Content uploaded by Philippe Jeandet
Author content
All content in this area was uploaded by Philippe Jeandet on Oct 13, 2015
Content may be subject to copyright.
308 — Marchal et al.
Am. J. Enol. Vitic. 53:4 (2002)
1Laboratoire d’œnologie, URVVC-UPRES EA 2069, Faculté des Sciences, Université de Reims,
BP 1039, 51687 Reims 2, France ; 2Société Chamtor, 20, route de Pomacle, 51110 Bazancourt,
France ; 3Institut Œnologique de Champagne, Z.I de Mardeuil, BP 25, 51201 Epernay, France.
*Corresponding author [Tel + fax: 33 (0) 326 91 3340; email: richard.marchal@univ-reims.fr]
Acknowledgments: The authors are grateful to Europol’Agro and the Association Recherche
Oenologie Champagne et Université for financial support.
The authors express gratitude to Chantal Radet for expert secretarial assistance and Guillaume
Venel for technical assistance.
Manuscript submitted January 2001; revised May 2002
Copyright © 2002 by the American Society for Enology and Viticulture. All rights reserved.
308
After alcoholic fermentation, wine becomes a colloidal so-
lution and suspension. Particle density, which is close to that
of the wine, electrical repulsion forces, and diffusion phenom-
ena lead to slow and insufficient spontaneous clarification.
Moreover, natural sedimentation, centrifugation, and clarifica-
tion with continuous alluviation filtration do not protect wines
against colloidal haze.
For these reasons, organic fining agents [2,5,11,14] and min-
eral fining agents [6,10,12] are still commonly used to clarify
and stabilize white wines, thus avoiding natural haze. Bento-
nite treatments are efficient for protein removal [6,12] and can
reduce browning [9]. However, this mineral fining agent is also
responsible for loss of wine aroma [8,13]. The influence of
treatments with tannins complexed with gelatin on wine com-
position and sensory perception has been largely reported in
scientific studies. Such studies have often focused on the fac-
tors contributing to protein-polyphenolic interactions, which are
responsible for the expected flocculation and clarification
[15,16]. However, fining procedures using tannins associated
with gelatin are still empirical. The observed failures can be
explained by insufficient knowledge of tannins and the gelatin
currently used [16] and the flocculation mechanisms involved.
Because of these difficulties, building a model has been impos-
sible and small-scale fining tests are systematically necessary.
The incidence of bovine spongiform encephalopathy has led
to concerns about the use of proteins derived from animal
sources in the food supply, and winemakers have been encour-
aged to discontinue use of bovine gelatin. In Europe, the con-
cern of transmitting this disease to humans has led to a ban on
the use of bovine plasma and blood cells, commonly but incor-
rectly called blood albumin and glue (regulation CE 2087/97,
European Council, 20 Oct 1997). Some winemakers have also
been hesitant to use caseins because of their bovine origin.
Given the above, it was important to develop treatments that
could replace gelatin and bentonite finings. We thus started our
investigation of wheat prolamins, commonly called gluten, as
clarifying agents of white musts and wines. Different experi-
mental procedures were established to compare gluten effi-
ciency with usual fining agents. We have also developed a reli-
Use of Wheat Gluten as Clarifying Agent of
Musts and White Wines
Richard Marchal,1* Laurence Marchal-Delahaut,1 Franck Michels,2
Maryline Parmentier,1 Armelle Lallement,3 and Philippe Jeandet1
The bovine spongiform encephalopathy crisis has led some winemakers to question gelatin as a fining agent
and to reject the use of animal proteins. Gluten was evaluated as a substitute for gelatin by comparing gluten
treatments to other fining agents currently used (casein, association gelatin-tannin, fish glue, bentonite).
The turbidity of a Chardonnay must treated by gluten (20 and 40 g/hL) was approximately 70% less than
that of the control. A gluten with high hydrolysis of prolamins gave poor flocculation. Better results were
obtained with partially hydrolyzed and deamidated wheat proteins and vital gluten. Gluten at 20 g/hL and
the mixed tannin-gelatin at 5 g/hL had similar clarifications. Must treated with bentonites at 60 g/hL had 50
to 60% lower turbidity than untreated must. Compared with gluten only, deamidated gluten associated with
tannin had poor clarifying efficiency. Fining of Chardonnay wine showed efficient clarification with gluten at
20 and 40 g/hL depending upon the gluten type. Wine fining with gluten was similar to fining with tannin-
gelatin and more efficient than bentonite treatment. However, lower turbidities were obtained with fish glue
at 1 g/hL and casein at 5 g/hL. The volume of lees generated by fining with gluten ranged between 0.2 and
0.4% (v/v), similar to the values obtained with casein, fish glue, and tannin-gelatin and much lower than the
value obtained with bentonites.
Key words:
Fining, clarifying, gluten, gelatin, bentonite, must, wine
Abbreviations:
B1, B2: bentonites. T2: tannins from chestnut tree. T1: tannins from tara. FG1: fish glue with
medium degree hydrolyzed collagens. FG2: fish glue with low level of hydrolysis. G1, G2: gelatins from porcine
hydrolyzed collagen. Si: silica gel (Si O2). Ca: milk proteins (caseins). Ca 10: casein used at a dose of 10
g/hL. G1 +T2 5/5: gelatin 1 (5 g/hL) combined with tannin 2 (5 g/hL).
Must and Wine Clarification Using Wheat Proteins — 309
Am. J. Enol. Vitic. 53:4 (2002)
able system to evaluate the lees heights generated by fining with
animal or vegetal proteins and bentonite.
Material and Methods
Must and wine. Chardonnay must and base wine used in
this study were from the Cooperative Vinicole de Nogent
l’Abbesse (France) and the Maison de Champagne Duval-Leroy
(Vertus, France). Grapes were hand-harvested in Sept 1999 and
pressed with an industrial 8,000-kg press (pressure between 1.5
and 2 bar). Sulfur dioxide (60 mg/L) was added to free-run juice.
The wine was racked 4 weeks after alcoholic fermentation in
stainless steel (temperature controlled at 18 to 20°C). Malolac-
tic fermentation had not started. Must and wine analyses are
reported in Table 1. Sugar content was determined using a
Dujardin-Salleron (Paris, France) 1060-1090 Mustimeter, with
a correction of the value to 20°C. Alcohol content was deter-
mined after distillation with a Dujardin-Salleron class II
alcohometer 9 to 16% volume (accuracy 1/10 % v/v). The pH
was determined by an Orion 420A pH-meter (Thermo Electron
Corporation, Waltham, MA). Total acidities were determined
by M/64 NaOH additions using blue bromothymol as a colori-
metric indicator. Wine malic acid content was quantified by
automatic enzymatic method (Intégral Plus apparatus, Alliance
Instrument, Mery sur Oise, France).
Enological products. Glutens (marked gluten 1 to gluten
6) were supplied by French companies, except for gluten 3,
which came from Australia. Producers guaranteed that the glu-
tens were not from genetically modified organisms. Except for
gluten 1, a vital gluten (Roquette, Lestrem, France), all other
glutens were obtained from gluten protein enzymatic hydroly-
sis or chemical deamidation. Glutens 4 and 5 were obtained from
vital gluten enzymatic hydrolysis (Chamtor, Bazancourt,
France), with protein pI (isoelectric point) between 6 and 8.
Gluten 4 and gluten 5 had the same molecular composition;
gluten 5 was prepared as an aqueous solution, and gluten 4 was
suspended in 2 g/L tartaric acid aqueous solution. Gluten 2 was
also an enzymatic hydrolyzed gluten (ARD, Pomacle, France),
with a high level of hydrolysis. Gluten 3 (Manildra Group,
Bomaderry, NSW, Australia) and gluten 6 (Chamtor,
Bazancourt, France) were obtained from vital gluten
deamidation. Gluten 6 had protein pI between 3.5 and 4.5. Glu-
tens 3 and 6 gave complete suspensions without particle sedi-
mentation.
All other enological products were supplied by Institut
Oenologique de Champagne (Epernay, France). Bentonites (B1
and B2) were two sodic activated bentonites. Tannins from
chestnut tree (T2), a brown powder with tannin content of 76%,
were partially water soluble. The yellow powder of tara tannin
(T1), which was totally water soluble, contained 90% tannin.
The food casein (Ca) contained less than 2% fat and had a pro-
tein content that was 92% of its dry weight. The two fish glues
(FG1 and FG2) were collagens exclusively used for wine fin-
ing. FG1 had a medium level of hydrolysis, and FG2 presented
a low level of hydrolysis. Fish glue proteins were suspended in
2 g/L tartaric acid aqueous solution to increase hydration and
solubility. Two gelatins (G1 and G2), from porcine hydrolyzed
collagen and supplied as aqueous solutions, were tested for must
and wine finings. The silica gel (Si) was used as a commercial
solution (30% p/v Si O2). Sulfur dioxide (SO2) was added at a
concentration of 1 g/L to all solutions, which were kept at 4°C.
Solutions were prepared 12 hours before must and wine fining.
For all fining agent solutions, concentration was 10 g/L.
Fining experiments. Fining tests were carried out in plas-
tic graduated cylinders (vol 500 mL and i.d. 52 mm), filled to
260 mm with must or wine. For the Chardonnay must, finings
were tested directly after pressing. For the Chardonnay base
wine, experiments were carried out 5 weeks after the end of
alcoholic fermentation. Enological products were introduced
with automatic BioHit pipettors (Bonelles, France) and sulfur
dioxide was added to musts (100 mg/L) and wines (60 mg/L)
to avoid an eventual alcoholic fermentation and vinegary taint
in wine (acetobacter fermentation), respectively. Graduated
cylinders were rotated twice to homogenize fining agents and
must or wine. (In Results and Discussion, the type of agent used
for fining and the quantity of fining agent added are abbrevi-
ated as follows: Ca 10 indicates that casein was used at a dose
of 10 g/hL; G1 +T2 5/5 indicates that gelatin 1 was used at a
dose of 5 g/hL combined with tannin 2 at a dose of 5 g/hL.) As
gluten clarifying activity was unknown, doses ranging from 5
to 40 g/hL of gluten were tested as they represented doses com-
monly used for other fining agents (casein, gelatin, and bento-
nite).
Clarification kinetics. A 15-mL sample was taken from the
graduated cylinder with a 25-mL plastic pipette. As liquid taken
from the top is clearer than that from the bottom, it does not
reflect the real turbidity. To avoid this problem, the pipette,
locked on top with a finger, was introduced into 200 mL of liq-
uid and then unlocked to allow the sample to fill it. Kinetics
were carried out at room temperature (20 ± 2°C). For the
Chardonnay must, turbidities were measured 18, 26, and 43 hr
after the addition of fining agents. For the Chardonnay wine,
turbidities were measured 28 and 48 hr after the addition of fin-
ing agents. Turbidities were measured with a Hach 2100AN
Turbidimeter calibrated with the Gelex® Secondary Standards
Kit (Hach Company, Loveland, CO) and expressed
in Nephelos turbidity units (NTU). A preliminary
study was conducted with a control wine to deter-
mine standard deviations. Three measures were made
per day (in three graduated cylinders), for 5 days.
Standard deviations were calculated for two and three
measures (Table 2). As values were always low, it
was decided to perform only one assay for the clari-
fying experiments.
Table 1 Analytical characteristics of Chardonnay must and wine before fining.
Must density (g/L) Sugar Total Malic
wine alcohol content acidity acid Turbidity
content (% v/v) (g/L) g/L H2SO4pH (g/L) (NTU)
Must 1068 157 6.1 3.13 —222
Still wine 11.2 —5.9 3.09 3.6 107
310 — Marchal et al.
Am. J. Enol. Vitic. 53:4 (2002)
Determination of lees volume after fining. As graduated
cylinders, with a 52 mm i.d., do not allow a precise measure-
ment of lees height, white glass bottles from Champagne were
used for fining experiments. Bottles were filled with wine and
a fining agent, closed with a cap, and suspended with the neck
down. Bottles were kept at room temperature and protected from
draft and light. After 20 days (time required for a total sponta-
neous clarification of untreated wine), bottles were rotated twice
round the vertical axis (Champagne riddling) to allow the floc-
culates to slide from the bulge into the bottle neck. Lees were
gathered on a small surface, permitting ease in measurement.
Lees height was measured with a ruler adapted to the bottle neck,
converted into mL and finally into percentage lees/liquid (v/v).
Preliminary calibration was carried out with four bottles closed
with caps.
Characterization of fining agents by SDS-PAGE. Mo-
lecular weight (MW) range of enological proteins (caseins, gela-
tins, glutens, and fish glues) was determined by sodium dodecyl
sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) ac-
cording to the Laemmli method [7] using slab gels (0.75 mm).
A Mini-protean vertical electrophoresis apparatus (Bio-Rad,
Hercules, CA) was used to run the gel at a constant setting of
150 V until the bromophenol blue tracker dye reached the bot-
tom of the gel (usually 65 min at room temperature). Standard
proteins from 14,000 to 94,000 were used as MW markers
(LMW Pharmacia, Uppsala, Sweden). Fining agents and stan-
dard proteins were treated as above with Laemmli buffer (v/v),
and 20-µL samples were loaded into each well. Molecular
weights of protein bands were calculated from the linear regres-
sion equation of log MW versus mobility. After electrophore-
sis, separated proteins were stained with 1.5% Coomassie Bril-
liant Blue in 50% (v/v) methanol and destained in acetic acid/
methanol/water (1:2:7).
Results and Discussion
Electrophoretic characteristics of fining agents. The mo-
lecular weight range of proteins from glutens and animal pro-
teins can be observed in Figures 1 and 2, respectively. The elec-
trophoretic profile of gluten 2 (lane 5) shows a significant deg-
radation of prolamins with a total disappearance of molecules
above 13,000 [1,3,4]. In contrast, gluten 1 (lane 6) has the clas-
sical profile of a vital gluten. Prolamins have a MW between
30,000 and 45,000 and at 16,000. Vital gluten also
contains numerous minor proteins between 10,000
and 100,000. Although gluten 5 (lane 3) has been
partially hydrolyzed, it is possible to observe a few
distinctive bands. Gluten 3 (lane 7) and gluten 6 (lane
2), two deamidated products, present a protein deg-
radation much higher than gluten 5 (lane 3) (Figure
1). They presented a discontinuous smear between
10,000 and >100,000 (glutens 5 and 4 are the same
product but with different preparations before use).
Milk “casein” (Figure 2, lane 7) consists of sev-
eral caseins; most have a MW close to 30,000, a few
between 10,000 and 23,000, and a few minor pro-
teins between 50,000 and 80,000. Fish glue FG1 (lane
6) presents a limited degradation without any col-
lagenic structure. Proteins appear as distinctive bands, most situ-
ated between 17,000 and 80,000. Unlike FG1, fish glue FG2
presents numerous nonhydrolyzed collagen molecules close to
110,000 to 120,000 and to 220,000 (double helix structure).
Some smears of gelatins were observed (G1 in lane 2 and G2
in lane 3), confirming that these products were obtained from a
Figure 1 SDS-PAGE analysis of wheat glutens used for Chardonnay
must and wine finings. Proteins stained with Coomassie Brilliant Blue.
Lanes 1 and 4, MW markers; lane 2, deamidated gluten (Chamtor); lane
3, hydrolyzed gluten (Chamtor); lane 5, hydrolyzed gluten (ARD); lane
6, native gluten (Roquette); lane 7, deamidated gluten (Manildra). Rela-
tive MW (x 10-3) of standard proteins given on left side of gel.
1 2 3 4 5 6 7
94
67
20
14
MW (10
-3
)-
+
30
45
Figure 2 SDS-PAGE analysis of enological proteins used for Chardonnay
must and wine finings. Proteins stained with Coomassie Brilliant Blue.
Lanes 1 and 4, MW markers; lanes 2 and 3, gelatins G1 and G2; lane 5,
FG2; lane 6, FG1; lane 7, cow caseins. Relative MW (x 10-3) of standard
proteins given on left side of gel.
MW (10
-3
)1 2 3 4 5 6 7
94
67
20
30
14
45
-
+
Table 2 Determination of turbidity standard deviations for static clarifying
kinetics in a control wine (no fining).
NTU after NTU after NTU after NTU after NTU after
22 hr 51 hr 78 hr 96 hr 115 hr
Cylinder 1 79.1 44.1 27.1 22.1 16.7
Cylinder 2 79.3 43.2 26.1 21.0 16.2
Cylinder 3 77.6 44.4 27.3 21.5 15.8
Avg value and
SDa for n = 2 79.2 ± 0.1 43.7 ± 0.45 26.6 ± 0.5 21.5 ± 0.55 16.4 ± 0.25
Avg value and
SD for n = 3 78.7 ± 0.54 43.9 ± 0.36 26.8 ± 0.36 21.5 ± 0.32 16.2 ± 0.26
aSD: standard deviation
Must and Wine Clarification Using Wheat Proteins — 311
Am. J. Enol. Vitic. 53:4 (2002)
high hydrolysis of porcine collagen. The continuity in the MW
distribution seems relatively similar for G1 and G2. Although
their analyses in SDS-PAGE do not present any noticeable dif-
ferences, results show variations in their efficiencies.
These enological products show a complex proteic compo-
sition with large differences between animal and vegetable pro-
teins. It is also clear that as a consequence of the industrial pro-
cess, glutens display large differences in proteic composition
and biochemical characteristics, such as isoelectric point, mo-
lecular weight, and hydrophobicity.
Comparison between gluten and other fining agents used
to clarify Chardonnay must. This test, carried out with a
Chardonnay must, showed that gluten clarifying efficiency
largely depends on the quantity used (Table 3). For doses be-
tween 10 and 40 g/hL, glutens 1 and 3 had similar activities,
despite different geographic origins (Australia and France) and
industrial processes (vital and deamidated glutens). The turbid-
ity obtained with gluten 2 at 40 g/hL represented 65% of the
nontreated must turbidity. This result was not very interesting
when compared with glutens 1 and 3, which gave a turbidity
equal to 30% of the control must. It seems that significant hy-
drolysis of prolamins can lead to difficult flocculation. In con-
trast, vital gluten, which is not hydrolyzed, rapidly forms a vis-
coelastic network that is impossible to use if prepared in ad-
vance. Despite its good fining effect, its enological use is lim-
ited, as the suspension has to be freshly prepared immediately
before addition into must or wine.
It is not apparent why one gluten is better than another. Glu-
tens are composed of numerous gliadins [1] and glutenins [4].
Gliadins vary in amino acid composition [3], the consequence
being a difference of wheat protein hydrophobicity [1] and pI
values [1]. These biochemical characteristics are responsible for
protein-phenolic interactions, leading to flocculation and clari-
fication. Moreover, must is a complex biochemical medium
containing colloidal particles and soluble colloids. To explain
the differences in clarifying efficiency, it would be necessary
to work with synthetic must that has a controlled composition.
It would also be necessary to experiment with proteic fractions
isolated from whole gluten.
Results obtained with gelatin 2 (associated with tannins T1
or T2) were not satisfying, regardless of the dose used (Table
3). Independent of the associated tannins, the clarifying effects
were systematically better with gelatin 1 than gelatin 2. Finings
with tannins associated with gelatin 1 (2 x 5 g/hL) were similar
to those obtained with glutens 1 and 3 at 20 g/hL. Milk casein
gave the best results with clear musts (Table 3). Results obtained
with the two bentonites were similar. With 15 g/hL, the turbid-
ity decreased by nearly 25%; however, with 60 g/hL, it still rep-
resented between 50 and 60% of the control must. For gluten
and casein, an increase of efficiency was noted when product
doses increased. In contrast, high doses of bentonite did not give
good clarification. Furthermore, a highly negative interaction
between bentonite and casein was observed (Table 3).
Combinations of gluten-adjuvants for Chardonnay must
fining. We also examined the effect of combinations of gluten
3 (deamidated) with tara tannins, chestnut tree tannins, or silica
gel on must clarification (Table 3). We expected a synergy be-
tween the two products, as was the case for gelatin and tannins.
We also experimented with low doses (1 and 2 g/hL). The tur-
bidities obtained with combinations of gluten + tannins were
higher than with gluten 3 alone. In several cases, the control
must was even clearer than the samples treated with these two
agents. Results also showed that chestnut tree tannins or tara
tannins used alone were not able to clarify this Chardonnay must
(Table 3). The affinity between deamidated gluten proteins and
tannins (whatever the tannins and the ratio tannins/gluten used)
was too low. The density of the flocculates was close to that of
the must and did not allow good sedimentation. Consequently,
these associations are not advisable.
The combined deamidated gluten-silica gel also showed a
negative interaction. Although gluten 3 at 20 g/hL and silica
gel at 2 g/hL decreased the turbidity by 64 and 47 NTU, re-
spectively, the combination of the two led to a decrease of only
72 NTU. As a result, there is no additional effect in their clari-
fying activity.
Influence of settling duration on clarifying efficiency.
Turbidities of Chardonnay must were also evaluated after 26
and 43 hr settling. For each treatment, turbidities obtained were
Table 3 Effect of different fining treatments on turbidity of Chardonnay
must. Deamidated gluten experimented in combination with cofining
agents. Measurements made 18 hr posttreatment.
Treatment NTU Treatment NTU
Control must 121 Ca 20 45
Gluten 1 5 108 Ca + B1 20/20 68
Gluten 1 10 85 Ca 40 23
Gluten 1 20 56 Ca + B1 40/40 59
Gluten 1 40 35 B1 15 94
Gluten 2 5 117 B1 30 86
Gluten 2 10 110 B1 60 59
Gluten 2 20 94 B2 15 91
Gluten 2 40 77 B2 30 88
Gluten 3 5 121 B2 60 70
Gluten 3 10 90
Gluten 3 20 58 G1+T1 2.5/5 85
Gluten 3 40 37 G1+T1 5/5 71
Gluten 3+T1 1/1 128 G1+T1 5/2.5 62
Gluten 3+T1 2/2 133 G1+T2 2.5/5 72
Gluten 3+T1 5/5 142 G1+T2 5/5 55
Gluten 3+T1 10/10 137 G1+T2 5/2.5 50
Gluten 3+T2 1/1 126 G2+T1 2.5/5 115
Gluten 3+T2 2/2 125 G2+T1 5/5 108
Gluten 3+T2 5/5 111 G2+T1 5/2.5 99
Gluten 3+T2 10/10 88 G2+T2 2.5/5 96
Gluten 3+T1 1/0.5 126 G2+T2 5/5 83
Gluten 3+T1 2/1 130 G2+T2 5/2.5 89
Gluten 3+T1 5/2.5 130
Gluten 3+T1 10/5 108 T1 2 120
Gluten 3+T1 20/10 102 T1 5 124
Gluten 3+T2 1/0.5 124 T2 2 115
Gluten 3+T2 2/1 127 T2 5 113
Gluten 3+T2 5/2.5 112 Si 2 73
Gluten 3+T2 10/5 91
Gluten 3+T2 20/10 65
Gluten 3 + Si 20/2 51
312 — Marchal et al.
Am. J. Enol. Vitic. 53:4 (2002)
lower in these conditions than after 18 hr settling (and also lower
after 43 hr than after 26 hr). The turbidity of the control must
was 121 NTU after 18 hr, 89 NTU after 26 hr, and 51 NTU
after 43 hr settling. To compare the different treatments more
efficiently, the untreated must turbidity was set at 100. The glu-
ten 3 efficiency at 10 and 20 g/hL appeared to be higher after
26 hr than after 18 hr (Table 4). We also obtained better results
with the combinations of tannin-gelatin 2 after 26 hr. On the
contrary, the efficiency after a treatment by bentonites was lower
after 26 hr than after 18 hr, except for bentonite 2, for which
the best activity was observed with the smallest dose after 43
hr settling. This indicates that bentonites presented a real inter-
est for fining when a rapid racking (of a few hours) was re-
quired. In contrast, treatments with tannin-gelatin or gluten re-
quired a longer settling time to give optimal results.
Flocculation rapidly appeared with bentonites, and the lag-
phase was much longer with organic fining agents. After 43 hr
settling, gluten 3 at 10, 20, and 40 g/hL resulted in a low effi-
ciency, considerably lower than that obtained after 26 and even
18 hr (Table 4), even if turbidities were lower. This can be ex-
plained by the fact that the results are compared to the control
must, which has a low turbidity. The efficiency of gluten 1 de-
creased only after 43 hr settling. Gelatin 1 gave better results
than gelatin 2, but its relative efficiency was lower after long
settling (43 hr). This phenomenon was less noticeable for gela-
tin 2.
These results show that fining treatments have to be applied
when the must turnover is rapid, which is the case in most press-
ing centers (generally between 12 and 36 hr). Fining treatments
seem less necessary for long settlings because of the low tur-
bidity naturally reached by the control must. Given the risk of
a spontaneous start of alcoholic fermentation, long settlings are
rarely made in actual practice, particularly for musts from late
harvest. This study also shows that these practices can gener-
ate great differences in the results. Glutens led to a high level
of clarification, which was similar to the casein efficiency.
However, casein cannot always be used, for it is reputed to give
meager and stripped wines. In some wine-growing regions such
as Champagne, its use is limited to oxidized must when grapes
are highly infected by Botrytis cinerea, the causal agent for gray
mold. Gluten as a vegetable protein agent gave better clarifica-
tion than bentonites. It can be used without any adjuvant and
as easily as bentonite and gelatin, when it is partially hydro-
lyzed. However, further experiments need to be carried out with
musts from different grape varieties and different wine-grow-
ing regions to verify and generalize these first conclusions.
Comparison of wheat gluten and other fining agents for
clarifying Chardonnay wine. Preliminary tests showed that
glutens 1 and 2 did not clarify the wine. In the case of gluten 1,
this is certainly related to the low dispersion of the viscoelastic
particles, leading to a small contact surface. Gluten 2 is com-
posed of prolamins with a low molecular weight, which do not
lead to the formation of flocculates. For each of the four glu-
tens tested (gluten 3 to gluten 6), an addition of 5 g/hL led to
an increase of the wine turbidity (Table 5). Yet the clarifying
effect (from 10 g/hL) increased proportionally to the gluten
quantity added to the wine samples. Also, differences between
glutens were observed. Gluten 3 at 20 g/hL was nearly as effi-
cient as gluten 4 at 40 g/hL. These variations can be explained
by the different treatments applied to vital gluten and to the
protein composition, which depends on the wheat species. As
for the must, the enology literature does not provide precise
explanations for the observed differences. This will be investi-
gated in further studies with synthetic wines and gluten proteic
fractions.
As observed for must, the combination of gelatin 1 + tannin
2 gave the best efficiency (Table 5). Differences between T1
and T2 are nevertheless more marked than for the must clarify-
ing. Turbidities obtained with gluten at 20 and 40 g/hL were
between the minimum and maximum values obtained with the
different associations of tannin-gelatin. The clarifying effect
obtained with bentonites was greater with 60 g/hL than with 30
g/hL, but the increase was minimal. As for must, wine fining
with glutens (especially gluten 3 and gluten 4 at 40 g/hL) re-
sulted in better clarification than with the two bentonites tested.
However the best results were obtained with 1 g/hL fish glue
(particularly when in combination with tannins) and with 10
and 20 g/hL casein (Table 5).
Combination of gluten-other fining agent for
Chardonnay wine clarification. Clarifications obtained with
the combination gluten 4-tannin 2 were always higher quality
than with gluten 4 only. For example, the turbidity obtained with
gluten 4 at 20 g/hL and tannin 2 at 4 g/hL was 60% lower than
with gluten 4 only. Contrary to the behavior in must, we also
observed that flocculation/clarifying appeared when tannins
Table 4 Clarifying of Chardonnay must: relative efficiency as a
function of settling time. Measurements made 18, 26, and 43 hr
posttreatment. All values transformed so that control must has
an arbitrary value of 100.
Turbidity (arbitrary units)
Treatment 18 hr 26 hr 43 hr
Control must 100 100 100
Gluten 1 10 70 69 78
Gluten 1 20 46 49 66
Gluten 1 40 29 32 42
Gluten 3 10 74 53 96
Gluten 3 20 48 35 62
Gluten 3 40 31 33 48
G1 + T1 5/5 59 51 64
G1 + T1 5/2.5 51 52 66
G1 + T2 5/5 45 49 58
G1 + T2 5/2.5 41 48 62
G2 + T1 5/5 89 77 84
G2 + T1 5/2.5 82 67 84
G2 + T2 5/5 69 65 72
G2 + T2 5/2.5 74 70 78
B 1 15 78 90 68
B 1 30 71 91 104
B 1 60 49 64 102
B 2 15 75 58 44
B 2 30 73 84 70
B 2 60 58 76 96
Must and Wine Clarification Using Wheat Proteins — 313
Am. J. Enol. Vitic. 53:4 (2002)
Table 5 Effect of different fining treatments on turbidity of Chardonnay
wine. Measurements made 28 hr posttreatment.
Treatment NTU Treatment NTU
Control wine 50 T2 2 41
T2 6 22
Gluten3 + T2 5/1 39 FG1 1 12
Gluten3 + T2 10/2 32 FG1 3 22
Gluten3 + T2 20/4 24 FG1 + T2 1/2 4
Gluten3 + T2 40/8 38 FG1 + T2 3/6 5
Gluten4 + T2 5/1 36 FG2 1 12
Gluten4 + T2 10/2 24 FG2 3 13
Gluten4 + T2 20/4 14 FG2 + T2 1/2 5
Gluten4 + T2 40/8 11 FG2 + T2 3/6 4
Gluten6 + T2 5/1 42 Ca 5 15
Gluten6 + T2 10/2 35 Ca 10 6
Gluten6 + T2 20/4 31 Ca 20 5
Gluten6 + T2 40/8 40 Ca + Gluten3 5/5 12
Ca + Gluten3 10/10 9
Gluten3 5 56 Ca + Gluten3 20/20 6
Gluten3 10 27 Ca + B1 10/10 9
Gluten3 20 15 G1+T1 5/5 22
Gluten3 40 10 G1+T1 5/2.5 23
Gluten4 5 55 G1+T2 5/5 6
Gluten4 10 47 G1+T2 5/2.5 9
Gluten4 20 32 G2+T1 5/5 26
Gluten4 40 13 G2+T1 5/2.5 28
Gluten5 5 52 G2+T2 5/5 15
Gluten5 10 44 G2+T2 5/2.5 11
Gluten5 20 35 B1 15 46
Gluten5 40 19 B1 30 24
Gluten6 5 53 B1 60 20
Gluten6 10 40 B2 15 29
Gluten6 20 28 B2 30 23
Gluten6 40 18 B2 60 16
were added in the wine (-56% for T2 at 6 g/hL, compared to the
control wine). In this experiment, there was synergy between
hydrolyzed gluten proteins and tannins. However, when the dose
of gluten 4 increased (40 g/hL), the synergy between wheat
proteins and tannins was less marked.
The combination of gluten 3 or gluten 6 with tannins gave
better results only for the dose of 5 g/hL, but the increase rep-
resented only 20 or 25% when compared to the untreated wine.
With higher doses, the clarification was better with gluten 3 and
6 alone than with the combinations of gluten-tannins (except
for gluten 6 at 10 g/hL and tannin 2 at 2 g/hL, but this treat-
ment gave poor turbidity decrease and cannot be applied in the
wine industry).
The combination of gluten-casein generates negative inter-
actions. When compared to the untreated wine, gluten 3 alone
or casein alone, both at 10 g/hL, led to a decrease of 24 and 44
NTU, respectively. But the combination of both in the same
conditions made the turbidity fall by only 41 NTU (Table 5).
Therefore, neither the combination of the two nor the combi-
nation of caseins + bentonite is advisable. These results show
that gluten is an efficient fining agent for clarifying white wines
after alcoholic fermentation.
Influence of kinetics on Chardonnay wine fining. Differ-
ences between the efficiencies at 28 and 48 hr were minimal or
nonexistent for numerous treatments (Table 6). Gluten 3 and
gluten 4 efficiency was significantly better at 48 hr, whatever
the dose. This was also the case for other protein finings such
as caseins, caseins + gluten, and tannins + gelatin treatments.
Even if the differences observed represent only a few turbidity
units, it points out a behavior contrary to mineral fining using
bentonite. Bentonite 1 did not present any variation, but bento-
nite 2 started its clarification process more rapidly. However,
gluten 3, gluten 4, and the four tannin-gelatin combinations
exhibited a longer delay before clarification started; this was
probably due to the time needed for the formation of floccu-
lates. For gluten 5, efficiency was better after 28 hr for 10 and
40 g/hL, and better after 48 hr for 20 g/hL. These results seem
to indicate that particle sedimentation does not follow a regu-
lar and simple mathematical model.
Table 6 Clarification of Chardonnay wine: relative efficiency as
a function of settling time. Measurements made 28 and 48 hr
posttreatment. All values transformed so that control must
has an arbitrary value of 100.
Treatment 28 hr 48 hr
Control wine 100 100
Gluten3 10 54 51
Gluten3 20 30 29
Gluten3 40 20 17
Gluten4 10 94 89
Gluten4 20 64 49
Gluten4 40 26 20
Gluten5 10 88 91
Gluten5 20 70 63
Gluten5 40 38 43
Gluten6 10 80 83
Gluten6 20 56 51
Gluten6 40 36 37
Gluten4 + T2 10/2 48 46
Gluten4 + T2 20/4 28 34
Gluten4 + T2 40/8 22 17
Gluten3 + Ca 5/5 24 20
Gluten3 + Ca 10/10 18 14
Gluten3 + Ca 20/20 12 11
FG1 1 24 29
FG1 3 44 37
FG2 1 24 20
FG2 3 26 31
Ca 5 30 29
Ca 10 12 11
Ca 20 10 9
G1+T1 5/5 44 37
G1+T2 5/5 12 11
G2+T1 5/5 52 43
G2+T2 5/5 30 20
B1 15 92 91
B1 30 48 49
B1 60 40 40
B2 15 58 60
B2 30 46 51
B2 60 32 34
314 — Marchal et al.
Am. J. Enol. Vitic. 53:4 (2002)
Variations observed in fining tests were smaller with wine
than with the must (Table 4). It is important to note that
winemakers use fining to reach as low as possible wine turbid-
ity, which is not always necessary for must. It seems that wine-
fining efficiency varies little during clarification, making it
possible to accomplish small-scale fining tests with short reac-
tion times.
Influence of the fining type on lees volume. This param-
eter has to be considered for the fining agent choice because it
is related to the loss of wine. The untreated wine presented the
lowest volume of lees (Table 7). Apart from gluten 3 at 40 g/
hL, which gave a lees volume of 0.6% (v/v), all glutens gener-
ated volumes of lees between 0.2 and 0.4%, which were simi-
lar to the values observed for casein, tannin-gelatin, and fish
glue. However, these volumes are less than those obtained with
bentonite-casein or with bentonites at 30 g/hL, which can reach
up to 2% (Table 7). Globally, glutens generate volumes of lees
comparable to those obtained with present animal-protein fin-
ing agents.
Conclusion
Gluten proteins at a concentration between 20 and 40 g/hL
produced good clarification of treated Chardonnay must when
compared to the control must with bentonite or tannin-gelatin
combinations. Gluten treatments only slightly differ from those
using cow caseins. Similar results were obtained for wine fin-
ing. The combination of fish glue (1 g/hL) + tannins (2 g/hL)
and casein (at 10 and 20 g/hL) were still the most efficient fin-
ing agents in terms of wine clarification. The combination of
Table 7 Clarification of Chardonnay wine. Volume of lees (% v/v)
obtained for different fining treatments using animal, vegetable, and
mineral agents.
Treatment Lees (v/v)
Control wine 0.09%
Gluten 6 20 0.13%
G1 + T2 5/2.5 0.16%
Gluten 6 40 0.16%
FG1 + T2 0.5/1 0.26%
Gluten 4 + T2 20/2 0.26%
Gluten 5 20 0.26%
FG1 0.5 0.26%
Gluten 4 20 0.28%
Gluten 5 40 0.31%
Ca 10 0.31%
Gluten 3 20 0.31%
Gluten 4 40 0.36%
Gluten 3 40 0.60%
Ca + B1 10/10 0.99%
B2 30 1.71%
B1 30 2.00%
hydrolyzed gluten and tannins improved the efficiency of this
fining treatment. A further study will focus on the influence of
fining on treated wine composition and sensory perception.
Literature Cited
1. Brown, J.W.S., and R.B. Flavell. Fractionation of wheat gliadin and
glutenin subunits by two dimensional electrophoresis and the role of
group 6 and group 2 chromosomes in gliadin synthesis. Theor. Appl.
Genet. 59:349-359 (1981).
2. Calderon, P., J.P. Van Buren, and W.B. Robinson. Factors influencing
the formation of precipitates and hazes by gelatin and condensed
hydrolysable tannins. J. Agric. Food Chem. 16:479-482 (1968).
3. Charbonnier, L., T. Terce-Laforgue, and J. Mosse. Some physico-
chemical properties of
Triticum vulgare
β, γ and ω-gliadins. Ann. Technol.
Agric. 29:175-190 (1980).
4. Gupta, R.B., and K.W. Shepherd. Two-step one-dimensional SDS-
PAGE analysis of LGM subunits of glutenin. I. Variation and genetic
control of the subunits in hexaploid wheats Theor. Appl. Genet. 80:65-
74 (1990).
5. Hrazdina, G., J.P. Van Buren, and W.B. Robinson. Influence of
molecular size of gelatin on reaction with tannic acid. Am. J. Enol. Vitic.
20:66-68 (1969).
6. Hsu, J.C., and D.A. Heatherbell. Heat-unstable proteins in wine. I.
Characterization and removal by bentonite fining and heat treatment.
Am. J. Enol. Vitic. 38:11-16 (1987).
7. Laemmli, U.K. Cleavage of structural protein during the assembly of
the head of Bacteriophage T4. Nature 227:680-685 (1970).
8. Lubbers, S., C. Charpentier, and M. Feuillat. Etude de la rétention
des composés d’arôme par les bentonites en moût, vin et milieux
modèles (Study of the binding of aroma compounds by bentonite in
must, wine and model systems). Vitis 35:59-62 (1996).
9. Main, G.L., and J.R. Morris. Color of Seyval blanc juice and wine as
affected by juice fining and bentonite fining during fermentation. Am. J.
Enol. Vitic. 45:417-422 (1994).
10. Marchal, R., J. Barret, and A. Maujean. Relation entre les
caractéristiques physico-chimiques d’une bentonite et son pouvoir
d’adsorption (Relation between physicochemical characteristics of a
bentonite and its adsorptive capacity). J. Int. Sci. Vigne Vin 29:27-42
(1995).
11. Marchal, R., C. Sinet, and A. Maujean. Etude des gélatines œnolo-
giques et du collage des vins de base champenois. Bull. OIV 751-
752:691-725 (1993).
12. Milisavljevic, D. Prévention des troubles protéiques du vin par
l’emploi des bentonites dans les moûts. Ann. Technol. Agric. 12:315-
330 (1963).
13. Miller, G.C., J.M. Amon, R.L. Gilson, and R.F. Simpson. Loss of
wine aroma attributed to protein stabilization by bentonite or ultra-
filtration. Aust. Grapegrower Winemaker 256:49-50 (1985).
14. Ribéreau-Gayon, J., and E. Peynaud. La clarification des vins. Rev.
Viticole 82:8-13 (1935).
15. Sarni-Manchado, P., A. Deleris, S. Avallone, V. Cheynier, and M.
Moutounet. Analysis and characterization of wine condensed tannins
precipitated by proteins used as fining agents in enology. Am. J. Enol.
Vitic. 50:81-86 (1999).
16. Waters, E.J., W. Wallace, and P.J. Williams. Identification of heat-
unstable wine proteins and their resistance to peptidases. J. Agric. Food
Chem. 40:1514-1519 (1992).
