Characterization of polyphenol oxidase in coffee

Article abstract of DOI:10.1016/S0031-9422(00)00332-0

Polyphenol oxidase (PPO) was characterized in partially purified extracts of leaves (PPO-L) and fruit endosperm (PPO-E) of coffee (Coffea arabica L.). PPO activity was higher in early developmental stages of both leaves and endosperm of fruits. Wounding or exposure of coffee leaves to methyl jasmonate increased PPO activity 1.5-4-fold. PPO was not latent and was not activated by protease treatment. PPO activity was stimulated 10-15% with sodium dodecyl sulphate (SDS) at 0.35-1.75 mM, but at higher concentrations activities were similar to the control samples, without detergent. Prolonged incubation of extracts-with trypsin or proteinase K inhibited PPO activity but pepsin had no effect. Inhibition of PPO with proteinase K was increased in the presence of SDS. PPO activity from both tissues was optimal at pH 6-7 and at an assay temperature of 30°C. Activity was highest with chlorogenic acid as substrate with a K(m) of 0.882 mM (PPO-L) and 2.27 mM (PPO-E). Hexadecyl trimethyl-ammonium bromide, polyvinylpyrrolidone 40, cinnamic acid and salicylhydroxamic acid inhibited PPO from both tissues. Both enzymes were inactivated by heat but the activity in endosperm extracts was more heat labile than that from leaves. The apparent M, determined by gel filtration was 46 (PPO-L) and 50 kDa (PPO-E). Activity-stained SDS-polyacrylamide gel electrophoresis (PAGE) gels and WeStern blots probed with PPO antibodies suggested the existence of a 67 kDa PPO which is susceptible to proteolytic cleavage that generates a 45 kDa active form. (C) 2000 Elsevier Science Ltd.

Full text of DOI:10.1016/S0031-9422(00)00332-0

Phytochemistry 55 (2000) 285±296
www.elsevier.com/locate/phytochem
Characterization of polyphenol oxidase in co€ee
Paulo Mazzafera a, Simon P. Robinson b,*
aDepartamento de Fisiologia Vegetal, Instituto de Biologia, CP 6109, Universidade Estadual de Campinas, 13087-970, Campinas, SP, Brazil
bCSIRO Plant Industry, Horticulture Unit, PO Box 350, Glen Osmond, SA 5064, Australia
Received 26 April 2000; received in revised form 20 July 2000
Abstract
Polyphenol oxidase (PPO) was characterized in partially puri®ed extracts of leaves (PPO-L) and fruit endosperm (PPO-E) of
co€ee (Co€ea arabica L.). PPO activity was higher in early developmental stages of both leaves and endosperm of fruits. Wounding
or exposure of co€ee leaves to methyl jasmonate increased PPO activity 1.5±4-fold. PPO was not latent and was not activated by
protease treatment. PPO activity was stimulated 10±15% with sodium dodecyl sulphate (SDS) at 0.35±1.75 mM, but at higher
concentrations activities were similar to the control samples, without detergent. Prolonged incubation of extracts with trypsin or
proteinase K inhibited PPO activity but pepsin had no e€ect. Inhibition of PPO with proteinase K was increased in the presence of
SDS. PPO activity from both tissues was optimal at pH 6±7 and at an assay temperature of 30ꢀC. Activity was highest with
chlorogenic acid as substrate with a Km of 0.882 mM (PPO-L) and 2.27 mM (PPO-E). Hexadecyl trimethyl-ammonium bromide,
polyvinylpyrrolidone 40, cinnamic acid and salicylhydroxamic acid inhibited PPO from both tissues. Both enzymes were inactivated
by heat but the activity in endosperm extracts was more heat labile than that from leaves. The apparent Mr determined by gel
®ltration was 46 (PPO-L) and 50 kDa (PPO-E). Activity-stained SDS±polyacrylamide gel electrophoresis (PAGE) gels and western
blots probed with PPO antibodies suggested the existence of a 67 kDa PPO which is susceptible to proteolytic cleavage that gen-
erates a 45 kDa active form. # 2000 Elsevier Science Ltd. All rights reserved.
Keywords: Co€ea arabica; Rubiaceae; Co€ee; Endosperm; Leaf; Methyl jasmonate; Polyphenol oxidase; Catechol oxidase; Protease; Wounding
1. Introduction
induced oxidation of phenols can modify the plant pro-
teins, reducing their nutritive value to herbivores.
For most plant tissues PPO is compartmentalized in
plastids, whereas its phenolic substrates are located in
the vacuoles (Vaughn and Duke, 1984; Vaughn et al.,
1988). Action of PPO only occurs when this compart-
mentation is disrupted after tissue wounding, as
observed in diseased tissues or those damaged by insect
attack. However, an increase in PPO mRNA has been
detected in arti®cially wounded tissues (Boss et al.,
1995; Thipyapong et al., 1995). In some plants the
increase in PPO activity and PPO mRNA have been
related to production of compounds (jasmonic acid and
methyl jasmonate) of the octadecanoid signal transduc-
tion pathway (Constabel et al., 1995; Czapski and
Saniewski, 1988), but this does not appear to be uni-
versal, as several plants had little or no induction of
PPO by wounding or treatment with methyl jasmonate
(Constabel and Ryan, 1998).
Polyphenol oxidase (PPO; EC 1.10.3.2) is a copper-
containing enzyme responsible for hydroxylation of
monophenols to o-diphenols and oxidation of o-diphe-
nols to o-diquinones. PPO is widely distributed in
higher plants but it still has no de®ned biological func-
tion, although many possible roles have been proposed
(Hind et al., 1995; Lax and Vaughn, 1991; Spruce et al.,
1987; Ste€ens et al., 1994; Trebst and Depka, 1995;
Vaughn and Duke, 1984; Vaughn et al., 1988). At pre-
sent, the most likely functions for PPO are its involve-
ment in plant resistance against diseases (Bashan et al.,
1985; Goy et al., 1992; Ray and Hammerschmidt, 1998)
and against insect herbivory (Felton et al., 1989, 1992).
While in the ®rst case, an increase in PPO activity has
been observed in incompatible interactions, in the sec-
ond, the enzyme seems to induce an antinutritive
defence. Upon wounding, the quinones formed by PPO
Due to their high reactivity, the quinone molecules
can react with each other as well as with other sub-
stances in the wounded tissue, forming a variety of
* Corresponding author. Fax: +618-83038601.
E-mail address: simon.robinson@pi.csiro.au (S.P. Robinson).
0031-9422/00/$ - see front matter # 2000 Elsevier Science Ltd. All rights reserved.
PII: S0031-9422(00)00332-0

286
P. Mazzafera, S.P. Robinson / Phytochemistry 55 (2000) 285±296
brown or black compounds, which are not the direct
result of PPO activity (Mayer and Harel, 1991). This
makes PPO a very important enzyme in the food
industry since during the processing of fruits and vege-
tables any wounding may cause cell disruption and lead
to quinone formation. The appearance of the product
may be a€ected, but also the taste and its nutritional
value, often decreasing the quality of the product
(Martinez and Whitaker, 1995; Whitaker, 1995).
In co€ee the relationship between co€ee quality and
PPO activity is not well known, although it is well
established that low quality co€ee has low PPO activity
(Amorim and Silva, 1968; Amorim and Amorim, 1977;
Amorim and Melo, 1991). Many factors associated with
reduced beverage quality, such as cultural aspects,
method of bean processing and time and storage condi-
tions, have been associated with low PPO (Cli€ord,
1985a). Upon storage in inadequate conditions, there is
deterioration of co€ee quality, which is followed by a
decolouration of the beans (green to yellowish) and
decrease of PPO activity (Melo et al., 1980).
The decrease of PPO in co€ee during quality dete-
rioration has been attributed to the exhaustion of sub-
strates, mainly chlorogenic acids, the major phenolics in
co€ee seeds (Cli€ord, 1985b). According to Cli€ord
(1985a), the sequence of events leading to low PPO are
membrane damage, enzyme activation, the oxidation of
chlorogenic acids to quinones, and enzyme inhibition by
the quinones. It has been observed that humidity lower
than 10% leads to marked disruption in membranes of
co€ee seeds (Begnami et al., 1999). Low quality co€ee
beans have marked alterations at the ultra-structural
level (Dentan and Illy, 1985). Therefore, the low bev-
erage quality may be related to membrane damage but
not with PPO per se (Cli€ord, 1985a). However, since
quinones are formed during the process, it cannot be
ruled out that they play some role in the loss of bev-
erage quality. Indeed, it has been shown that soluble
protein contents decline during co€ee storage (Cli€ord,
1985a), and this could be related to quinone complexa-
tion. To our knowledge, the contribution of the oxida-
tion and polymerization products to co€ee quality have
not yet been determined.
2. Results and discussion
2.1. Developmental changes
PPO was investigated in leaves and fruit endosperms
of co€ee at di€erent developmental stages (Fig. 1). As
observed in other plants (Dry and Robinson, 1994;
Flurkey and Jen, 1978; Rathjen and Robinson, 1992),
young tissues showed the highest activities on a fresh
weight basis in leaves of co€ee plants (Fig. 1A). PPO
activity (PPO-L) was high in the youngest leaves and
declined with increasing leaf length, both on a fresh
weight and per mg protein basis (Fig. 1A). This indi-
cates that PPO is only synthesized very early in leaf
development and the enzyme is reasonably stable, result-
ing in the decrease in enzyme activity at later stages of
leaf development as the leaf expands.
In co€ee fruits, the ®rst cellular divisions of the
endosperm start 20±30 days after anthesis and pollina-
tion (Mendes, 1941). As the endosperm grows, it occu-
pies the inner fruit space, replacing the perisperm, until
it reaches full size, at approx. 140±150 days after polli-
nation. At 160 days, the embryo is di€erentiated, the
seed is completely formed, and the ripening process
takes place without any further change in size of the
Except for the works on beverage co€ee quality
(comprehensively reviewed in Amorim and Amorim,
1977; Amorim and Melo, 1991; Cli€ord, 1985a) there is
little information about PPO in this crop. PPO was
investigated in compatible and incompatible interac-
tions of co€ee with the causal agent of co€ee leaf rust
(Maxemiuc-Nacache and Dietrich, 1985) and with
nematodes (Mazzafera et al., 1989). PPO was partially
characterised in co€ee beans (Draetta and Lima, 1976)
but there has been no detailed study of PPO in co€ee.
We report here a detailed study in which PPO activity
has been characterized in leaves and endosperm of cof-
fee fruits.
Fig. 1. PPO activity in (A) co€ee leaves and (B) endosperm of fruit at
di€erent developmental stages.

P. Mazzafera, S.P. Robinson / Phytochemistry 55 (2000) 285±296
287
endosperm. The endosperm starts to lose water and
become hard and maturation is completed 34±36 weeks
after pollination. Since total fruit size and weight do not
change signi®cantly during these later stages of devel-
opment, these are not useful parameters to denote fruit
developmental stage and a better indication is given by
plotting the percentage of the whole fruit occupied by
the endosperm. PPO activity in endosperm was lowest
on a fresh weight basis at the earliest developmental
stage collected, when the fruits were dark green and the
endosperm comprised only 10% of the total fruit weight
(Fig. 1B). Activity increased in later stages of fruit
development and this may re¯ect synthesis of PPO or
the decreasing water content of the endosperm. When
PPO was calculated as speci®c activity (per mg of pro-
tein), activity was highest in the endosperm of young
fruits, and decreased with fruit development. This pre-
sumably re¯ects that increase in storage protein levels
during endosperm development (Rogers et al., 1999). As
with the leaves, PPO appears to be synthesized in the
early stages of endosperm development in co€ee.
this induction is not systemic in co€ee, at least among
leaves of di€erent branches.
Although we selected plants with leaves of similar
length to be used in the experiments, the variation in PPO
activity was considerable among plants, as demonstrated
by the controls and the treated plants themselves. The
co€ee plants used in this study were maintained in a
glasshouse and received the same treatments (watering,
fertilization, etc.), however, these variations may re¯ect
the previous history of the plants, since they were
obtained from a commercial nursery and transplanted
to new pots. It has been shown that PPO can respond to
fertilizers and stresses (Bower and van Lelyveld, 1985;
Mayer and Harel, 1991). Constabel and Ryan (1998),
who analysed PPO in several plant species also observed
large variations among repetitions.
Leaf PPO was also induced by methyl jasmonate but
activities in leaves of the untreated opposite branches
were similar to control leaves, again suggesting that
there was not a systemic response among leaves of dif-
ferent branches (Table 1). PPO increased 2.2, 1.5 and
2.4-fold following wounding in the three plants used
here. In a survey for plants in which PPO was induced
by methyl jasmonate, Constabel and Ryan (1998)
detected increases up to 24-fold in tobacco and 46-fold
in tomato following wounding. However, several plants
did not show signi®cant increases and among those
which did respond to the treatment, most had a limited
increase in PPO activity of 2 to 4-fold. Considering the
data in Table 1, co€ee could be grouped with those
plants which have relatively high constitutive levels of
PPO activity, but limited induction by wounding or
methyl jasmonate.
2.2. Wounding and methyl jasmonate induction of PPO
Co€ee leaves wounded by crushing with a haemostat
showed an increase of PPO activity compared to control
leaves on opposite branches (Table 1). The control
leaves had similar PPO activities to untreated plants,
suggesting that PPO is induced by wounding but that
Table 1
Activation of co€ee leaf PPO by wounding and methyl jasmonate
PPO activity (Zkat mg 1 protein)a
2.3. Characterization of PPO activity
Wounding
Methyl jasmonate
Several works have shown that in some plants PPO is
found in a latent state, even after cell disruption, when it
is released from plastids and interacts with vacuolar
phenolics (Ste€ens et al., 1994). In some cases PPO
continues to be latent even after extraction from the
tissue and activation is observed after treatment with
proteases, detergents, (NH4)2SO4, etc. One explanation
for this activation has been that PPO has a proenzyme
form, which is bound to membranes, and needs activa-
tors (Whitaker, 1995). Activation of PPO by SDS was
observed by several researchers (King and Flurkey,
1987; Robinson et al., 1993; Sanchez-Ferrer et al.,
1989a,b; Sojo et al., 1998), although signi®cant variation
may be seen depending on the plant source (Flurkey,
1986). In general, activation is observed at low
concentrations of SDS, and higher concentrations may
inhibit activity (Sanchez-Ferrer et al., 1989b; Sojo et al.,
1998). Interestingly, co€ee PPO was slightly activated
by SDS only between 0.35 and 1.75 mM and higher
concentrations did not have a major inhibitory e€ect
Plant 1
Plant 1
Treated leaf
Control leaf
0.465
0.215
Treated leaf
Control leaf
1.205
0.291
Plant 2
Plant 2
Treated leaf
Control leaf
0.442
0.294
Treated leaf
Control leaf
0.361
0.243
Plant 3
Plant 3
Treated leaf
Control leaf
0.921
0.386
Treated leaf
Control leaf
0.461
0.420
Absolute controls
Plant 4
1
2
0.271
0.360
Treated leaf
Control leaf
Absolute controls
0.636
0.466
1
2
3
4
0.633
0.576
0.411
0.388
a
Treated leaf indicates the leaf that was wounded or treated with
jasmonate; control leaf indicates the leaf of the opposite (untreated)
branch in the same plant; absolute control indicates the leaf of an
untreated plant.

288
P. Mazzafera, S.P. Robinson / Phytochemistry 55 (2000) 285±296
(Fig. 2A). Since the activities at higher concentrations
were comparable to the controls, this would indicate that
the activation at low SDS would not be a consequence of
membrane solubilization or protein unfolding. All
subsequent experiments were carried out with 3.5 mM
SDS in the reaction bu€er.
Optimum PPO activity was observed between pH 6
and 7 (Fig. 2B) and the optimum temperature was
between 25 and 30ꢀC for both leaf and endosperm PPO
(Fig. 2C). This is in agreement with the optima of pH 6.0
and 30ꢀC determined by Draetta and Lima (1976) for
mature co€ee beans. It is interesting to note that at
pH 8, PPO activity was signi®cantly lower in Tris-HCl
than in Na±Pi bu€er (Fig. 2B). Galeazzi et al. (1981)
studied the e€ect of the extracting and dialysing media
on the banana PPO activity, and observed an activation
by Na±Pi bu€er. The authors suggested that Pi would
stabilize the enzyme and this may explain the higher
activity in Na±Pi bu€er. Alternatively, co€ee PPO
might be inhibited by Cl as has been reported for some
other tissues (Robert et al., 1998). PPO-E was more
sensitive to higher temperature than PPO-L (Fig. 2C).
The best substrate for both co€ee leaf and endosperm
PPO was 5-ca€eoylquinic acid (5CQA; chlorogenic
acid), followed by 4-methylcatechol (Table 2). However,
PPO-E showed lower activity with 4-methylcatechol
than PPO-L. In addition, l-3,4-dihydroxyphenylalanine
(DOPA) was only oxidized by PPO-E. Since there was
not activity towards hydroquinone, a para-diphenol,
and syringaldazine, substrates for laccases, both of the
co€ee PPO enzymes appear to be of the catechol-
oxidase type.
Results from previous work with PPO from mature
co€ee beans di€er considerably with our results, regard-
ing the substrate anity. Draetta and Lima (1976)
observed in spectrophotometric assays that catechol was
the best substrate for PPO (100%), followed by DOPA
(39%), pyrogaloll (3.7%), 5CQA (0.5%), guaiacol
(0.45%) and tyrosine (0.2%). However, Draetta and
Lima (1976) used a spectrophotometric assay in which
Table 2
Substrate anity of PPO from leaves and endospems of co€eea
Relative PPO activityb
Substrates
Leaves
Endosperm
Chlorogenic acid
4-Methylcatechol
Ca€eic acid
100
87.9
60.6
18.2
0
100
74
67.5
15.6
3.9
0
Catechol
DOPA
Protocatechuic acid
Tyrosine
0
0
0
Hydroquinone
p-Coumaric acid
Syringaldazine
0
0
0
0
0
0
a
Fig. 2. (A) E€ect of SDS on PPO activity and (B) determination of
optimum pH and (C) optimum reaction temperature of the enzyme
from co€ee leaves and endosperm. The optimum pH was investigated
with the bu€ers citrate±Pi (pH 3, 4, 4.5, 5 and 6), Na±Pi (pH 6, 7 and
8) and Tris±HCl (pH 8 and 9).
PPO activity was determined as oxygen uptake with each of the
substrates at a ®nal concentration of 1 mM.
b
Activity is expressed relative to chlorogenic acid which was 0.23
1
1
Zkat mg protein in leaves and 0.25 Zkat mg protein in endo-
sperms.

P. Mazzafera, S.P. Robinson / Phytochemistry 55 (2000) 285±296
289
the extracts were incubated for 30 min. Although sev-
eral reports in the literature used spectrophotometric
assays to measure PPO activity, this method has several
disadvantages. The presence of reducing agents in the
assay can cause a delay in colour development, under-
estimating enzyme activity, and polymerization of the
quinone products or reaction with other components of
the extract can also result in incorrect estimates of
activity (Mayer and Harel, 1991). In both cases, the
colour formation may not be linear with time and,
therefore, initial rates of reaction must be measured. We
repeated the measurement of PPO activity in mature
co€ee endosperm extracts using spectrophotometric
assays and measuring the initial rate of reaction. We
found that 5CQA was a better substrate (12.9 absor-
Robinson, 1994) and apple (Murata et al., 1992). On the
other hand, Robinson et al. (1993) observed that PPO
activity in the skin and sap of mango fruits measured
with 5CQA as substrate was only 20 and 76%,
respectively, of the activity detected with 4-methylca-
techol.
In general, PPO has low anity for its substrates and
the reported Km values are often in the 1±10 mM range
(Shomer et al., 1979; Golbeck and Cammarata 1981;
Sanchez-Ferrer et al., 1989b) although Km's as low as
0.02±0.12 mM were reported for PPO from apple
(Murata et al., 1992) and potato tuber (Patil and
Zucker, 1965). When the concentration of 5CQA was
varied in the normal assay for co€ee extracts, the cal-
culated Km for 5CQA was 0.88 mM for PPO-L and
2.27 mM for PPO-E (data not shown). Thus both of the
co€ee PPOs appear to have a relatively low anity for
their substrates.
Chlorogenic acids are the main phenolics in Co€ea
arabica (5±7% on a dry weight basis), and 5CQA repre-
sents approximately 70% of the total phenolics (Cli€ord,
1985b). Estimates of 5CQA in leaves of the same species
also showed signi®cant contents, ranging from 0.45%
(Baumann et al., 1991) to 0.56% (Sartor and Mazzafera,
2000). Other co€ee species showed 1.7% of this mono-
isomer (Sartor and Mazzafera, 2000). Therefore, it is very
likely that chlorogenic acids are the major substrates of
1
bance units min mg protein 1) than pyrogallol (1.9),
DOPA (1.65) and guaiacol (0) when the initial activity
in the ®rst minute was used to calculate the activities
(data not shown), con®rming the results obtained by
measuring oxygen uptake.
These observations are consistent with those made by
measuring PPO activity as oxygen uptake (Table 2) and
con®rm that 5CQA is the preferred substrate for co€ee
PPO. The relative activity of PPO with di€erent sub-
strates varies considerably from plant to plant. For
example, 5CQA was the best substrate for potato tubers
(Patil and Zucker, 1965), sugarcane (Bucheli and
Fig. 3. E€ect of (A) CTAB, (B) PVP-40, (C) cinnamic acid and (D) salicylhydroxamic acid on PPO activity from co€ee leaves and endosperm of
fruit.

290
P. Mazzafera, S.P. Robinson / Phytochemistry 55 (2000) 285±296
co€ee PPO in vivo, and this is consistent with the
observed anities for the di€erent substrates in vitro
(Table 2).
It has been reported that salicylhydroxamic acid,
PVP-40 and cinnamic acid inhibit plant catechol oxi-
dases (Allan and Walker, 1988; Walker and McCallion,
1980). Both co€ee PPOs were inhibited by hexadecyl
trimethyl-ammonium bromide (CTAB), PVP-40, sali-
cylhydroxamic acid and cinnamic acid (Fig. 3A±D).
PPO-E was less sensitive for the ®rst three inhibitors.
Therefore, in addition to substrate anities observed
here (see Table 1), our data indicate co€ee PPO as a
catechol oxidase type. However, it has been also repor-
ted that CTAB inhibits laccases but not catechol oxi-
dases (Walker and McCallion, 1980). This cationic
detergent did not activate PPO from the mango skin
(Robinson et al., 1993), but activated PPO from grapes
(Sanchez-Ferrer et al., 1989a), broad bean (Jimenez and
Garcõa-Carmona, 1996), banana (Sojo et al., 1998) and
potato leaf (Sanchez-Ferrer et al., 1993). On the other
hand, this is not a general rule, since spinach PPO
(Sanchez-Ferrer et al., 1989b) was strongly inhibited at
similar concentrations to those used in this study. Cof-
fee PPOs were inhibited by 30±40% with 11 mM CTAB
(Fig. 3A).
In the assays of co€ee PPO with SDS and CTAB, the
enzyme preparations were exposed to these detergents a
few seconds before addition of the substrate. In both
cases small volumes of the protein extracts (5±10 ml)
were mixed in the O2 electrode chamber with bu€er
(950 ml) containing the detergents at the desired con-
centrations. Just after that, 50 ml of a 20 mM 5CQA
stock solution was added and the initial rate of reaction
was determined. Under these conditions, inhibition of
PPO activity by CTAB was much higher than with SDS
(see Fig. 2A), although SDS inhibition was also
observed with prolonged exposure (see below).
Fig. 4. Heat stability of PPO from (A) co€ee leaves and (B) endo-
sperm of fruits. The heat stability of PPO was investigated by incu-
bating aliquots of the extracts at 50, 70, 80 and 90ꢀC for di€erent
intervals. After incubation, the extracts were chilled on ice and PPO
activity was determined in the standard assay at 25ꢀC.
The thermal stabilities of the co€ee PPOs were stu-
died by incubating the protein extracts at di€erent tem-
peratures for various periods of time and then
determining PPO activity in the standard assay at 25ꢀC
(Fig. 4). As observed in the determination of optimum
reaction temperature (Fig. 2C), PPO-E was more sensi-
tive to high temperatures than PPO-L (Fig. 4B). While
almost no alteration of the activity was observed for
PPO-L at 50ꢀC after 30 min, there was a decrease of
20% in activity with PPO-E, although this was observed
in the ®rst minutes of incubation and the activities
remained stable after that (Fig. 4B). At 80ꢀC, PPO-L
lost 50% of activity after 5 min, while this took less than
2.5 min for PPO-E. Both enzymes lost 50% of activity
at 90ꢀC in less than 1 min incubation. Comparatively,
the co€ee PPOs are more stable to high temperature
than apple (Murata et al., 1992) and sugarcane PPOs
(Bucheli and Robinson, 1994), but less than mango skin
PPO (Robinson et al., 1993).
2.4. Protease treatment
Depending on the plant source, PPO can be present in
a latent form and may then be activated in vitro by salt
treatment, detergents and acid pH (Ste€ens et al., 1994).
There are several reports of activation of PPO in vitro
by proteases (King and Flurkey, 1987; Sanchez-Ferrer
et al., 1989b; Ste€ens et al., 1994; Tolbert, 1973). To
investigate activation of PPO in co€ee, extracts were
incubated with proteases (1:2, wt/wt, total protein in the
extract/protease) for 3 h at room temperature then
assayed for activity using the standard assay, with 0.1%
SDS in the Na±Pi bu€er. Trypsin, proteinase K and
pepsin did not activate the enzyme, conversely, the ®rst
two led to a decrease of activity, proteinase K being
more e€ective, while pepsin did not alter PPO activity in
extracts from leaves or endosperms under these condi-
tions (data not shown). These results are consistent with

P. Mazzafera, S.P. Robinson / Phytochemistry 55 (2000) 285±296
291
the small increase in activity observed with SDS (Fig.
2A) and suggest that PPO activity was not latent in
either of the co€ee extracts.
(Robinson and Dry, 1992). Extracts were separated by
SDS±PAGE and stained for PPO activity using 5CQA.
To investigate the e€ects of proteolysis, extracts of cof-
fee leaves or endosperms were prepared in the presence
and absence of a cocktail of protease inhibitors. The
extracts prepared with protease inhibitors were partially
puri®ed by (NH4)2SO4 precipitation and desalting in
mini-columns of Sephadex G25, then digested with
proteases (1:1, wt/wt protease:protein extract) for 3 h at
room temperature. A di€erent banding pattern was
observed for leaf and endosperm (Fig. 6A, compare
lanes 1 and 6). The leaf gel showed prominent bands of
activity at apparent Mr of 38, 56 and 85 kDa, while
endosperm extracts had several active bands between 35
and 150 kDa. When extracted in the presence of pro-
tease inhibitors the banding patterns were altered. In the
leaf extract, the band at 38 kDa was less intense and the
bands at 53±56 kDa were increased in intensity com-
pared to extracts without protease inhibitors (Fig. 6A,
compare lanes 1 and 2). The band at 56 kDa was also
more intense in the endosperm extracts prepared with
protease inhibitors (Fig. 6A, compare lanes 6 and 7).
Following treatment of the extracts with proteases, the
band at 56 kDa was no longer detected in either leaf
(Fig. 6A, lanes 3±5) or endosperm extracts (Fig. 6A,
An additional experiment was carried out to deter-
mine the combined e€ect of proteases and SDS during
prolonged incubation periods. Protein extracts sub-
jected to digestion with proteinase K and trypsin (1:2,
wt/wt, total protein/protease) were incubated for 24 h
with or without detergent, and the PPO activity deter-
mined at di€erent intervals using the standard activity
assay (Fig. 5). After 24 h at room temperature in the
absence of SDS, PPO activity was only decreased by 30±
50%, indicating that the partially puri®ed preparations
were relatively stable, and there was little e€ect of tryp-
sin in either leaf or endosperm extracts (Fig. 5A and B).
In contrast, proteinase K signi®cantly reduced PPO
activity in both extracts (Fig. 5C and D). PPO activity
was inhibited when the extracts were incubated with
SDS to a greater extent than the inhibition observed
with proteinase K. Trypsin did not cause any additional
inhibition in the presence of SDS (Fig. 5), whereas the
combination of proteinase K and SDS gave a greater
inhibition than with each alone.
PPO activity can be detected in plant extracts follow-
ing separation by partially denaturing SDS±PAGE
Fig. 5. Combined e€ect of (A, B) trypsin and (C, D) proteinase K with SDS on PPO activity from (A, C) co€ee leaves and (B, D) endosperm of
fruits.

292
P. Mazzafera, S.P. Robinson / Phytochemistry 55 (2000) 285±296
were observed at an apparent Mr of 46 kDa for leaf
extracts and 50 kDa for endosperm extracts (data not
shown). To further clarify the di€erent forms of PPO in
co€ee extracts, fully denatured samples were separated
by SDS±PAGE, blotted onto PVDF membranes and
probed with antibodies to puri®ed grape PPO (Fig. 6B).
Grape antibodies revealed two main bands with Mr of
approx. 67 and 45 kDa in the undigested extracts (Fig.
6B, lanes 1, 2, 4 and 5). In the leaf extract, the 67 kDa
form was only signi®cant when the extracts were pre-
pared in the presence of proteinase inhibitors. In the
extracts subjected to trypsin digestion, the 67 kDa band
disappeared and an additional one of 41 kDa was
observed (Fig. 6B, lanes 3 and 6).
PPO is synthesized as a pre-protein of 65±68 kDa
which includes a plastid transit peptide of 10±12 kDa
(Dry and Robinson, 1994). The transit peptide is
removed during import of the PPO pre-protein into the
plastid to yield a mature protein of approx. 60 kDa.
Recently, it was demonstrated that a stromal peptidase
is responsible for the cleavage of the transit peptide of
PPO, during its import into the chloroplast (Kousse-
vitzky et al., 1998). Robinson and Dry (1992) showed
that puri®cation of PPO from broad bean leaves in the
presence of protease inhibitors yielded a 60 kDa pro-
tein, but in vitro proteolysis converted this into a 42 kDa
protein, with no loss of enzyme activity. It appears that
a region of the mature protein is sensitive to proteolysis,
resulting in removal of a 16±18 kDa peptide from the C-
terminal of the protein without diminishing its catalytic
capacity (Robinson and Dry, 1992; Dry and Robinson,
1994). This proteolysis can occur in vitro during pre-
paration, and this may explain numerous reports of
puri®ed PPO proteins of 40±45 kDa in plants (Golbeck
and Cammarata, 1981; King and Flurkey, 1987; Mar-
ques et al., 1994; Rathjen and Robinson, 1992; Dry and
Robinson, 1994). It is not yet clear whether this pro-
teolytic cleavage can also occur in vivo in some plants.
The results in Figs. 5 and 6 demonstrate that co€ee PPO
can also undergo proteolytic cleavage, without sig-
ni®cant loss of enzyme activity. The 67 kDa protein in
co€ee extracts was converted to a 45 kDa protein by
trypsin treatment (Fig. 6B), indicating that the 45 kDa
form in extracts may have arisen as a result of proteo-
lysis during extraction. The presence of the 45 kDa form
in extracts made in the presence of proteinase inhibitors
suggests that inhibition of proteolysis during extraction
was incomplete or that this proteolysis may also occur
in vivo. It is clear that the co€ee PPO proteins contain a
proteolytically sensitive region outside of the catalyti-
cally active part of the protein, as has been observed
with other plant PPO proteins. Proteolysis can occur
during extraction, resulting in 40±45 kDa PPO proteins
which retain full enzyme activity.
Fig. 6. Partially denaturating SDS±PAGE of protein extracts (A) of
leaf (lanes 1±5) and endosperm of fruits (lanes 6±10) stained for poly-
phenol oxidase activity. Lanes 1 and 6 Ð extracts obtained without
proteinase inhibitors; lanes 2 and 7 Ð extracts obtained with protei-
nase inhibitors; lanes 3 and 8, 4 and 9, and 5 and 10 Ð extracts
obtained with proteinase inhibitors and incubated, respectively, with
trypsin, pepsin and proteinase K. Western blot of fully denatured
co€ee leaf (lanes 1±3) and endosperm (lanes 4±6) extracts separated by
SDS±PAGE and probed with grape antibodies (B). Lanes 1 and 4 Ð
extracts obtained without proteinase inhibitors in the extraction buf-
fer; lanes 2 and 5 Ð extracts obtained with proteinase inhibitors in the
extraction bu€er; lanes 3 and 6 Ð extracts obtained with proteinase
inhibitors in the extraction bu€er and incubated with trypsin.
lanes 8±10). In the leaf extracts, protease treatment
resulted in more intense activity staining in the 38 kDa
bands whereas in endosperm extracts there was also the
appearance of new bands between 67 and 93 kDa, which
varied according to which protease was used. An addi-
tional band at approx. 51 kDa was also observed fol-
lowing incubation of the endosperm extract with
proteinase K.
Although activity stained gels can give valuable
information on the presence of di€erent isoforms of
PPO, the results need to be interpreted with great cau-
tion. Because the extracts are not fully denatured, the
mobilities only indicate apparent molecular weight,
which can di€er signi®cantly from actual size (Robinson
and Dry, 1992). When co€ee extracts were analysed by
gel ®ltration in Sephadex S300, single peaks of activity
To our knowledge, there has never been a detailed
comparative investigation of PPO from leaves and fruits

P. Mazzafera, S.P. Robinson / Phytochemistry 55 (2000) 285±296
293
in the same plant. In addition, most of the reports in the
literature are of ¯eshy fruits. In this study, leaf and
endosperm PPO from co€ee has been characterized in
detail. The results suggest that both tissues contain a
catechol oxidase type enzyme, similar to those found in
other plants. The co€ee leaf and endosperm PPOs have
very similar properties and may represent the same or
very similar proteins. In addition, it seems likely that, as
in other plants, co€ee PPO is synthesized as a precursor
protein, which can be hydrolysed to a 45 kDa form
while still retaining catalytic activity.
mini-columns (Pharmacia) using 50 mM Na±Pi, pH 6,
for the elution. The protein concentration in the extracts
was determined using the BioRad protein assay.
For PPO characterization leaves of 2±3 cm and
endosperms from immature fruits were used. Protein
extracts were obtained as above, however, before cen-
trifugation they were passed through two layers of
Miracloth. The supernatants recovered from the cen-
trifugation were treated with (NH4)2SO4 between 20
and 80% saturation, and the 80% pellet was solubilized
in a small volume of 50 mM Na±Pi, pH 6. The extracts
were desalted in a Sephadex G-25 column (160 ml bed
volume, ¯ow rate 1.3 ml min 1, 50 mM Na±Pi, pH 6 as
elution bu€er), glycerol was added to 20% ®nal con-
centration and aliquots of the extracts were frozen and
stored at 30ꢀC. The protein concentration was deter-
mined using the BioRad protein assay.
3. Experimental
3.1. Plant material
The plants used in this study were obtained from a
commercial nursery and transplanted to pots (500 ml)
containing a mixture of soil, sand and composted pul-
verized pine bark. They were maintained in a glasshouse
and at the time leaves were collected, they were approx.
1-year-old. The leaves were classi®ed according their
length.
PPO activity was determined as oxygen uptake in an
oxygen electrode (Hansatech), at 25ꢀC. The reaction
mixture (®nal volume=1 ml) contained 1 mM 5-caf-
feoylquinic acid (5CQA), 3.5 mM SDS, and 50 mM
Na±Pi, pH 6. The reaction was initiated with the addi-
tion of 5CQA and the rate of oxygen uptake over the
®rst minute was used to calculate the activity. Experi-
ments were carried out at least two times with separate
plant extracts and typical results are presented.
Immature fruits were obtained from a co€ee (C. ara-
bica L.) plantation in Brazil and transported to Aus-
tralia refrigerated (8ꢀC). At that developmental stage
the endosperms represented approx. 10% of the fruit
fresh weight and their appearance was ivory and milky.
Immediately after arrival the fruits were cut in two
halves, the endosperms removed with a spatula, and
then frozen in liquid N2 and kept at 80ꢀC. While
removing the endosperms care was taken to avoid peri-
sperm contamination. Mature fruits were obtained from
a co€ee plantation (C. arabica) in Queensland, Aus-
tralia. During the transport to Adelaide the material
was kept refrigerated. They were cut in two halves and
the endosperm removed with a spatula. Immature and
mature fruits were classi®ed according to their colour,
being dark green, green and yellow-red, and the pro-
portion between endosperm and total fruit fresh weight
was determined.
3.3. PPO characterization
Activation of PPO by SDS was tested at a range of
concentrations and in all subsequent activity measure-
ments 3.5 mM SDS was included in the reaction mix-
ture, with 1 mM 5CQA as substrate. Where present,
inhibitors were dissolved in the reaction bu€er (50 mM
Na±Pi bu€er, pH 6), at appropriate concentrations or
added as concentrated solutions just before the addition
of 5CQA.
For the determination of the optimum reaction tem-
perature of co€ee PPO, oxidation of 5CQA was fol-
lowed spectrophotometrically at 420 nm (UV 2401PC-
Shimadzu). In addition to the control of the cell cuvette
temperature, the bu€er was also previously chilled or
heated. The initial 10±20 s interval was used to calculate
the activity. All other PPO activity measurements were
made with an oxygen electrode as described above.
The native Mr of co€ee PPO was estimated with a
S300-Sephacryl column (Pharmacia, 440 ml bed
volume, ¯ow rate of 2 ml min 1) using the Sephadex G-
25 desalted extracts stored at 30ꢀC, as described
above. Fractions (7 ml) were collected and assayed for
protein concentration and PPO activity. The Mr mar-
kers were catalase (238 kDa), BSA (67 kDa), ovalbumin
(44 kDa) and ribonuclease A (13.7 kDa).
3.2. Protein extraction and PPO assay
For the detection of PPO in leaves and endosperm of
fruits at di€erent developmental stages or to investigate
the e€ect of wounding and methyl jasmonate, the sam-
ples were frozen in liquid N2 and ground to a powder
with a mortar and pestle. Endosperm from mature fruits
was ground in a co€ee grinder. Proteins were extracted
with 100 mM Na±Pi, pH 7, containing 2% ascorbic
acid, and 20% PVPP (w/w) and the extracts were kept
at 4ꢀC for 30 min with occasional agitation. After cen-
trifugation (32,600 g, 20 min, 4ꢀC) the supernatants
were recovered and desalted in PD-10 Sephadex G25
PPO from leaves and co€ee endosperms was incu-
bated with proteinase K, trypsin and pepsin at room
temperature (25ꢀC). Proteases were added at 2:1 (wt/wt)

294
P. Mazzafera, S.P. Robinson / Phytochemistry 55 (2000) 285±296
the total protein concentration in the extracts, i.e. 2 mg
protease per mg extract protein. Incubations were car-
ried out for 3 h or at di€erent intervals, up to 24 h
incubation, aliquots were taken and immediately
assayed for the enzyme activity. When present, SDS was
included in the incubation mixtures at a ®nal con-
centration of 0.1%.
puncture made by the needle was immediately closed
with adhesive tape. The plants were maintained under
laboratory conditions and after 24 h a second 50 ml
application was made. After a further 24 h the leaves of
the ®rst pair of the treated branches were collected for
PPO activity determination with a oxygen electrode, as
well as the ®rst leaves of the opposite branches of the
same plants and from untreated plants. Untreated
branches were also enclosed in plastic bags.
In another experiment, protein extracts were obtained
with or without a cocktail of protease inhibitors (Com-
plete, Boehringer Mannheim) and 1 mM phenyl-
methylsulfonyl ¯uoride (PMSF) in the extraction
bu€er. The proteins were pptd with (NH4)2SO4 between
20 and 80% saturation, desalted in PD-10 columns and
assayed for activity. Extracts obtained with protease
inhibitors were incubated with proteases at room tem-
perature (1:1, wt/wt, total protein in the extract/pro-
tease) for 3 h and together with the other extracts, they
were used for activity-stained electrophoresis (partially
denaturing, Robinson and Dry, 1992), using 12% poly-
acrylamide Tris±HCl Ready gels (BioRad) and a mini-
gel system (Mini-Protean Cell, BioRad). The
electrophoresis was carried out at 100 V, and 0.1% SDS
was included in the tank bu€er. For the activity stain-
ing, the gels were washed for 10 min with 50 mM Na±Pi
bu€er, pH 6.0, and the reaction carried out in the same
bu€er containing 2 mM 5CQA and 0.5 mM p-phenyle-
nediamine. In addition, protein extracts obtained with
and without protease inhibitors, as well as those incu-
bated with trypsin, were fully denatured (5 min, 95ꢀC),
separated using the same electrophoretic conditions
described before, then electrotransferred to PVDF
membrane using a Mini Trans-Blot Transfer Cell
(BioRad) and probed with grape antibodies (Rathjen
and Robinson, 1992).
Acknowledgements
P.M. was supported by Fundacao de Amparo a Pes-
quisa do Estado de Sao Paulo (grant 98/0550-9) during
a leave of absence at CSIRO-Plant Industry, Adelaide,
Australia. We wish to thank Melissa Pickering for
excellent technical assistance.
References
Allan, A.C., Walker, J.R.L., 1988. The selective inhibition of catechol
oxidases by salicylhydroxamic acid. Phytochemistry 27, 3075±3076.
Amorim, H.V., Silva, D.M., 1968. Relationship between the poly-
phenol oxidase activity of co€ee beans and the quality of beverage.
Nature 219, 381±382.
Amorim, H.V., Amorim, V.L., 1977. Co€ee enzymes and co€ee qual-
ity. In: Ory, R.L., St Angelo, A.J. (Eds.), Enzymes in Food and
Beverages Processing, ACS Symposium Series, 47. American Che-
mical Society, pp. 27±56.
Amorim, H.V., Melo, M., 1991. Signi®cance of enzymes in non-alco-
holic co€ee beverage. In: Fox, P.F. (Ed.), Food Enzymology, vol. 2.
Elsevier, London, pp. 189±209.
Bashan, Y., Okon, Y., Henis, Y., 1985. Peroxidase, poly-
phenoloxidase, and phenols in relation to resistance against Pseu-
domonas syringae pv. tomato in tomato plants. Canadian Journal of
Botany 65, 366±372.
3.4. Wounding and methyl jasmonate induction
Baumann, T.W., Rodriguez, M.F., Kappeler, A.W., 1991. In: XIV
Colloque Scienti®que International sur la Chimie du Cafe, Associa-
tion Scienti®que International du Cafe, San Francisco, pp. 465±
469.
Both leaves of the second leaf pair (2±3 cm in length)
of a branch with at least four leaf pairs were wounded
by crushing with a haemostat (Constabel et al., 1995).
The main vein was avoided and approx. 50% of the leaf
blade was wounded. This was done in four plants and
only these leaves were wounded in the plants. After 48 h
in glasshouse conditions the leaves were collected for
PPO measurement, using an oxygen electrode, as
described above. As controls, leaves of the second pair
of the opposite branch of the wounded branches were
collected, as well as leaves of the second pair from
unwounded plants. Care was taken to select branches
with leaves of the second pair similar in length.
Begnami, C.N., Cortelazzo, A.L., Mazzafera, P., Dolder, M.A.H.,
1999. Structural and biochemical alterations during the loss of via-
bility of seeds of Co€ea arabica cv. Catuaõ Vermelho. Acta Micro-
scopica 8, 803±814.
Boss, P.K., Gardner, R.C., Janssen, B.J., Ross, G.S., 1995. An apple
polyphenol oxidase cDNA is upregulated in wounded tissues. Plant
Molecular Biology 27, 429±433.
Bower, J.P., van Lelyveld, L.J., 1985. The e€ect of stress history and
container ventilation on avocado fruit polyphenol oxidase activity.
Journal of Horticultural Science 60, 545±547.
Bucheli, C.S., Robinson, S.P., 1994. Contribution of enzymic brown-
ing to colour in sugarcane juice. Journal of Agricultural Food
Chemistry 42, 257±261.
Cli€ord, M.N., 1985a. Chemical and physical aspects of green co€ee
and co€ee products. In: Cli€ord, M.N., Wilson, K.C. (Eds.), Co€ee:
Botany, Biochemistry and Production of Beans and Beverage. AVI
Publishing, Westport, CN, pp. 305±374.
For methyl jasmonate treatment, a branch containing
at least four leaf pairs (three plants) was enclosed with
plastic bags, containing inside a small piece of cotton
wool. Adhesive tape was used to seal the bag at the
botton. A syringe was used to wet the cotton wool with
50 ml methyl jasmonate (Boehringer Mannheim) and the
Cli€ord, M.N., 1985b. Chlorogenic acids. In: Clarke, R.J., Macrae, R.
(Eds.), Co€ee, vol. 1 Chemistry. Elsevier Science, London, pp. 153±
202.

P. Mazzafera, S.P. Robinson / Phytochemistry 55 (2000) 285±296
295
Constabel, C.P., Bergey, D.R., Ryan, C.A., 1995. Systemin activates
synthesis of wound-inducible tomato leaf polyphenol oxidase via the
octadecanoid defense signaling pathway. Proceeding of the National
Academy of Science USA 92, 407±411.
Maxemiuc-Nacache, V., Dietrich, S.M.C., 1985. Changes in phenols
and oxidative enzymes in resistant and susceptible Co€ea arabica
inoculated with Hemileia vastatrix (co€ee rust). Revista Brassica
Botanica 8, 185±190.
Constabel, C.P., Ryan, C.A., 1998. A survey of wound- and methyl
jasmonate-induced leaf polyphenol oxidase in crop plants. Phy-
tochemistry 47, 507±511.
Mayer, A.M., Harel, E., 1991. Polyphenoloxidases and their sig-
ni®cance in fruits and vegetables. In: Fox, P.F. (Ed.), Food Enzy-
mology, vol. 1. Elsevier, London, pp. 373±398.
Czapski, J., Saniewski, M., 1988. The e€ect of methyl jasmonate on
polyphenol oxidase and peroxidase activities in tomato fruit. Bul-
letin of the Polish Academy of Science-Biological Science 36, 127±
132.
Mazzafera, P., Goncalves, W., Fernandes, J.A.R., 1989. Fenois, per-
oxidase e polifenoloxidase na resistencia do cafeeiro a Meloidogyne
incognita. Bragantia 48, 131±142.
Melo, M., Fazuoli, L.C., Teixeira, A.A., Amorim, H.V., 1980. Alter-
acoes fõsicas, quõmicas e organolepticas dos graos de cafe armaze-
nados. Ciencia Cultura 32, 468±471.
Dentan,E., Illy, A. 1985. Etude microscopique de grains de cafe matures,
immatures et immatures fermenteas arabica Santos. XI Colloque
International sur la Chimie du Cafe. Lome, Togo pp. 341±350.
Draetta, I.S., Lima, D.C., 1976. Isolamento e caracterizacao das poli-
fenoloxidases do cafe. Coletanea du Instituto de Tecnologia de Ali-
mentos 7, 13±28.
Mendes, A.J.T., 1941. Cytological observations in Co€ea. VI. Embryo
and endosperm development in Co€ea arabica L. American Journal
of Botany 28, 784±789.
Murata, M., Kurokami, C., Homma, S., 1992. Puri®cation and some
properties of chlorogenic acid oxidase from apple (Malus pumila).
Bioscience Biotechnology Biochemistry 56, 1705±1710.
Patil, S.S., Zucker, M., 1965. Potato phenolases. Puri®cation and
properties. Journal of Biological Chemistry 240, 3938±3943.
Rathjen, A.H., Robinson, S.P., 1992. Aberrant processing of poly-
phenol oxidase in a variegated grapevine mutant. Plant Physiology
99, 1619±1625.
Dry, I.B., Robinson, S.P., 1994. Molecular cloning and characterisa-
tion of grape berry polyphenol oxidase. Plant Molecular Biology 26,
495±502.
Felton, G.W., Donato, K., Delvecchio, R.J., Du€ey, S.S., 1989. Acti-
vation of plant foliar oxidases by insect feeding reduces nutritive
quality of foliage for noctuid herbivores. Journal of Chemical
Ecology 15, 2667±2694.
Felton, G.W., Donato, K.K., Broadway, R.M., Du€ey, S.S., 1992.
Impact of oxidized plant phenolics on the nutritional quality of
dietary protein to a noctuid herbivore, Spodoptera exigua. Journal
of Insect Physiology 38, 277±285.
Ray, H., Hammerschmidt, R., 1998. Responses of potato tuber to
infection by Fusarium sambucinum. Physiology and Molecular Plant
Pathology 53, 81±92.
Robert, C., Rouch, C., Cadet, F., 1998. Inhibition of palmito poly-
phenoloxidase by halide salts. Jurnal of Enzyme Inhibition 13, 285±
290.
Flurkey, W.H., 1986. Polyphenoloxidase in higher plants. Immunolo-
gical detection and analysis of in vitro products. Plant Physiology
81, 614±618.
Robinson, S.P., Dry, I.B., 1992. Broad bean leaf polyphenol oxidase is
a 60 kilodalton protein susceptible to proteolytic cleavage. Plant
Physiology 99, 317±323.
Flurkey, W.H., Jen, J.J., 1978. Peroxidase and polyphenol oxidase
activities in developing peaches. Journal of Food Science 43, 1826±
1831.
Robinson, S.P., Loveys, B.R., Chacko, E.K., 1993. Polyphenol
oxidase enzymes in the sap and skin of mango fruit. Australian
Journal of Plant Physiology 20, 99±107.
Galeazzi, M.A.M., Sgarbieri, V.C., Constantinides, S.M., 1981. Isola-
tion, puri®cation and physicochemical characterization of poly-
phenoloxidases (PPO) from a dwarf variety of banana (Musa
cavendishii, L.). Journal of Food Science 46, 150±155.
Golbeck, J.H., Cammarata, K.V., 1981. Spinach thylakoid polyphenol
oxidase: isolation activation and properties of the native chloroplast
enzyme. Plant Physiology 67, 977±984.
Rogers, W.J., Bezard, G., Deshayes, A., Meyer, I., Petiard, V., Mar-
raccini, P., 1999. Biochemical and molecular characterization and
expression of the 11 S-type storage protein from Co€ea arabica
endosperm. Plant Physiology and Biochemistry 37, 261±272.
Sanchez-Ferrer, A., Bru, R., Garcia-Carmona, F., 1989a. Novel pro-
cedure for extraction of a latent grape polyphenoloxidase using
temperature-induced phase separation in Triton X-114. Plant Phy-
siology 91, 1481±1487.
Goy, P.A., Felix, G., Metraux, J.P., Meins, F. Jr, 1992. Resistance to
disease in the hybrid Nicotiana glutinosaÂNicotiana debneyi is asso-
ciated with high constitutive levels of b-1,3-glucanase, chitinase,
peroxidase and polyphenoloxidase. Physiology and Molecular Plant
Pathology 41, 11±21.
Sanchez-Ferrer, A., Villalba, J., Garcõa-Carmona, F., 1989b. Triton
X-114 as a tool for purifying spinach polyphenol oxidase. Phy-
tochemistry 28, 1321±1325.
Hind, G., Marshak, D., Coughlan, S., 1995. Spinach thylakoid poly-
phenol oxidase: cloning, characterization, and relation to a putative
protein kinase. Biochemistry 34, 8157±8164.
Sanchez-Ferrer, A., Laveda, F., Garcõa-Carmona, F., 1993. Substrate-
dependent activation of latent potato leaf polyphenol oxidase by anio-
nic surfactants. Journal of Agricultural Food Chemistry 41, 1219±1223.
Sartor, R.M., Mazzafera, P., 2000. Ca€eine formation by suspension
cultures of Co€ea dewevrei. Brazilian Archives of Biological Tech-
nology 43 (1), 61±69.
Jimenez, M., Garcõa-Carmona, F., 1996. The e€ect of sodium dodecyl
sulphate on polyphenol oxidase. Phytochemistry 42, 1503±1509.
King, R.S., Flurkey, W.H., 1987. E€ects of limited proteolysis on
broad bean polyphenoloxidase. Journal of Science and Food Agri-
culture 41, 231±240.
Shomer, I., Ben-Shalom, N., Harel, E., Mayer, A.M., 1979. The
intracellular location of catechol oxidase in the olive fruit. Annals of
Botany 44, 261±263.
Koussevitzky, S., Neeman, E., Sommer, A., Ste€ens, J.C., Harel,
E., 1998. Puri®cation and properties of a novel chloroplast stro-
mal peptidase. Processing of polyphenol oxidase and other
imported porecursors. Journal of Biological Chemistry 273,
27064±27069.
Sojo, M.M., Nunez-Delicado, E., Garcõa-Carmona, F., Sanchez-Fer-
rer, A., 1998. Partial puri®cation of a banana polyphenol oxidase
using Triton X-114 and PEG 8000 for removal of polyphenols.
Journal of Agricultural Food Chemistry 46, 4924±4930.
Spruce, J., Mayer, A.M., Osborne, D.J., 1987. A simple histochemical
method for locating enzymes in plant tissue using nitrocellulose
blotting. Phytochemistry 26, 2901±2903.
Lax, A.R., Vaughn, K.C., 1991. Colocalization of polyphenol oxidase
and photosystem II proteins. Plant Physiology 96, 26±31.
Marques, L., Fleuriet, A., Cleyet-Marel, J.-C., Macheix, J.-J., 1994.
Puri®cation of an apple polyphenoloxidase isoform resistant to
SDS-proteinase K digestion. Phytochemistry 36, 1117±1121.
Martinez, M.V., Whitaker, J.R., 1995. The biochemistry and control
of enzymatic browning. Trends in Food Science Technology 6, 195±
200.
Ste€ens, J.C., Harel, E., Hunt, M.D., 1994. Polyphenol oxidase. In:
Ellis, B.E., Kuroki, G.W., Sta€ord, H.A. (Eds.), Recent Advances
in Phytochemistry, Genetic Engineering of Plant Secondary Meta-
bolism, vol. 28. Plenum Press, New York, pp. 275±312.

296
P. Mazzafera, S.P. Robinson / Phytochemistry 55 (2000) 285±296
Thipyapong, P., Hunt, M.D., Ste€ens, J.C., 1995. Systemic wound
induction of potato (Solanum tuberosum) polyphenol oxidase. Phy-
tochemistry 40, 673±676.
Vaughn, K.C., Lax, A.R., Duke, S.O., 1988. Polyphenol oxidase: the
chloroplast oxidase with no established function. Physiology of
Plants 72, 659±665.
Tolbert, N.E., 1973. Activation of polyphenol oxidase of chloroplasts.
Plant Physiology 51, 234±244.
Walker, J.R.L., McCallion, R.F., 1980. The selective inhibition of
ortho- and para-diphenol oxidases. Phytochemistry 19, 373±377.
Whitaker, J.R., 1995. Polyphenol oxidase. In: Wong, D.W.S. (Ed.),
Food Enzymes Ð Structure and Mechanism. Chapman and Hall,
New York, pp. 271±307.
Trebst, A., Depka, B., 1995. Polyphenol oxidase and photosynthesis
research. Photosynthesis Research 46, 41±44.
Vaughn, K.C., Duke, S.O., 1984. Function of polyphenol oxidase in
higher plants. Physiology of Plants 60, 106±112.