The reduction of oxidation of food products using dioxygenases

Article abstract of DOI:10.1021/op0255175

A novel way of deoxygenating food products by oxygen scavenging enzymes has been developed, especially suitable for food products comprising mono or polyunsaturated fatty acids or both which are very sensitive to oxidation. It has been found that dioxygen

Full text of DOI:10.1021/op0255175

Organic Process Research & Development 2002, 6, 562−568
The Reduction of Oxidation of Food Products Using Dioxygenases
Liesbeth C. M. Bouwens* and Yvonne E. Bruggeman
UnileVer Research and DeVelopment, Vlaardingen, The Netherlands
Abstract:
A novel way of deoxygenating food products by oxygen
scavenging enzymes has been developed, especially suitable for
food products comprising mono or polyunsaturated fatty acids
or both which are very sensitive to oxidation. It has been found
that dioxygenases, which are enzymes capable of oxidising
substrates by inserting one or two oxygen atoms into the
Figure 1. Reaction mechanism of quercetinase with the
substrates quercetin and oxygen.
substrate using molecular oxygen, are very effective in oxygen
removal in food products. Moreover, no reactive species are
formed during this reaction. Furthermore, the substrates for
dioxygenases are antioxidants, which have the advantage that
a combination-effect can be achieved, particularly because the
substrate retains its antioxidant activity after being oxidised.
The use of quercetinase and catechinase, both dioxygenases, as
oxygen-scavenging enzymes for off-flavour prevention in oil-
in-water (o/w) emulsions was found to be highly efficient, and
the oxygen concentration could be reduced to zero. The enzymes
quercetinase and catechinase were able to deoxygenate o/w
emulsion in the presence of the antioxidants quercetin and
catechin, respectively. Especially the combination of querceti-
nase and quercetin resulted in a very effective off-flavour
reduction of 80-97% as determined by GC analysis of 7 out
of 10 volatiles labeled as off-flavour.
where hydrogen peroxide is formed and in the case of
laccase4,5 where reactive radical cations or quinones can be
formed. Hydrogen peroxide can be effectively removed using
catalase but oxygen is a reactive product in this reaction
which makes the glucose oxidase/catalase system less
efficient.
Dioxygenases consume oxygen via a different mechanism,
that is, by incorporation of oxygen atoms into their substrates,
which has the advantage that no reactive species are formed.
Quercetinase from Aspergillus japonicus is an extracellular
copper containing dioxygenase that is able to incorporate
oxygen to certain flavonoids, such as quercetin and kaempfer-
ol, which leads to decolourisation6,7. Quercetinase is a dimer,
having a molecular weight of 110 kDa, containing two
identical monomers of 55 kDa and containing 1 Cu molecule
per monomer. Catechinase, also from Aspergillus japonicus,
uses flavonoids, such as catechin as substrate to remove
oxygen. Catechinase is a monomeric Fe-containing enzyme
having a molecular weight of 58 kDa (1 Fe/molecule
enzyme). Since dioxygenases are very substrate-specific and
thus it rarely happens that the food product contains the
substrate naturally, substrates have to be added. These
substrates are often antioxidants; for example catechinase
and quercetinase can use catechin and quercetin as substrates,
respectively. This is shown in the reaction mechanism of
the reaction of quercetinase with quercetin and oxygen, given
in Figure 1 and the reaction of catechinase with catechin in
Figure 2.
Introduction
Food products often contain mono or polyunsaturated fatty
acids or both which are very sensitive to oxidation. Oxygen
is one of the origins of product deterioration due to oxidation.
Polyunsaturated fatty acids (PUFAs) are particularly sensitive
to oxidation when applied in emulsions. Options to solve
this problem are: (1) the reduction of oxygen concentration,
(2) addition of antioxidants, (3) lowering storage temperature,
and (4) scavenging of metal ions (mainly Fe, Cu) with
sequestrants. We have found two unique enzymes (dioxy-
genases) that are active to consume oxygen in oil-in-water
emulsions.1 Moreover, these enzymes can use a number of
antioxidants as substrates which has the advantage that a
combination effect can be achieved of 1 and 2. Furthermore,
it is important that no reactive species that can give rise to
further deteriorative reactions are formed during this reaction.
This occurs when applying other proposed oxygen-scaveng-
ing enzyme systems: glucose oxidase2 or lipoxygenase3
The reaction scheme shows that 1 molecule of quercetin
or catechin is needed for the consumption of 1 molecule of
(3) Hildebrand, D. F.; Kemp, T. R.; Andersen, R..; Loughrin, J. H. Method of
reducing odor associated with hexanal production in plant products.
(University of Kentucky Research Foundation). U.S. Patent 5,017,386, 1991.
(4) Petersen B. R.; Mathiasen, T. E. Deoxygenation of a food item using a
laccase. (Novo Nordisk A/S). Patent WO 96/31133, 1996.
* To
whom
correspondence
should
be
addressed.
E-mail:
(5) Petersen, B. R.; Mathiasen, T. E.; Peelen, B.; Andersen, H. Use of laccase
for deoxygenation of oil-containing product such as salad dressing. (Novo
Nordisk A/S). Patent WO 9635768, 1996.
(6) Oka, T.; Simpson, F. J. Quercetinase, a dioxygenase containing copper.
Biochem. Biophys. Res. Commun. 1971, 43, 1-5.
liesbeth.bouwens@unilever.com.
(1) Bruggeman, Y. E.; Bouwens, E. C. M.; Heiden, M.; van der, Kesselring,
M.; Mooren, A. Dioxygenase-based method to suppress oxidation in food
products. (Unilever, N.V.). Patent WO 0072 704, 2002.
(2) Mistry, B.; Min, D. B. Reduction of dissolved oxygen in model dressing
by glucose oxidase-catalase dependent on pH and temperature. J. Food
Science 1992, 57, 196-199.
(7) Oka, T.; Simpson, F. J.; Child, J. J.; Mills, C. Degradation of rutin by
Asperillus flaVus. Purification of the dioxygenase quercetinase. Can. J.
Microbiol. 1971, 17, 111-118.
562
•
Vol. 6, No. 4, 2002 / Organic Process Research & Development
10.1021/op0255175 CCC: $22.00 © 2002 American Chemical Society
Published on Web 06/26/2002

Materials
YSI biological oxygen monitor, oxygen (model 5300),
oxygen electrodes, and oxygen permeable membranes were
used. YSI Bath Assembly, from Yellow Springs Incorporated
Co., U.S.A connected with pen recorder and water bath 25
°C was used.
HPLC model HP1090 with diode array detector from
Hewlett-Packard, U.S.A. was used. RP-HPLC column: ODS
Hypersil C18 column: 3 µm (4.6 mm × 100 mm) and
precolumn ODS Hypersil C18 column: 5 µm (4.6 mm ×
10 mm).
Figure 2. Reaction mechanism of catechinase with catechin
and oxygen.
oxygen. Only in the reaction with quercetinase, one molecule
of CO is formed for each molecule of oxygen that is
consumed. The fact that during the quercetinase reaction
oxygen is replaced by carbon monoxide might be a important
disadvantage for the use of quercetinase in food products.
Quercetinase and catechinase are both extracellular fungal
dioxygenases which are obtained by performing a 100-L
fermentation of Aspergillus japonicus using respectively
quercetin and catechin (10 g/L) as inducer.8 Overexpression
of quercetinase and catechinase in Aspergillus awamori
delivered good quantities of enzyme (at least 0.5 g/L; van
der Helm et al, unpublished data). They are unique because
in general dioxygenases, such as catechol dioxygenases and
protocatechuate dioxygenase, are part of complex intracel-
lular multienzyme systems, which are often membrane-bound
and require external cofactors. In contrast, quercetinase and
catechinase do not depend on an external cofactor. The
crystal structure of quercetinase (quercetin 2,3-dioxygenase)9
and the characterisation of catalytic states of the enzyme by
EPR10 became available only recently. The structure and
kinetics of catechinase have not yet been investigated.
The dioxygenase/antioxidant combination might be a good
method to use for deoxygenation of food and nonfood
products which are sensitive to oxidation, particularly in
emulsions such as mayonnaise and spreads. The aim of this
work is to scope the use of the combination of oxygen-
scavenging dioxygenases and antioxidants for the prevention
of autoxidation of food products, focused on oil-in-water
emulsions. Potential usage of dioxygenases for enhanced
stability and shelf life (but also the antioxidant value) will
be discussed. The emphasis is on the use of dioxygenase/
antioxidant combinations for the reduction in off-flavour
formation which will be evaluated analytically, by measuring
oxygen levels and analyzing the volatiles which are repre-
sentative for off-flavour and, more close to the consumer,
by surveying taste-perception by means of a taste panel of
experts. The effect of deoxygenation of products on the
antioxidant properties will be discussed using antioxidant
activity assays.
Gas chromatograph model GC-17A with Static-Headspace
and FID detector was used from Shimadzu. A capillary CP-
Sil 5CB column (length 25 m, i.d. 0.25 mm, o.d. 0.39 mm,
film thickness 1.2 µm) from Chrompack was used.
Enzymes. Quercetinase from Aspergillus japonicus 2.1
mg/mL and catechinase from Aspergillus japonicus 0.84 mg/
mL, both overexpressed in Aspergillus awamori.8 Both
enzymes were identified, produced, and isolated (>90% pure,
stored in 30 mM MES buffer pH 6 at -80 °C) at Unilever
R&D Vlaardingen, The Netherlands.9
Substrates/Antioxidants. Quercetin dihydrate, (+)-cat-
echin hydrate, and kaempferol were from Fluka Belgium,
and instant green tea (IGT) was from Unilever R&D
Colworth, UK.
Oil. Safflour/linseed oil (SA/LN 85/15) was obtained from
Chempro, The Netherlands, and extra virgin olive oil (first
press) was obtained from Puget, Vitrolles, France.
Methods
Oxygen Consumption Assay. Oxygen levels were de-
termined using an biological oxygen monitor, oxygen
electrodes, and oxygen-permeable membranes calibrated with
air-saturated water at 25°. Oxygen uptake (removal) by the
addition of enzymes (5.6 mg/L catechinase and 0.14 mg/L
quercetinase) was measured in time, and initial oxygen
uptake activity was calculated. For catechinase, catechin was
added in an amount of 1 mM (0.3 g/L). For quercetinase,
quercetin was added as substrate in an amount of 1 mM (0.3
g/L) The specific activity of quercetinase is expressed in units
(U) defined as µmol min-1 mg-1 of oxygen consumed by
quercetinase. The activity of quercetinase for the indicated
batch was: 293 µmol min-1 mg-1 oxygen consumed and
for catechinase: 6.4 µmol min-1 mg-1 oxygen consumed.
Analysis of Reaction Components by HPLC. Samples
were analysed on reverse phase HPLC. An ODS Hypersil
column (C18) 3 µm 4.6 mm × 100 mm was used for
analytical determination of the reaction components (e.g.,
quercetin and catechin). A linear gradient of solvent A (2%
acetic acid/2% acetonitrile/96% Milli Q water pH 2.8) and
solvent B (100% acetonitrile) was used under the following
conditions: t ) 0 min 100% solvent A, t ) 30 min 40%
solvent B, flow: 1 mL/min. Sample preparation was as
follows: samples were 1:1 diluted in a block solution (60%
CH3CN/10%HAc/30% MilliQ) to stop the reaction. Prior to
injection the samples were centrifuged for 10 min at 14 000
rpm using an Eppendorf centrifuge. Sample (20 µL) was
injected on the column using an autosampler. Calibration of
the components was performed using external standard
(8) van der Helm, M. J.; van der Heiden, M.; Hondmann, D. H.; Smits, A.;
Swarthoff, T.; Verrips, C. T. Enzymic bleach composition. (Unilever N.V.).
Patent WO 98 28 400 A2, U.S. Patent 6,107,264, 1998.
(9) Fusetti, F.; Schro¨ter K. H.; Steiner, R. A.; van Noort, P. I.; Pijning, T.;
Rozeboom, H. J.; Kalk, K. H.; Egmond, M. R.; Dijkstra, B. W. Crystal
structure of the copper-containing quercetin 2,3-dioxygenase from As-
pergillus japonicus. Structure 2002, 10, 259-268.
(10) Kooter, I. M.; Steiner, R. A.; van Noort, P. I.; Egmond, M. R.; Dijkstra, B.
W.; Huber, M. EPR characterisation of the mononuclear Cu-containing
Aspergillus japonicus quercetin 2,3-dioxygenase reveals dramatic changes
upon anaerobic binding of substrates. Eur. J. Biochem. 2002, 269, 1-9.
Vol. 6, No. 4, 2002 / Organic Process Research & Development
•
563

Table 1. Oxygen uptake activity of dioxygenases in 25 mM
acetate buffer pH 6.0 and in 60% o/w emulsion (olive
oil/acetate buffer) using 2 % (w/w) Tween as emulsifier
calibration with catechin, detection at 280 nm, and quercetin
at 360 nm. Reference samples with catechin and quercetin
were checked to be clearly resolved. The retention times (tR)
of the elution of catechin, quercetin and their reaction
products were: tR (2-[2-(1,4-dicarboxy-1,3-butadienyl)]-3,4-
dihydro-2H-1-benzopyrane-3,5,7-triol) ) 5 min, tR (catechin)
) 6 min, tR (2-(3′,4′-dihydroxybenzoyloxy)-4,6-dihydroxy-
benzoic acid) ) 12.5 min and tR (quercetin) ) 20 min.
Preparation of Emulsions. Emulsions were made with
40% oil (SA/LN 85/15), 59% w/w buffer (0.15 M adipine
acid pH 5) and 1% w/w/ Tween 60 using an Ultra-Turrax.
Two series of experiments were carried out for off-flavour
analysis by static headspace gas chromatography and the
following samples were:
(1) The effects of quercetinase (17 mg/L) and the substrate
quercetin (6 g//L or 20 mM) were tested, flushed, and filled
with pure oxygen in the headspace. Samples were made of
17 mL of headspace and 3 mL of emulsion. The samples
were capped and incubated for 24 h at 40 °C and then further
incubated at 60 °C for 14 days.
(2) Effects of quercetinase (0.014 mg//L)/quercetinase
(1.5mM) and catechinase (0.14 mg//L)/catechin (1.5mM)
were tested in air-filled headspace. Samples were made of
10 mL of headspace and 10 mL of emulsion. The samples
were capped and incubated for 24 h at 40 °C and then further
incubated at 60 °C for 4 weeks.
Static-Headspace Gas Chromatography (GC) Analysis.
The lipid oxidation of samples was followed by static-
headspace GC analysis of the following volatiles: acetal-
dehyde, propenal, propanal, pentane, 2-tert-butenal, 1-pentene-
3-one, 1-pentene-3-ol, pentanal, pentenal, and hexanal. The
column used was a capillary CP-Sil 5 CB column. The
following settings for the autosampler and for the chromato-
graph were used so that the volatiles were well resolved.
The GC Cycle time for one sample is 25 min including the
cool step to 0 °C.
oxygen uptake activitya
[µmol O2/min‚mg]
[enzyme]
(mg/L)
0% oil
60% oil
substrate
quercetinase
catechinase
0.14
5.6
293 ( 22
351 ( 2
266 ( 34
284 ( 42
1 mM quercetin
1 mM kaempferol
6.99 ( 0.07 6.74 ( 0.46 0.3% (w/v) IGTb
6.40 ( 0.61 6.84 ( 1.41 1 mM catechin
a [oxygen]initial (mM) for 0% oil samples was 0.29 mM and for 60% oil
samples was 0.80 mM. b IGT ) Instant green tea.
colloid mill. Glass jars were filled completely, almost without
a headspace, and immediately closed. The pH of the citric
mayonnaise was 3.4.
Antioxidant Activity Assay. Antioxidant activity was
measured using the standard Trolox equivalent antioxidant
capacity (TEAC) assay and the ferric reducing ability of
plasma (FRAP) method as described by Miller, N. J. et al11
and Benz, I. F. F. et al.,12 respectively. Samples were made
of 0.1 mM quercetin or catechin with and without enzyme
addition in concentrations of 0.1 mg/L quercetinase or 10
mg/L catechinase, respectively, in 20% methanol and 80%
demineralised water, pH 6. Samples were incubated for 60
min at room temperature. In the samples with enzyme, the
product formation was checked to be 100% using HPLC
analysis.
Results
Activity of Dioxygenases. The activity of the two
dioxygenases, quercetinase and catechinase, with and without
the presence of oil is shown in Table 1.
As shown from Table 1 the dioxygenases are both active
in oil in water emulsions, even up to 75% oil (not shown).
The presence of oil did not inhibit the enzyme significantly.
The specific uptake activity of quercetinase per mg enzyme
is 40-50 times higher compared to that of catechinase which
makes quercetinase more preferable for applications (less
enzyme dosage, low cost price). Quercetinase is most active
towards kaempferol, which is in agreement with the results
of Oka.13. They found the reactivity of the substrate is greatly
influenced by the distribution of the hydroxyl substituents
(kaempferol 4′ OH at B-ring where quercetin 3′- and 4′-OH
at B-ring, see Figure 1).
The optimal activity of both dioxygenases is in the pH
range 5-6.5 (data not shown). Their activity at low pH (the
relevant pH for spreads and dressings) is given in Table 2.
Although the activity of both dioxygenases decreased at
more acidic conditions, quercetin was most stable and
showed to be still 92% active at pH 3.5. Thus, oxygen
consumption by quercetinase can be applied in products with
Temperature program: Maximum temperature: 399 °C.
Initial temperature 0 °C. Inititial time 4.00 min. Rate 40 °C/
min. Final temperature 40 °C. Final time 0.00 min. Rate A
5.0 °C/min. Final temperature A 60 °C. Final time A 0.00
min. Rate B 40 °C/min. Final temperature B 200 °C. Final
time B 3.00 min. Total program time 15.5 min.
Pressure program: Initial value:115 kPa. Initial time:
4.00 min. Rate: 10 kPa/min. Final pressure: 125 kPa. Final
time: 0.00 min. Rate A: 1.5 kPa/min. Final pressure A:131
kPa. Final time A: 0.00 min, Rate B: 10.3 kPa/min. Final
pressure B: 167 kPa. Final time B: 3.00 min.
Detector FID: Range: 0. Polarity: 2. Current: 0. H2
Flow: 50 mL/min. Air flow: 500 mL/min. Make up gas
flow: 30 mL/min.
Mayonnaise Preparation. Citric mayonnaise was made
according to a standard protocol for the preparation of
mayonnaise. The following ingredients were used: 80%
vegetable oil, egg yolk, water, vinegar, sugar, glucose, syrup,
mustard, nutritious acid E-270, spices, pigment E-160a,
aroma components. Quercetin (2 g) was added to 6 kg of
mayonnaise, and as the last ingredient 6 mg of quercetinase
was added just before the mayonnaise passed through the
(11) Miller, N. J. et al. Spectrophotometric determination of antioxidant. Redox
Report 1996, 2, 161-171.
(12) Benz, I. F. F. et al. The ferric reducing ability of plasma (FRAP) as a
measure of “antioxidant power” the FRAP assay. Anal. Biochem. 1996,
239, 70-76.
(13) Oka, T.; Simpson, F. J.; Krishnamurty, H. G. Degradation of rutin by
Asperillus flaVus. Studies on the specificity, inhibition, and possible reaction
mechanism of quercetinase. Can. J. Microbiol. 1972, 18, 493-508.
564
•
Vol. 6, No. 4, 2002 / Organic Process Research & Development

Table 2. Activity of quercetinase and catechinase at low pHa
Table 3. Antioxidant activity of the substrates quercetin and
catechin of quercetinase and catechinase, respectively, and
their reaction products
activity
activity
catechinase
quercetinase
% activity
% activity
pHst [µmol O2/min‚mg] [µmol O2/min‚mg] catechinase quercetinase
samples
TEAC (mM)
FRAP (mM)
3.5
4
5
298 ( 3
319 ( 19
324 ( 7
4.8
8.6
15.2
32
57
100
92
98
100
quercetin
quercetin + quercetinase
catechin
0.44
0.44
0.38
0.35
0.31
0.39
0.32
0.24
catechin + catechinase
a Activity at pH 5 was taken as 100 % value.
Antioxidant Activity. The presence of antioxidants is a
key factor in off-flavour formation, and therefor the anti-
oxidant activity of the used antioxidants was measured in
the presence and absence of the dioxygenases. Antioxidant
activity is difficult to quantify since it can be directed towards
the prevention of various radical reactions. We have chosen
to use two assays: the Trolox equivalent antioxidant capacity
(TEAC) assay and the ferric reducing ability of plasma
(FRAP) assay; the data of the antioxidant activity are given
in Table 3.
The antioxidant activity of quercetin is similar to the
antioxidant value of 2-(3′,4′-dihydroxybenzoyloxy)-4,6-di-
hydroxybenzoic acid which is the product of the quercetinase
reaction with quercetin. This means that there is no loss of
the antioxidant value during the reaction!
The antioxidant value of catechin showed a small decrease
when oxygen was incorporated in catechin by the enzyme
catechinase, but the product contained antioxidant value. This
shows that both the antioxidant activity of the substrate as
its oxidation product can contribute to prevent off-flavour
formation.
The antioxidant active moiety of catechin and quercetin
is the B ring described by Jovanovic et al.14 They concluded
that the ring whose radical has the lowest reduction potential
is the antioxidant moiety in any flavenoid. Since the
incorporation of oxygen in quercetin by quercetinase oc-
curred at the C ring, the antioxidant moiety of the B ring
remained intact, although the reduction potential of the B
ring is likely to be different. In the case of catechin the
decrease of antioxidant value could be due to the fact that
the incorporation of oxygen by catechinase did occur at the
B ring.
Off-Flavour Prevention: Oil-in-Water Emulsions. The
production of off-flavour of 40/60 oil-in-water emulsions was
measured in time in absence and presence of quercetin with
and without quercetinase using static headspace GC analysis.
The experiment was done under standard conditions for
optimal off-flavour generation using a large amount (85%
v/v) of headspace filled with pure oxygen and a high
incubation temperature of 60 °C. To scavenge all the oxygen
present it was needed to add a high antioxidant concentration
(20 mM). The amount of the volatiles of the samples after
15 days is given in Table 4.
Figure 3. Oxygen consumption in time by quercetinase (0.14
mg/L) in 12.5% w/v oil-in-water emulsions with 1 mM quer-
cetin, 2% w/v Tween at pH 3.5, 4, and 5.
an acidic pH: for instance mayonnaise (pH 3.8), spreads
(pH between 3.8 and 4.5). As shown, catechinase is also
active at low pH but less than quercetinase. Compared with
that of queretinase, the specific activity of catechinase was
already lower at pH 5 and decreased even more at pH 3.5
which makes this enzyme less preferable.
The oxygen concentration could be reduced to zero by
the addition of the enzymes as shown in the example in
Figure 3 where oxygen concentration against time is given
of oil-in-water emulsions in the presence of quercetinase and
quercetin. This demonstrates the potential usage of enzymes
for a deoxygenation of products on a molecular level,
whereas others (e.g., deoxygenation via packaging or filters)
often reduces oxygen slowly to a certain amount.
The dioxygenases are active at room temperature. Their
initial activity was most optimal at 40 °C, and they are still
very active at 60 °C.
Stability. Both dioxygenases are very stable; at room
temperature their activity remained about 100% for 50 days.
At higher temperatures (till 60 °C) quercetinase was still
100% active after 24 h incubation at 60 °C, and catechinase
kept 85% of its activity. Both enzymes were unstable at 80
°C, within 10 min activity decreased to zero. In presence of
0.5% NaCl their activity remained about 95 %. In presence
of 60% oil their activity was 100%. When 0.5% NaCl was
added to a 60/40 o/w emulsion, at the local NaCl concentra-
tion of 1.2%, their activity remained 80%. The excellent
stability towards salt is a clear benefit for the use of
dioxygenases in applications as spreads and dressings,
especially since other enzymes (as laccases) are generally
very unstable under these circumstances.
As shown in the reference sample high levels of off-
flavour were produced. The combination of quercetin and
(14) Jovanovic, S. V.; Steenken, S.; Hara, Y.; Simic, M. G. Reduction potentials
of flavonoid and model phenoxyl radicals. Which ring in flavonoids is
responsible for anti oxidant activity. J. Chem. Soc., Perkin Trans. 1996, 2,
1-8.
Vol. 6, No. 4, 2002 / Organic Process Research & Development
•
565

Figure 4. Pentane production, representative for most flavour components in oil: 40/60 o/w emulsion at 60 °C (40% SA/LN, 1%
Tween 60, 59% pH 5 adipic acid (0.15 M) buffer solution).
Table 4. Flavour production (in ppm) after 15 days incubation of 40% o/w emulsions filled with pure oxygen
acetaldehyde
propenal
pentane
2-tert-butenal
1-pentene-3-ol
pentanal
pentenal
hexanal
reference
quercetin + quercetinase
quercetin
67.5
4.4
73.3
1.1
0.4
4.5
243.4
12.5
258.9
7.9
1.1
12.9
17.9
4.2
28.4
337.7
6.7
100.4
37.3
8.2
68.3
1627.1
84.2
923.9
Table 5. Flavour production (in ppm) after 4 weeks
incubation of 40% o/w emulsions filled with pure oxygen
prevented the formation of three of the five volatiles, while
in the reference sample high levels of these volatiles were
produced.
samples
acetaldehyde propenal pentane pentanal hexanal
Other volatiles were produced in very low concentra-
tions.
Off-Flavour Prevention in Mayonnaise. For the fol-
lowing taste experiment four citric mayonnaise samples were
prepared, filled in glass jars almost without a headspace:
reference
quercetin/
quercetinase
quercetin
catechin/
catechinase
catechin
3.0
1.6
9.8
8.3
35.3
33.2
2.3
0.9
17.5
8.2
3.6
0.7
9.6
6.5
29.9
43.1
1.5
1.4
12.3
13.3
1.7
8.5
32.2
1.3
10.3
sample 1
sample 2
sample 3
sample 4
no additions
with 1 mM quercetin and 1 mg/L quercetinase
with 1 mM quercetin
quercetinase addition not only reduced the formation of the
volatiles, it also prevented the formation of almost all
volatiles! The addition of quercetin alone reduced the
volatiles for some extent. However after 2 weeks incubation
at 60 °C the addition of quercetin led to enhanced off-flavour
production for most volatiles and reduced only the production
of two volatiles. The formation of pentane, which is a
representative off-flavour component, is shown as function
of time in Figure 4.
A second experiment was carried out under more realistic
storage conditions for food products using an air-filled
headspace instead of pure oxygen and lower concentrations
of antioxidants and enzyme. In this experiment both quer-
cetinase and catechinase were tested on their ability to
prevent off-flavour formation. In this experiment the forma-
tion of 10 volatiles was analysed, but only five volatiles were
produced at reasonable levels. The amount of these volatiles
in the samples after 4 weeks is given in Table 5. As shown
the combination of either quercetinase (0.014 mg/L)/quer-
cetin (1.5 mM) or catechinase (0.14 mg/L)/catechin (1.5 mM)
with 1 mM EDTA
Sample 4 is the standard mayonnaise with the addition of
EDTA as a metal scavenger and can be considered as positive
control. The mayonnaise was stored at 5 °C, at room
temperature (20 °C), and at 37 °C.
Quercetin levels were analysed by RP-HPLC analysis and
showed that in sample 2 quercetin was oxidised for 90%.
This indicates that quercetinase is very active in the mayon-
naise product. The mayonnaise samples were judged for
rancidity every 2 weeks during a total period of 16 weeks
by three experienced persons. The taste results of the products
stored at 5, 20, and 37 °C are presented in Table 6 of which
the data of the 5 °C storage trial is also presented in Figure
5.
As shown the mayonnaise without additives was not stable
as shown by the rapid decrease of the taste scores. The
addition of only quercetin addition had a slight positive
effect. The addition of the combination quercetin/quercetinase
enhanced the stability. As shown in Figure 5, the taste of
566
•
Vol. 6, No. 4, 2002 / Organic Process Research & Development

Table 6. Results of the taste of mayonnaise by a taste panel: high numbers mean a better taste and less off-flavour (score 8 )
good, score 5 ) off)a
1. no additives
2. quercetinase + quercetin
3. quercetin
4. EDTA
storage time (weeks) 5 °C 20 °C 37 °C
5 °C
20 °C
37 °C
5 °C 20 °C 37 °C 5 °C 20 °C 37 °C
2
4
6
8
8
6
6.5
5
7.5
8
8
8
7
7
8
7
8
6.5
8
7
6
7
6
7.5
8
7.5
8
8
6
6.5
6.5
8
8
6
5
5
5.5
5
5
5
n.d.
8
8
8
8
8
8
8
8
8
8
7.5
7
7
6.5
8
7
7
6
5
n.d.
7.5
6.5
5.5
6
6
5
6.5
6.5
6.5
5.5
5
6.5
5.5
5
8
n.d.
10
12
16
n.d.
n.d.
5
a n.d. not determined because sample was too rancid.
80-97% as determined by GC analysis of 7 out of 10
volatiles, while in the references the off-flavour caused
volatile increase.
As a practical example the effect of quercetinase/quercetin
on off-flavour formation in mayonnaise was tested. On the
basis of the taste results it is concluded that quercetinase/
quercetin gave good protection against oxidation, while
quercetin alone did not.
Furthermore, the enzymes are highly stable in oil-in-water
emulsions, active at low dosage levels (0.01-0.1 mg/L for
quercetinase, 0.1-1 mg/L for catechinase), stable against salt
(0.5% NaCl), and still active at relatively low pH, that is,
pH 3.5-5 which is relevant for application in spreads and
dressings.
Discussion
Despite the fact that the combination dioxygenases/
antioxidant as oxygen-scavenging method to prevent off-
flavour in oil-in-water emulsions was found to be highly
efficient, the use of the dioxygenases tested have major
bottlenecks: they are both not food-grade, and therefore
safety clearance is required. However, even if the enzymes
are proven to be safe for consumption, the fact that they are
produced via recombinant DNA technology might hamper
their application in food products due to public opinion.
Quercetinase is preferable over catechinase because it can
be added at very low levels. However, the fact that carbon
monoxide is formed in the quercetinase reaction might be a
disadvantage for the use of this enzyme in food which has
to be investigated further in terms of concentrations, and so
forth. Catechinase would be a better alternative. Another
bottleneck is the limitation on the levels of antioxidant which
may be added to a product. However, levels of 1 mM (0.3
g/L) quercetin or catechin are allowed and in combination
with dioxygenases (0.01-0.1 mg/L quercetinase or 0.1-1
mg/L catechinase) found to be effective for prevention of
off-flavour generation. Therefore we recommend using these
conditions in products where deoxygenation on a molecular
level is desired. Furthermore, they can be used in vitamin
and antioxidant-containing products to prevent them from
being oxidised as, for instance, in tomato paste containing
vitamin C, or in nonfood products containing antioxidants,
such as retinol in skin cream.
Figure 5. Stability of mayonnaise samples, stored at 5 °C.
the mayonnaise with quercetin and quercetinase was observed
to be good during the whole 16 weeks storage at 5 °C. The
best prevention of off-taste was observed when EDTA was
added.
On the basis of these results, especially at 5 and 20 °C,
it is concluded that the combination quercetin and querceti-
nase gives good protection against oxidation, while quercetin
alone does not give a sufficient result (results were compa-
rable with the reference, sample 1). EDTA addition provided
the highest off flavour prevention; however, EDTA is not
suited as general additive for mayonnaise (not accepted in
all countries).
Conclusions
The use of dioxygenases, quercetinase and catechinase,
as oxygen-scavenging enzymes to prevent off flavour in oil-
in-water emulsions was found to be highly efficient. The
enzymes achieve their antioxidant effect, in part, by reducing
the oxygen concentration in the sample to zero, so that no
autooxidation can take place. In combination with the
antioxidant activity of the substrate and its oxidation product,
this has resulted in a decreased formation of off-flavours in
a model system. The combination of quercetinase and
quercetin resulted in very effective off-flavour reduction of
Vol. 6, No. 4, 2002 / Organic Process Research & Development
•
567

Acknowledgment
terdam, Y. Heydari for the production of mayonnaise and
his cooperation in the taste experiment.
At Unilever R&D, Vlaardingen, we thank M. v. d. Heiden
for the isolation and purification of the enzymes, M. S.
Chaara for the GC analysis, and W. van Nielen for the
antioxidant activity assays and at Unilever Bestfoods, Rot-
Received for review February 6, 2002.
OP0255175
568
•
Vol. 6, No. 4, 2002 / Organic Process Research & Development