-hydroxy-
The enzyme stability was investigated at different tem-
peratures and with various additives. Samples were stored
either with glycerol 5% and b-mercaptoethanol 30 mM or
without any additive at room temperature, +4, ꢀ20 or at
ꢀ80 ꢀC. The stability study was carried out in sodium
phosphate buffer (50 mM, pH 6.5). Samples stored at
ꢀ80 ꢀC with glycerol and b-mercaptoethanol conserved
considerable enzyme activity for more than 6 months.
After 5 weeks only three of the samples still exhibited
any enzyme activity, namely the samples stored at ꢀ20
and ꢀ80 ꢀC without additives, as well as the above men-
tioned sample stored at ꢀ80 ꢀC with glycerol and b-merca-
ptoethanol. Relative remaining activities were 60%, 80%
and 95%, respectively. Samples stored in b-mercap-
toethanol only lost their enzyme activity already after 3
days of storage at either ꢀ20 or ꢀ80 ꢀC, whereas samples
at RT and at +4 ꢀC still showed 90% and 100% remaining
activity. Except at RT, no significant change occurred
between the first and the third week of storage. In sum-
mary, Dp21MaT can be stored for more than 6 months
at ꢀ80 ꢀC without loss of activity and it is stable for 3
weeks when stored at 4 ꢀC in the presence of 30 mM b-
mercaptoethanol and 5% glycerol.
pregnane 21-O-malonyltransferase (Dp21MaT).
is a specific reaction in all cardenolides (Scheme 1).
Preliminary studies demonstrated that ammonium sulphate
was highly detrimental to Dp21MaT activity (data not
shown). The combination of dithiothreitol (DTT) and
ascorbic acid used as antioxidants during the purification,
especially in hydrophobic interaction chromatography,
proved to be disadvantageous. Using only one of these
agents at a time was less damaging to the enzyme activity
which suggests the formation of an inhibitory complex
between both compounds. Preliminary assays showed that
most of the chlorophyll still present in the crude extract
was pelleted between 0% and 20% saturation using
(NH4)2SO4 precipitation, whereas most of the Dp21MaT
activity was found in the 20–60% ammonium (NH4)2SO4
fraction. The purification procedure described here omits
cation exchange chromatography. At pH 5.5 the enzyme
was only partly adsorbed to a SP Sepharose matrix and
at lower pH values an irreversible loss of activity was
observed. The isoelectric point of Dp21MaT was 4.66,
which corroborates the above observation that the enzyme
could not be fully attached to the ion exchange resin at pH
5.5. Changes in the ionization state of some amino acid side
chains in the enzyme do not contribute to the expression of
the Dp21MaT activity. Amino acids, like histidine, glu-
tamic acid and to some extent aspartic acid, where the side
chain pKa is 6.0, 4.25 and 3.65, respectively (Voet et al.,
2006) may play an essential role in the catalytic activity
of the enzyme (Suzuki et al., 2003).
The Dp21MaT is insensitive to salt addition up to
5 mM. Even Cu2+, known to impair the activity of many
transferases at low concentrations (Suzuki et al., 2001,
2002; Akermoun et al., 2002) did not affect Dp21MaT
activity.
Acylation is a common and biochemically significant
modification of plant secondary metabolites. BAHD acyl-
transferases constitute a large family of acyl CoA-utilizing
enzymes involved in the formation of many plant second-
ary metabolites (St-Pierre and De Luca, 2000). To date,
more than 45 acyltransferases of the BAHD family includ-
ing 10 malonyltransferases have been genetically and bio-
chemically characterized. They were all found to be
monomeric enzymes with a molecular mass ranging from
48 to 55 kDa (D’Auria, 2006). Only the tomato fruit vari-
ety showed a lower molecular mass of 38 kDa (Martin and
Saftner, 1995). Malonyltransferases have also been purified
3. Discussion
Enzyme purification requires careful selection of a suit-
able source for the enzyme (Ersson et al., 1989). The malo-
nyltransferase Dp21MaT is abundant in leaves but not in
suspension-cultured cells of D. lanata and D. purpurea.
However, other cell suspensions have been successfully
used as the starting material for the purification of enzymes
involved in the cardenolide metabolism (Wendroth and
Seitz, 1990; Warneck and Seitz, 1990; Finsterbusch et al.,
1999). Dp21MaT activity in cell suspension might have
been obscured by the presence of the highly active malony-
lesterase during screening of different putative sources of
the enzyme. Attempts to separate Dp21MaT completely
from the esterase have only recently been successful
(Kuate, Hahn, Kreis, unpublished data). Young leaves of
D. purpurea proved to be the best starting material for
enzyme purification taking into account the abundance of
Dp21MaT combined with a low malonylesterase activity.
The Dp21MaT catalyses the transfer of a malonyl moi-
ety from malonyl-CoA to a 21-hydroxypregnane. This is
supposed to initiate the butenolide ring formation, which

624
S.P. Kuate et al. / Phytochemistry 69 (2008) 619–626
from bacteria yielding molecular masses of around 35 kDa.
Dp21MaT measuring only 27 kDa appears to be one of the
smallest malonyltransferases described so far. At a glance it
seems that many authors have investigated plant mal-
onyltransferases. However, taking a closer look, it appears
that the majority of them only concentrated on the biosyn-
thesis of anthocyanins and flavonoids (Callebaut et al.,
1996; Suzuki et al., 2001, 2002, 2003, 2004). Only few have
looked at this type of enzyme in others metabolic pathways
(Martin and Saftner, 1995; Revill et al., 1995; Liu et al.,
2006).
tubes were shaken thoroughly for 5 s and phase separation
was facilitated by centrifugation (2 min at 13,000 · g) (Cen-
triguge Biofuge 13; Heraeus, Nuremberg, Germany). The
organic phase was dried under a stream of air in a dry
block (Dry-Block DB 2A; Techne, Cambridge, England)
and the residue was taken up in 60 ll methanol. This solu-
tion was either analysed by thin layer chromatography or
by high performance liquid chromatography. For the for-
mer, the solution was spotted on silica gel 60 TLC F254
plates which were subsequently developed with dichloro-
methane:methanol:water (80:18:2). Pregnane spots were
dyed with anisaldehyde reagent (3 min, 100 ꢀC) (York,
1990), whereas the method recently described (Kuate
et al., unpublished) was employed for HPLC analysis.
The separation of substrate and product was achieved on
Reprosil-Pur 120-C18, particle size 5 lM, 4.6 · 250 mm).
UV absorbance was measured at 195 and 228 nm. A gradi-
ent (acetonitrile: H2O, adjusted to pH 2.5 with H3PO4) was
used for the separation.
4. Experimental
4.1. Chemicals
3b-Benzoyloxy-5b-pregnane-14b,21-diol-20-one (1), the
substrate for malonyltransferase assay is not commercially
available and was thus synthesized according to the methods
of Euw and Reichstein (1964), Rabitzsch (1971) and Kuate
et al. (unpublished). 3b-Acetoxy-5b-pregnane-14b,21-dihy-
droxy-20-one (2), 5b-pregnane-3b,4b,21-trihydroxy-20-one
(3), 21-hydroxy ketone from digitoxin tretraacetate (6) and
21-hydroxy ketone obtained from digitoxin (7) were
obtained in a similar way. Malonyl-coenzyme A as well as
the compounds 3b,21-dihydroxypregn-5-en-20-one (21-
hydroxypregnenolone) (4) and 21-hydroxypregn-4-en-3,20-
dione (21-hydroxyprogesterone) (5) were purchased from
Sigma (Deisenhofen, Germany) and dimethylsulfoxide from
Serva (Heidelberg, Germany). Chromatography columns
and media used throughout the purification were from
Amersham Biosciences (Freiburg, Germany).
4.4. Purification of Dp21MaT
Every purification step was performed in a cold lab (+3
¨
to 5 ꢀC) unless otherwise stated. Akta prime (Amersham
Biosciences, Freiburg, Germany) equipment was employed
for all column chromatography steps. Buffers used
throughout the enzyme purification contained 30 mM b-
mercaptoethanol. For the concentration of protein solu-
tions Amiconꢁ tubes 10 kDa (Millipore, Billerica, USA)
were used. The amount of protein as well as enzyme activ-
ity were assessed after each purification step.
4.4.1. Preparation of the crude extract
Fresh leaves of D. purpurea (150 g), collected from the
greenhouse, were washed thoroughly with tap water and
dried with tissue paper. The central leaf veins were excised
and leaves cut into small pieces and finally ground in a
Waring blender (Waring, Torrington, USA) in the presence
of 600 ml of Na–Pi buffer (50 mM, pH 7.0) and polyvinyl-
polypyrrolidone (Aug. Hedinger, Stuttgart, Germany)
(20%, w/w). The slurry preparation was homogenized for
30 min and squeezed through four layers of cotton gauze
Mullroꢂ (Hartmann AG, Heidenheim, Germany). The fil-
trate was centrifuged at 20,000 · g for 20 min. The crude
extract was finally filtered through two layers of Mira-
clothꢂ (Calbiochem, La Jolla, USA). Protein concentration
was determined with the method of Bradford (1976) using
bovine serum albumin (BSA) as the protein standard.
4.2. Plant material
D. lanata Ehrh., D. purpurea L., I. canariensis (L) Lindl.
ex G. Don, I. sceptrum L., G. glauca Cav. were grown in
the institute’s greenhouse. Cell suspension cultures of
D. lanata (strains K3OHD and W.1.4) and D. purpurea
were provided by our tissue culture laboratory.
4.3. Standard enzyme assay for Dp21MaT
Dp21MaT activity was assayed routinely using the
method described by Stuhlemmer and Kreis (1996). The
assay contained 0.2 mM substrate in 96% ethanol,
0.2 mM malonyl-coenzyme A in DMSO and 240 ll of pro-
tein solution in a final volume of 250 ll. Samples denatured
for 10 min at 100 ꢀC served as controls. The reaction was
run in 2 ml-tubes for 1 h at 42 ꢀC under agitation
(650 rpm) in a thermomixer (Eppendorf, Hamburg, Ger-
many). At the end of the enzyme reaction 15 ll of testoster-
one (10 mM) were introduced as internal standard.
Subsequently, the medium was acidified with an equal vol-
ume of citrate buffer (50 mM, pH 3.5) and immediately
extracted with 1 ml of dichloromethane. The Eppendorf
4.4.2. Enzyme purification
(NH4)2SO4 was added as a fine powder to the crude
extract while stirring for 30 min in order to achieve 20%
salt saturation. After centrifugation (20,000 · g, 20 min)
the pellet was discarded, the concentration of (NH4)2SO4
in the supernatant increased to 60% saturation and the
solution then treated as described above. The resulting
supernatant was discarded and the pellet obtained was dis-

S.P. Kuate et al. / Phytochemistry 69 (2008) 619–626
625
solved in the 50 mM Na–Pi pH 7.0 containing 0.5 M
(NH4)2SO4 (buffer A) and centrifuged again. The superna-
tant was applied to a Phenylsepharose 6 FF column previ-
ously equilibrated with buffer A and washed with the same
buffer. Proteins were eluted with 150 ml of decreasing
(NH4)2SO4 (0.5–0.0 M) in 0.02 M Na–Pi pH 7.0 (buffer
B) followed by 225 ml of buffer B without (NH4)2SO4.
Fractions of 5 ml each were collected at a flow rate of
0.5 ml minꢀ1. Fractions were analysed by TLC and active
fractions were pooled and concentrated.
The protein was exchanged into buffer C (20 mM Tris/
HCl, pH 9.5). The solution was loaded onto column Source
30 Q equilibrated with buffer D (20 mM Tris/HCl, pH 9.5,
0.5 M NaCl) prior to equilibration with buffer C. The
bound proteins were eluted with a linear gradient of buffer
D at a flow rate of 0.7 ml minꢀ1. Fractions of 2.5 ml each
were collected and checked for Dp21MaT activity. Pooled
active fractions were concentrated and loaded onto the
same column again, employing the same experimental con-
ditions as described above.
The most active fractions were exchanged into buffer E
(20 mM Na–Pi, pH 7.0) and applied to Cibacron Blue
F3G-A FF column (5 ml) previously equilibrated with
20 mM Na–Pi, pH 7.0 containing 0.5 M NaCl (buffer F).
Unbound proteins were washed with buffer E until the pro-
tein content approached zero. Dp21MaT was eluted with a
gradient formed from buffers E and F: 0–6 ml (30% buffer
F), isocratic until 10 ml, 10–20 ml (80% buffer F). Fractions
of 1.0 ml each were collected and enzyme activity was
checked.
4.6. SDS–polyacrylamide gel electrophoresis and isoelectric
focusing
SDS–polyacrylamide gel electrophoresis (PAGE) was
carried out using the BioRad mini PROTEAN 3 slab gel
apparatus (BioRad, Richmond, CA, USA) and the
Laemmli buffer system (Laemmli, 1970). Polyacrylamide
gels of 0.75 mm thickness consisting of 10% separating
gel (pH 8.8) and 5% stacking gel (pH 6.8) were cast and
samples of 17 ll (0.5–0.6 lg) of protein were loaded. The
proteins were denatured by heating them to 100 ꢀC in the
presence of 2-mercaptoethanol for 3 min. Molecular weight
markers (BioRad) ranging in size from 10 to 250 kDa were
used as molecular weight standards. Gels were silver-
stained according to Heukeshoven and Dernick, 1985.
Native PAGE was treated similarly, only denaturing agents
and conditions were avoided.
The isoelectric point (pI) of Dp21MaT was determined
by isoelectric focusing on a Criterionꢃ IEF gel (pH 3–
10). Gels 13.3 cm · 8.7 cm · 1.0 mm (W · L · thickness),
as well as the IEF Markers were from BioRad (Hercules,
CA, USA). Isoelectric focusing was performed using the
Protean IEF Cell and Criterion Precast System (BioRad)
according to the instruction manual provided by the manu-
facturer. Using an electrophoresis power supply (ISCO
Model 494, Lincoln, Nebraska, USA). The experiment
was run at room temperature with a constant voltage step
change of 100 V for 1 h, followed by 250 V for 1 h and
finally focusing at 500 V for 30 min. The voltage was sup-
plied by an electrophoresis power supply (ISCO Model
494, Lincoln, Nebraska, USA). The gel was silver-stained
as described earlier.
The most active fractions were pooled, concentrated and
loaded onto Superdex 75 (Amersham Biosciences, Frei-
burg, Germany), already equilibrated with Na–Pi buffer
(20 mM, pH 7.0) containing 0.15 M NaCl. The enzyme
was eluted using the same buffer. Fractions of 1 ml each
Acknowledgements
were collected at a flow rate of 0.2 ml minꢀ1
.
S.P.K. and R.M.P. acknowledge the financial support
from the Deutscher Akademischer Austauschdienst
(DAAD). We thank Barbara White for linguistic advice.
4.5. Enzyme characterization
Enzyme characterization was carried out using a 34-fold
purified enzyme obtained from the second anion exchange
chromatography on Source 30 Q. The isoelectric point (pI)
of Dp21MaT was determined by isoelectric focusing on a
Criterionꢃ IEF gel (pH 3–10) using the Criteron cell system.
Gels 13.3 cm · 8.7 cm · 1.0 mm (W · L · thickness) as well
as the IEF Markers were from BioRad (Hercules, CA,
USA). Buffers employed to determine the optimal pH were
50 mM McIlvaine for pH 5.5–6.5, 50 mM Na–Pi for pH
6.5–7.5, 50 mM Tris/HCl for pH 7.5–9.0 and 50 mM Na–
Borate for pH 9–9.5. Buffer exchange was carried out with
PD 10 columns (Amersham Biosciences, Freiburg,
Germany). The concentrations of divalent cations tested
were 0.1, 0.5, 1 and 5 mM. Assays without salt addition
served as controls. Hyper32, a hyperbolic regression pro-
gramme for the analysis of enzyme kinetic data (http://
homepage.ntlworld.com/john.easterby/software.html) was
used to determine kinetic constants.
References
Akermoun, M., Testet, E., Cassagne, C., Bessoule, J.J., 2002. Inhibition of
the plastidial phosphatidylcholine synthesis by silver, copper, lead and
mercury induced by formation of mercaptides with the lyso-PC
acyltransferase. Biochim. Biophys. Acta 1581, 21–28.
Bradford, M.M., 1976. A rapid and sensitive method for the quantifica-
tion of microgram quantities of protein utilizing the principle of
protein–dye binding. Anal. Biochem. 72, 248–254.
Callebaut, A., Terahara, N., Decleire, M., 1996. Anthocyanin acyltrans-
ferases in cell cultures of Ajuga reptans. Plant Sci. 118, 109–118.
D’Auria, C.J., 2006. Acyltransferases in plants: a good time to be BAHD.
Curr. Opin. Plant Biol. 9, 331–340.
Dec, G.W., 2003. Digoxin remains useful in the management of chronic
heart failure. Med. Clin. N. Am. 87, 317–337.
´
Ersson, B., Ryden, L., Janson, J.-C., 1989. Introduction to protein
´
purification. In: Janson, J.-C., Ryden, L. (Eds.), Protein Purification:
Principles, High Resolution Methods and Applications. VCH Pub-
lishers, New York, pp. 3–32.

626
S.P. Kuate et al. / Phytochemistry 69 (2008) 619–626
Euw, J.V., Reichstein, T., 1964. Die Biosynthese des Digitoxigenins,
Herkunft des C-20. Helv. Chim. Acta 74, 711–724.
of Metabolic Pathways, vol. 34. Elsevier Science Publishing, Oxford,
pp. 285–315.
Finsterbusch, A., Lindemann, P., Grimm, R., Eckerskorn, C., Luckner,
M., 1999. D5-3b-Hydroxysteroid dehydrogenase/D5-D4-ketosteroid
isomerase from Digitalis lanata. A multifunctional enzyme in steroid
metabolism. Planta 209, 478–486.
Stuhlemmer, U., Kreis, W., 1996. Does malonyl-coenzyme A: 21-
Hydroxypregnane 21-hydroxymalonyltransferase catalyse the first step
in the butenolide ring of cardenolides? Tetrahedron Lett. 37, 2221–
2224.
Heukeshoven, J., Dernick, R., 1985. Simplified method for silver staining
of proteins in polyacrylamide gels and the mechanism of silver
staining. Electrophoresis 6, 103–112.
Jessup, M., Brozena, S., 2003. Hearth failure. N. Engl. J. Med. 348, 2007–
2018.
Suzuki, H., Nakayama, T., Yonekura-Sakakibara, K., Fukui, Y.,
Nakamura, N., Nakao, M., Tanaka, Y., Yamaguchi, M., Kumusi,
T., Nishino, T., 2001. Malonyl-CoA:anthocyanin 5-O-glucoside-60-O-
malonyltransferase from scarlet sage (Salvia splendens) flowers. J. Biol.
Chem. 272, 49013–49019.
Kreis, W., Hensel, A., Stuhlemmer, U., 1998. Cardenolide biosynthesis in
foxglove. Planta Med. 64, 491–499.
Suzuki, H., Nakayama, T., Yonekura-Sakakibara, K., Fukui, Y.,
Nakamura, N., Yamaguchi, M., Tanaka, Y., Kumusi, T., Nishino,
T., 2002. cDNA cloning, heterologous expressions and functional
characterization of malonyl-coenzyme A:anthocyanidin 3-O-glucoside-
60-O-malonyltransferase from dahlia flowers. Plant. Physiol. 130,
2142–2151.
Suzuki, H., Nakayama, T., Nishino, T., 2003. Proposed mechanism and
functional amino acid residues of malonyl-coA:anthocyanin 5-O-
glucoside-60-O-malonyltransferase from flowers of Salvia splendens, a
member of the versatile plant acyltransferase family. Biochemistry 42,
1764–1771.
Laemmli, U.K., 1970. Cleavage of structural proteins during the assembly
of the head of bacteriophage T4. Nature 225, 680–685.
Liu, W., Han, C., Hu, L., Chen, K., Shen, X., Jiang, H., 2006.
Characterization and inhibitor discovery of one novel malonyl-
CoA:Acyl carrier protein transacylase (MCAT) from Helicobacter
pylori. FEBS Lett. 580, 697–702.
Luber, E., 2002. Reinigung der Malonyl-coenzym A: 21-Hydroxypregnan
21-malonyltransferase und Versuche zur Isolierung einer Steroid-21-
hydroxylase aus Digitalis lanata Ehrh. Doctorate Thesis, Universita¨t
Erlangen-Nurnberg.
¨
Suzuki, H., Nakayama, T., Nagae, S., Yamaguchi, M., Iwashita, T.,
Fukui, Y., Nishino, T., 2004. cDNA cloning and functional charac-
terization of flavonol 3-O-glucoside-60-O-malonyltransferase from
flowers of Verbena hybrida and Lamium purpureum. J. Mol. Catal. B:
Enzym. 28, 87–93.
Luckner, M., Wichtl, M., 2000. Digitalis: Geschichte, Biologie, Chemie,
Physiologie, Molekularbiologie, Pharmakologie, Medizinische Anwen-
dung. Wissenschaftliche Verlags-Gesellschaft, Stuttgart.
Martin, N.M., Saftner, R.A., 1995. Purification and characterization of 1-
aminocyclopropane-1-carboxylic acid N-malonyltransferase from to-
mato fruit. Plant Physiol. 108, 1241–1249.
Voet, D., Voet, J.G., Pratt, C.W., 2006. Fundamentals of Biochemistry,
2nd ed. John Wiley & Sons, Hoboken, USA.
Pervaiz, M.H., Dickinson, M.G., Yamani, M., 2006. Is digoxin a drug of
the past?. Cleveland Clin. J. Med. 73 821–824.
Rabitzsch, G., 1971. Ozon/pyridin-abbau der cardenolidtridigitoxoside
digitoxin und digoxin zu isopregnanolontridigitoxosiden. Pharmazie
26, 592–597.
Revill, P.W., Bibb, M.J., Hopwood, D.A., 1995. Purification of a
malonyltransferase from Streptomyces coelicolor A3(2) and analysis
of its genetic determinant. J. Bacteriol. 177, 3946–3952.
Rising, S.E., Welch, J.M., Koo, J.M., 2006. Association of mortality risk
with high serum digoxin concentrations. Hosp. Pharm. 41, 342–347.
St-Pierre, B., De Luca, V., 2000. Evolution of acyltransferase genes: origin
and diversification of the BAHD superfamily of acyltransferases
involved in secondary metabolism. In: John, R.I., Romeo, T., Varin,
L., De Luca, V. (Eds.), Recent Advances in Phytochemistry. Evolution
Warneck, H.M., Seitz, H.U., 1990. 3b-Hydroxysteroid oxidoreductase in
suspension cultures of Digitalis lanata Ehrh.. Z. Naturforsch. 45c, 963–
972.
Wendroth, S., Seitz, H.U., 1990. Characterization and localization of
progesterone 5a-reductase from cell cultures of foxglove (Digitalis
lanata Ehrh.). Biochem. J. 266, 41–46.
Withering, W., 1785. An Account of the Floxglove and Some of its
Medical Uses with Practical Remarks on Dropsy and Other Diseases.
M. Swinney, Birmingham, UK.
York, H., 1990. Thin layer chromatography: reagents and detection
methods. In: Jork, H., Funk, W., Fischer, W., Wimmer, H. (Eds.),
Thin-Layer-Chromatography, vol. 1. VCH Publishers, New York, pp.
195–198.