Trends in Food Science & Technology 12 (2002) 230–239

Review

The wine proteins

Introduction
The past 40 years have seen the greatest change in the
Ricardo B. Ferreira,*,y,{ making, the quality, and the distribution of wine in its
8000-year history. This is not surprising. Which art, and
Maria A. Piçarra-Pereira,{,x which science, has not altered beyond recognition in the
Sara Monteiro,{ era of technology? Even though the application of
modern analytical techniques has permitted significant
Virgı́lio B. Loureiro{ advances in wine protein research, wine producers are
still facing a number of unsolved problems.
and Artur R. Teixeira{ Wines, like many other natural food products, con-
y tain varying amounts of different nitrogenous sub-
Instituto de Tecnologia Quı́mica e Biológica,
stances, the most important of which are proteins.
Universidade Nova de Lisboa, Apartado 127, 2781-901
These polymers do not contribute significantly to the
Oeiras, Portugal (tel: +351-2144-69651; fax: +351-
nutritive value of wines since their concentration varies
2144-33644; e-mail: ferreira@itqb.unl.pt)
{
typically from 15 to 230 mg per liter. However, they
Departamento de Botânica e Engenharia Biológica, assume a considerable technological and economical
Instituto Superior de Agronomia, Universidade importance because they greatly affect the clarity and
Técnica de Lisboa, 1349-017 Lisboa, Portugal stability of the wine. Limpidity (translucency) is of vital
x
Escola Superior Agrária de Castelo Branco, 6001 importance to wine quality because this property makes
Castelo Branco Codex, Portugal the first impression on the consumer, so it is necessary
that wines should stay stable and clear, regardless of the
conditions of storage.
Proteins are typically present in wines in low concentra- Instability of proteins in white wines is one of the
tions, contributing little to their nutritive value. However, most common non-microbial defects of commercial
they assume a considerable technological and economical wines (Bayly & Berg, 1967; Hsu & Heatherbell, 1987a;
importance because they greatly affect the clarity and sta- Waters, Wallace, & Williams, 1992). Coagulation of
bility of wines. Although exhibiting a large diversity, the proteins in white wines may result from unfavorable
majority of the wine proteins are structurally related and storage conditions, leading to their aggregation. The
have been identified as pathogenesis related (PR) proteins. denatured protein can subsequently precipitate to form
Thus, different wines are essentially composed by identical an amorphous sediment or deposit or flocculate and
sets of polypeptides. They derive from the grape pulp, and thus produce a suspended and unattractive haze in the
survive the vinification process simply because they are bottled wine that reduces its commercial value, making
highly resistant to proteolysis and to the low pH character- it unacceptable for sale. Apart from the fact that they
istic of wines. There is increasing evidence suggesting that are known to be important for clarity and stability, the
although protein-dependent, the development of turbidity role of peptides and proteins in wine qualities such as
in wines is controlled by a number of factors of non-protein flavor still remains obscure.
origin, such as polyphenols, the wine pH and the presence Naturally occurring proteins of grapes have been
of polysaccharides. A variety of procedures has been known for many years to cause clouding or haze in
developed and tested for the specific removal of proteins wines. Wine protein research and the influence of pro-
from wines. Even though bentonite fining is nonspecific and teins on wine stability was initiated in California, in
can impair the quality of wine, it remains the only effective 1896, by Colby. However, the first reports on the nature
method to stabilize wines. # 2001 Elsevier Science Ltd. All of the compounds responsible for wine turbidity were
rights reserved. made by Koch and Sajak (1959) and Moretti and Berg
(1965). Most of these early investigations on nitrogen-
ous substances were carried out by determining the total
* Corresponding author. nitrogen content by the Kjeldahl method. The protein
0924-2244/02/$ - see front matter # 2001 Elsevier Science Ltd. All rights reserved.
PII: S 0 92 4 - 2 24 4 (0 1 ) 0 0 08 0 - 2
 R.B. Ferreira et al. / Trends in Food Science & Technology 12 (2002) 230–239 231

content was subsequently estimated using an appro- among wines are primarily caused by factors which
priate conversion factor. The protein values obtained in influence the total nitrogen in grapes, such as climate,
this way are not representative and are overestimated soil and the variety itself (Bayly & Berg, 1967). ‘Terroir’
owing to the fact that grapes and wines contain a con- and biotic stresses influence not only the quantity but
siderable amount of non-protein nitrogenous substances also the quality of the proteins synthesized during grape
(Koch & Sajak). development and maturation (see below). Dorrestein,
The actual protein levels at which a wine may be Ferreira, Laureano, and Teixeira (1995) obtained vari-
considered stable are not known. Also, there is a sig- ety specific chromatograms when the protein profiles of
nificant amount of information in the literature, in some four different varietal wines were analysed by FPLC ion
cases reporting conflicting data, on the nature of the exchange chromatography. Similar results were repor-
proteins responsible for wine turbidity. Increasing evi- ted by other authors (see, for example, Dizy & Bisson,
dence is accumulating suggesting that factors of non- 1999; Pueyo, Dizy, & Polo, 1993). Other studies showed
protein origin may play a fundamental role in triggering that fruit maturity also relates to the irregular occur-
wine haze formation. rence of protein clouding in wines (Murphey, Spayd, &
Various methods of determining protein stability in Powers, 1989). Indeed, the soluble protein in juice and
wines have been reported. These include the heat test wine increases with increasing grape maturity (Murphey
(Berg & Akiyoshi, 1961), as well as the use of protein et al.). Koch and Sajak (1959) were unable to detect any
precipitants, such as tannin, trichloroacetic acid, or soluble protein in the juice of green grapes prior to ver-
phosphomolybdic acid tests (Amerine & Joslyn, 1970). aison (from the French word Veraison, is a term used by
Heating tests, however, correlate most closely with the viticulturalists to signify the onset of berry ripening),
behavior of bottled wine (Amerine, Berg, Kunkee, but this probably relates to difficulties encountered in
Ough, Singleton, & Webb, 1980). the protein extraction (see below). Protein synthesis
The presence of proteins in wines may also be of proceeds rapidly after veraison and parallels the rapid
interest to the winemaker. A close, positive correlation accumulation of sugars (Luis, 1983). Nevertheless, the
has been found between protein concentration and proteins present in wines do not correspond to a repre-
foam quality in sparkling wines (Brissonnet & Maujean, sentative fraction of the pulp proteins, since most of
1993). Proteins are particularly important for foam for- these are lost during vinification (Ferreira et al., 2000).
mation and foam stability in champagne wines (Bris- Fermentation is primarily responsible for the difference
sonnet & Maujean, 1993). Their surface properties, between grape juice and wine protein content (Murphey
hydrophobicity in particular, rather than isoelectric et al.). The lower protein levels typically found in wines
point or molecular size, are important in determining are mainly due to proteolysis and denaturation of the
foamability (Brissonnet & Maujean, 1991). grape proteins during fermentation, caused by proteases
and changes in the pH, respectively (Bayly & Berg;
The origin of the wine proteins Feuillat et al., 1980; Murphey et al.). In fact, the pro-
Wine proteins have long been considered as a mixture teins that end up in wines are those that are highly
of grape proteins and proteins from autolyzed yeast. resistant to proteolysis and to the low pH values char-
Yeasts may affect the wine protein composition in two acteristic of these beverages (Waters et al., 1992). In
ways: there may be a transfer of proteins to the wine addition, it has been estimated that approximately one-
during the process of yeast autolysis and/or the presence half of total wine protein is bound to grape phenolics.
of exocellular protease enzymes in the yeasts may con- During vinification, part of the soluble grape proteins
tribute to the hydrolysis of the must proteins (Feuillat, are precipitated via interaction with tannins (Somers &
Brillant, & Rochard, 1980). Bayly and Berg (1967), Ziemelis, 1973).
studying protein fractions in varietal white wines,
observed the release of peptides by the yeast during fer- The grape proteins
mentation but concluded that the yeast contributed no Grapes contain naturally a wide range of different
soluble heat-unstable proteins to the wine. More proteins. They are not high in protein compared with
recently, Ferreira, Monteiro, Piçarra-Pereira, Tan- some other fruits. The profile of soluble proteins found
ganho, Loureiro, and Teixeira (2000) employed immu- in the juice of ripe grapes often appears surprisingly
nological methods to show that at least the vast simple, with a predominance of a few low molecular
majority of the polypeptides present in wines derive mass proteins (Hsu & Heatherbell, 1987b; Murphey et
entirely from the grape pulp. al., 1989; Pueyo et al., 1993; Yokotsuka, Ebihara, &
Colby was the first to point out, in 1896, that the Sato, 1991). Difficulties in extracting proteins from
grape variety, the environment where the grapes are grapes at different stages of berry development (Tatter-
grown, ‘terroir’, and the fermentation process are the sall, Heeswijck, & Hoj, 1997) may well contribute, and
three main factors affecting the stability of young wines must be acknowledged as a possible contributor to this
(Colby 1986). The variations in protein content detected simple picture.
232 R.B. Ferreira et al. / Trends in Food Science & Technology 12 (2002) 230–239

There is a significant increase in total protein content (Jacobs, Dry, & Robinson, 1999). Enhanced resistance
after veraison and analysis of proteins by electrophor- to fungal pathogens has also been demonstrated in
esis indicates that a small number of proteins are syn- transgenic plants overexpressing chitinase or b-1,3-glu-
thesized in significant amounts during ripening canase, with a synergistic benefit when both genes are
(Tattersall et al., 1997). present (Jongedijk et al., 1995). Such an approach in
Some of the proteins synthesized in grapes during grapevine may well augment the problems associated to
ripening have recently been identified. The most abun- wine turbidity.
dant are pathogenesis-related (PR) proteins, including A thaumatin-like protein, with a molecular mass of 24
chitinases and thaumatin-like proteins (Robinson & kDa, is one of the most abundant proteins present in
Davies, 2000). Chitinases have been reported to account extracts of mature grapes (Tattersall et al., 1997). The
for ca. 50% of the soluble proteins in the grape berries, protein is encoded by a single Vitis vinifera gene, which
with thaumatin-like proteins comprising the majority of is expressed in a berry- and ripening-specific manner.
the other proteins (Waters, Hayasaka, Tatersall, Southern, northern and western analyses revealed that it
Adams, & Williams, 1998). These proteins constitute a is found in the berry only and is expressed in conjunc-
defence mechanism of plants against fungal attack (Van tion with the onset of sugar accumulation and berry
Loon, 1985). The constitutive levels of PR proteins are softening (Tattersall et al.). Thaumatin-like proteins
typically low in healthy plants, but the proteins are exhibit sequence homology to the thaumatins, a group
induced in response to wounding as well as to pathogen of intensely sweet tasting proteins from the fruit of an
attack. In support of this is the result of a recent study African shrub, Thaumatococcus danielli (Edens et al.,
showing that mature grapes grown under field condi- 1982). Although taste tests have shown that the thau-
tions and harvested from the same vineyard in con- matin-like proteins from tobacco are not sweet tasting
secutive years (and thus subjected to different types and (Singh et al., 1987), the possibility that a protein with
intensities of both biotic and abiotic stresses) accumu- sweet taste properties might occur naturally in wine is
late different proportions of the same set of PR proteins tantalizing and of potential significance to the wine
(osmotin, thaumatin-like, and chitinase) as their major industry (Waters, Shirley, & Williams, 1996). Thauma-
protein component (unpublished data). tin-like proteins have antifungal properties, possibly by
A number of chitinase isoforms [poly (1,4-N-acetyl-b- their ability to permeabilize cell membranes. Their pre-
d-glycosaminide) glycanohydrolase] (EC 3.2.1.14) have cise mechanism of action has not yet been elucidated
been found to be constitutively expressed or stress- and their exact role in the grapevine remains unknown.
induced in grape berries (Derckel, Legendre, Audran, However, the timing of its accumulation correlates with
Haye, & Lambert, 1996). Their expression is devel- the inability of the fungal pathogen U. necator to initi-
opmentally regulated in a sometimes tissue-specific ate new infections of the grape (Tattersall et al.).
manner and can be stimulated by wound or pathogen Chitinase and thaumatin-like protein are the pre-
attack. With a molecular mass of 32 kDa, chitinase is dominant proteins in free-run juice from grapes (Tat-
present in grapevine leaves, roots, stems and berries, but tersall et al., 1997) and have been identified as the major
the highest activity is located in the berries (Derckel et proteins present in white wines (Waters et al., 1996).
al.). The enzyme activity increases markedly at the onset Because they are resistant to proteolysis and to the low
of ripening in grapes and continues to increase wine pH, these PR proteins survive vinification and can
throughout the sugar accumulation phase of berry cause haze and sediment formation in wines (Waters et
development. In contrast, no b-1,3-glucanase[1,3- al.).
(1,3;1,4)-b-d-glucanohydrolase] (EC 3.2.1.6) activity is Other grape proteins whose synthesis undergoes a
detected in berries at any stage of development (Robin- rapid increase at the onset of ripening include four pro-
son, Jacobs, & Dry, 1997). The two PR proteins, chit- line-rich proteins, a small protein that is similar to the
inase and b-1,3-glucanase, are thought to provide non-catalytic, N-terminal domain of some pectin
defence against fungal pathogens due to the anti-fungal methylesterases, two glutamate-rich proteins, and some
properties that arise from their catalytic ability to putative stress response proteins such as a metallothio-
hydrolyse chitin and b-1,3-glucans, respectively, which nein, a transcription factor, a cytochrome P450-enzyme,
are abundant structural components of the cell walls of and proteins induced by water, sugar, and/or cold stress
invading fungal hyphae. In addition, the enzymes exhi- in other species (Davies & Robinson, 2000).
bit antifungal activity in vitro (Mauch, Mauch-Mani, &
Boller, 1988). Uncinula necator, the causal agent for Characterization of the wine proteins
grapevine powdery mildew, was reported to elevate the Advances in scientific knowledge are typically closely
activities of chitinase and b-1,3-glucanase in the leaves followed by the development of new technological
and berries of a number of susceptible grapevine culti- methods. In the case of wine protein research, the
vars. The increase in hydrolytic activity was directly application of modern biological techniques has con-
related to the severity of fungal infection on both organs tributed considerably to elucidate the structure of the
 R.B. Ferreira et al. / Trends in Food Science & Technology 12 (2002) 230–239 233

wine proteins and the problems associated with their demonstrated that the intense microheterogeneity that
presence in wines. Techniques applied include: ion appears to characterize the major wine polypeptides is
exchange FPLC (Dorrestein et al., 1995; Waters, Wal- also present in the grape juice polypeptides (unpub-
lace, Tate, & Williams, 1993; Waters et al., 1992), chro- lished data).
matofocusing on FPLC (Dawes, Boyes, Keene, & Immunological and N-terminal sequencing experi-
Heatherbell, 1994), HPLC (Santoro, 1995; Tyson, Luis, ments have also allowed the identification of the struc-
Day, Walker, & Lee, 1981), size exclusion chromato- turally related wine polypeptides—they are essentially
graphy (Pellerin, Waters, & Brillouet, 1993), affinity homologous to PR proteins, namely chitinase, thauma-
chromatography (Pellerin et al., Waters et al., 1993), tin-like protein and osmotin (Monteiro et al., in press;
two-dimensional electrophoresis (Lamikanra & Inyang, Waters et al., 1996).
1988), capillary electrophoresis (Dizy & Bisson, 1999; Furthermore, wines prepared from different grape
Ledoux, Dulau, & Dubourdieu, 1992), isoelectric varieties or subjected to different vinification procedures
focusing (Dawes et al.; Hsu & Heatherbell, 1987a; are essentially composed by identical sets of PR proteins
1987b; Hsu, Heatherbell, Flores, & Watson, 1987; (Ferreira et al., 2000).
Pueyo et al., 1993; Santoro), protein blotting (Hsu & The stability of the PR proteins to acidic pH and their
Heatherbell, 1987a, 1987b; Hsu et al.), and immu- very high resistance to proteolytic attack (Linthorst,
nological methods (Ferreira et al., 2000; Monteiro et 1991) means that winemaking functions as a selective
al., 1999). However, the isolation, separation and extraction procedure for grape berry, releasing vacuolar
characterization of the proteins of musts and wines acids and hydrolytic enzymes that precipitate and
is a difficult task due to their low concentration degrade many of the grape proteins. Subsequent fer-
and strong interaction with endogenous polyphenols mentation of the must by yeast further augments the
and other non-protein compounds. Therefore, con- proteolytic pool (Lagace & Bisson, 1990). This combi-
ventional extraction and concentration procedures nation of low pH (3.0–3.8) and proteolytic activity
may lead to protein precipitation by phenols and oxi- ensures that only proteins resistant to these conditions,
dases, to degradation and/or to modification without such as PR proteins, survive the winemaking process,
precipitation. becoming the troublesome proteins of wines.
Wines have been reported to contain polypeptides There are reports in the literature about the glycosy-
ranging in molecular mass from 9 to 62 kDa and iso- lation status of wine proteins. Thus, whilst some
electric points from 3 to 9 (Brissonet & Maujean, 1993; authors claim that all wine proteins are glycosylated
Hsu & Heatherbell, 1987b; Lamikanra & Inyang, 1988). (Paetzold, Dulau, & Dubourdieu, 1990; Yokotsuka et
However, the vast majority of the wine proteins exhibit al., 1991), the majority of studies suggest that the
low molecular masses (20–30 kDa) and low isoelectric occurrence of glycosylated proteins in wines is not
points (4.1 < pI< 5.8), possessing a net positive charge at common (Hsu & Heatherbell, 1987b; Waters, 1991;
the pH values encountered in wines (Brissonet & Mau- Waters et al., 1993).
jean, 1993; Ferreira et al., 2000; Hsu & Heatherbell, Two weakly acidic, arabinogalactan proteins (AGPs)
1987b). A simple denaturing electrophoretic analysis were purified and characterized from red wines (Pellerin
suggests that the wine total protein fraction is mainly et al., 1993). They represent, respectively, 4.3 and 5.2%
composed of only a few polypeptides. However, a more of the total alcohol-precipitable colloids. Their protein
detailed examination reveals the presence of many tens content is less than 10%. These AGPs, which reduce the
of polypeptides with similar molecular masses but subtle filterability of the wine (Belleville, Brillouet, Fuente, &
differences in electric charge (Monteiro, Piçarra-Pereira, Moutounet, 1990), are probably derived from native
Mesquita, Loureiro, Teixeira, & Ferreira, in press). grape pectins by the action of endogenous pectolytic
Immunological and N-terminal sequencing experiments enzymes.
revealed that these polypeptides are structurally related,
possibly deriving from one or a few common precursors Which proteins are responsible for wine turbidity
synthesized during grape formation, which may have Although there have been numerous investigations on
undergone limited proteolysis in the later stages of grape, juice and wine proteins in recent years, the nature
grape maturation or during vinification, to generate the of the proteins responsible for, or the factors that trig-
variety of distinct polypeptides present in wines (Mon- ger, wine turbidity remains unclear.
teiro et al., in press). It is well known that protein con- The presence of proteins in wine is certainly a pre-
figuration is subjected to several modifications during requisite for haze formation and it seems generally
winemaking, as suggested by the fact that while the accepted that the higher the wine total protein content,
molecular masses of some protein bands in the wines are the higher its tendency to become unstable (Mesquita,
the same of those found in their juices, their isoelectric Piçarra-Pereira, Monteiro, Loureiro, Teixeira, & Fer-
points differ (Lamikanra & Inyang, 1988; Murphey et reira, in press). However, some authors claim that pro-
al., 1989; Pueyo et al., 1993). However, it was recently tein instability does not correlate well with total protein
234 R.B. Ferreira et al. / Trends in Food Science & Technology 12 (2002) 230–239

concentration, and, therefore, the potential of a wine to formation is controlled by factors of non-protein origin.
form haze is not predictable from its total protein con- Indeed, it has been suggested that the interaction of
tent (Bayly & Berg, 1967; Moretti & Berg, 1965). If this proteins with other components in wine must be con-
is correct, two alternative hypotheses may be advanced sidered in order to determine factors resulting in wine
to explain the insolubilization of proteins in wines: (1) instability (Dawes et al.).
individual proteins behave differently in their sensitivity Waters et al. (1996) were the first to identify the pro-
to heat denaturation, contributing differentially to haze teins that cause haze in wines as PR proteins from the
formation, in which case, only part of the protein mix- grape berry. Dawes et al. (1994) fractionated the total
ture is responsible for instability rather than the entire wine proteins in several groups on the basis of their
protein content; (2) although protein-dependent, the isoelectric points, and observed that all protein groups
development of turbidity in wines is controlled by a were thermally unstable contributing to haze formation.
number of factors of non-protein origin. Different wines exhibit distinct patterns of turbidity
Several studies have reported that total protein con- formation when subjected to increasing temperatures
tent is a poor index of the tendency of a wine to protein (Mesquita et al., in press). The observation that these
clouding, in the sense that some protein fractions are wines are essentially composed by identical sets of PR
heat-labile while others contribute little to wine proteins suggests that the pattern of protein insolubili-
instability (Bayly & Berg, 1967; Koch & Brethauer, zation is not determined by the protein molecules
1957; Ledoux et al., 1992; Moretti & Berg, 1965; Paet- themselves, depending on some other non-protein fac-
zold et al., 1990; Waters et al., 1991, 1992). However, tors, such as procyanidins, the wine pH and the pre-
the exact nature of the proteins that are responsible for sence of polysaccharides. In a recent study, Mesquita et
wine turbidity is still subject to debate. al. showed that the typical pattern of haze formation
The incomplete understanding is largely due to diffi- with increasing temperature is not altered when the wine
culties in targeting particular wine proteins involved in is depleted of its own protein and back-added with a
heat instability. An ideal research strategy to overcome similar amount of bovine serum albumin. Furthermore,
this problem is one that could demonstrate haze pre- the typical pattern of turbidity formation with increas-
vention upon removal of specific protein fractions and ing temperature remains essentially unaltered when the
haze induction upon back-addition of the material wine is depleted of its own protein and back-added with
(Waters et al., 1991). In such studies, bovine serum an equivalent amount of protein from another wine.
albumin may be used as a suitable model in wine haze The work reported by Mesquita et al. (in press) also
experiments because it has a haze potential similar to showed that protein instability is induced by low pH
that of wine proteins (Waters et al., 1991). and high concentrations of polysaccharides. Increasing
Molecular size and isoelectric properties are protein the pH or decreasing the amount of polysaccharides
characteristics that have been analysed in relation to reduced protein precipitation at high temperatures,
their susceptibility to solubility limitations. Some of whereas alcohol content had no effect. However, it has
these experiments produced contradictory conclusions. also been reported that protein instability induced by
Thus, some studies claim that the lower molecular mass, high ethanol and low pH can be a source of haze for-
lower pI proteins are the major and most important mation in wine (Lagace & Bisson, 1990).
fractions contributing to protein instability in wines Polysaccharides and polyphenols are essential com-
(Hsu & Heatherbell, 1987a, 1987b; Hsu et al., 1987; pounds in wines because they contribute to the organo-
Mesrob, Gorinova, & Tzakov, 1983). leptic properties and influence clarification and
Other studies indicate that it was not until the lower stabilization phenomena (Vernhet, Pellerin, Prieur,
molecular mass and higher pI (Heatherbell et al., 1984; Osmianski, & Moutounet, 1996). They may participate
Lee, 1985; Ngaba & Heatherbell, 1981), the high mole- in haze formation or modify the efficiency of fining
cular mass (Lamikanra & Inyang, 1988; Millies, 1975) treatments (Mesquita et al.; Santoro, 1995; Segarra,
or the glycoprotein (Hsu & Heatherbell, 1987a) frac- Lao, Löpez-Tamames, & Torre-Boronat, 1995).
tions were removed by bentonite fining that the wines Some wine polysaccharides (such as yeast mannopro-
became protein stabilized to heat testing. teins, and grape arabinogalactan-proteins and rhamno-
The recent observations that wines are essentially galacturonan) carry negative charges in the wine pH
composed by identical sets of polypeptides, which have range. As a consequence, these wine polysaccharides
been identified as PR proteins (Ferreira et al., 2000; may establish electrostatic and ionic interactions with
Monteiro et al., 2001; Waters et al., 1996), and that the other wine components (Vernhet et al., 1996), resulting
haze-forming wine proteins are apparently similar in in the formation of either soluble or insoluble complexes
wines vinified from different grape varieties (Dawes in a process that is strongly dependent on their net
et al., 1994; Hsu & Heatherbell, 1987b; Murphey et al., electrical charge and on the structure of their functional
1989; Paetzold et al., 1990; Pueyo et al., 1993; Waters groups (Samanti, Singhal, Kulkaml, & Rege, 1993).
et al., 1992) add support to the hypothesis that haze Therefore, some polysaccharides have been described as
 R.B. Ferreira et al. / Trends in Food Science & Technology 12 (2002) 230–239 235

protective colloids because they prevent protein haze than 50 years ago. Bentonite, carrying a net negative
formation (Waters, Pellerin, & Brillouet, 1994a), charge at the wine pH, interacts electrostatically with
whereas others exhibit no effect or increase wine the positively charged wine proteins, producing their
instability (Mesquita et al., 2001; Pellerin, Waters, Bril- flocculation (Hsu & Heatherbell, 1987a; Lamikanra &
louet, & Moutounet, 1994). The improvement of wine Inyang, 1988).
thermal stability by wine lees is not due to either The relationship between wine pH and protein iso-
removal of the unstable protein fractions or to the pro- electric point governs the cationic nature of the wine
teolytic activities present in yeasts but to the addition of proteins and thus their tendency to be adsorbed elec-
yeast mannoproteins (Ledoux et al., 1992). Waters et al. trostatically by the negatively charged plate surfaces of
(1993) identified and isolated a natural haze protective bentonite (Bayly & Berg, 1967; Hsu & Heatherbell,
wine macromolecule composed of a polysaccharide 1987a; Koch & Sajak, 1959; Lee, 1985; Moretti & Berg,
component and a protein component. 1965; Somers & Ziemelis, 1973).
In contrast to polysaccharides, polyphenols carry no However, bentonite is not equally effective in remov-
charge or negligible negative charges at the wine pH, so ing all proteins from the wine. A number of studies have
that electrostatic and ionic forces are not determinant to been reported showing that different protein fractions
their physico-chemical reactivity. The major interac- may require distinct bentonite concentrations for
tions rat seem more likely are hydrogen bonding and removal (Hsu & Heatherbell, 1987a; Murphey et al.,
hydrophobic interactions, as has been shown in the case 1989; Paetzold et al., 1990; Santoro, 1995). In another
of the complexation between proanthocyanidins and study, Dawes et al. (1994) observed that the amount of
proteins (Vernhet et al., 1996). Procyanidins have been protein depletion correlates approximately linearly with
shown to be implicated in the formation of protein haze the level of bentonite addition.
in wines because many soluble wine proteins previously The amounts of bentonite required to stabilize wines
liberated of procyanidins precipitate to give haze in has increased over the last 10–20 years. Doses of 0.2–0.4
wine but not in model wine that is devoid of phenolic g l1 bentonite were generally enough to stabilize the
compounds (Waters, Peng, Pocock, & Williams, 1995a). wines. However, nowadays, to ensure a secure stabili-
Also, isolated protein fractions added to model solu- zation, doses of 0.8–1 g l1 bentonite are often
tions at wine pH interact with fractions of grape con- employed (Hsu & Heatherbell, 1987a; Paetzold et al.,
densed tannin to produce significant turbidity 1990).
(Yokotsuka et al., 1991). Koch and Sajak (1959) were As a cation exchanger, bentonite adsorption is not
among the first to propose that wine hazes contain tan- specific for proteins, removing other charged species or
nin. It was later suggested that about 50% of white wine aggregates. As a result, the large amounts of added
protein is bound to a much smaller quantity of flavo- bentonite are responsible for decreasing the organolep-
noid material, and that the interaction of wine protein tic properties of the wines by removal of important
with flavonoids makes the proteins prone to precipita- aroma and flavor components (Voilley, Lamer, Dubois,
tion (Somers & Zremelis, 1973). In fact, both heat- & Feuillat, 1990). In addition, anywhere from 5 to 20%
induced and natural hazes isolated from various white of the wine volume treated with bentonite may be lost as
wines were all shown to contain procyanidins, with a bentonite lees, depending upon dowstream processing
content ranging from < 0.02 to 4.9% (w/w) (Waters et steps (Lagace & Bisson, 1990). Therefore, an increasing
al., 1995a). interest in alternative practices to bentonite treatment
Wine proteins may also act as deposition centers for for protein stability has developed. So far, bentonite
many cations or can bind to ferric phosphate leading, in treatment remains the only effective method to stabilize
both cases, to sediments or suspensions affecting the wines. Other fining treatments, such as silica sol/gelatin,
limpidity of the wine. Such turbidity may be due to the also modify the organoleptic quality of the wines (Mil-
formation of protein-tannin complexes in the presence lies, 1975; Waters et al., 1994a, 1994b).
of traces of heavy metals (Krug, 1968). Considerable attention was given to the use of pro-
teolytic enzymes as an alternative protein depletion
Removal of proteins from wines technique to remove wine proteins through enzymatic
Bentonite fining (the term ‘fining’ is used in wine- hydrolysis into small peptides and their component
making to describe the deliberate addition of an amino acids. These studies, which included the use of
adsorptive compound that is followed by the settling or grape proteases (Cordonnier & Dugal, 1968), of yeast
precipitation of partially soluble components from the proteases (Feuillat et al., 1980; Ledoux et al., 1992), and
wine) is currently and commonly utilized by the wine of exogenous proteases (Modra, 1989), must be per-
industry as a clarifying technique to remove proteins formed at the low temperatures appropriate to wine-
that are a source of potential haze in wines. making. It is now clear that treatment of juices and
The use of the cation exchanger bentonite to remove wines with proteolytic enzymes does not confer protec-
the unstable proteins from wines was proposed more tion against protein precipitation, because the wine
236 R.B. Ferreira et al. / Trends in Food Science & Technology 12 (2002) 230–239

proteins are not hydrolysed under these conditions lerin et al., 1994). The presence of these polysaccharides
(Heatherbell et al., 1984; Modra & Williams, 1988; in wines does not prevent the thermal denaturation of
Waters et al., 1992, 1995b). the proteins. Rather, they reduce the visible haziness
This is not due to the presence of protease inhibitors induced by heating wine proteins by decreasing the
in wines because the protease preparations are highly particle size of the haze (Waters et al., 1993). These
active on exogenous protein substrates (e.g. casein, col- observations explain the lee-induced protein stabiliza-
lagen, bovine serum albumin) introduced into wines tion of white wines (Ledoux et al., 1992), which is
(Modra & Williams, 1988; Waters et al., 1992). attributed to the release of mannoproteins from the
These observations suggest that grape and wine yeasts (Pellerin et al., 1994).
proteins are extremely resistant to proteolytic attack Tannins are well known to interact with proteins,
under winemaking conditions (Waters et al., 1992). resulting in precipitation. Furthermore, tannin extracted
Furthermore, this proteolytic resistance is not due to from the grape seed and subsequently immobilized also
phenolic association or glycosylation (Waters et al., exhibit the ability to bind proteins (Oh, Hoff, & Hoff,
1995b), indicating that it is an inherent property of these 1985). The capacity of immobilized grape proanthocya-
polymers. nidins to bind proteins from wine results in protein-
This is not surprising considering that the wine pro- stable wines. However, their use is apparently limited by
teins are the only grape proteins that survived the reduction in protein-binding capacity after a small
winemaking process, which included the exposure of the number of regeneration cycles (Powers, Nagel, &
berry proteins to natural grape and yeast proteases for Weller, 1988).
very long periods of time. The recent finding that most Development of other methodologies capable of spe-
wine proteins are PR proteins explains the inherent cifically removing the proteins from wines have been
resistance of the wine proteins to proteolytic degrada- hampered by a number of factors. Removal with spe-
tion. In conclusion, strategies based on the proteolysis cific, immobilized antibodies, for example, has to take
of wine proteins have proved unsuccessful in practice into account the low pH values typical of wines,
and, given the nature of PR proteins, are probably incompatible with the antigen-antibody interactions.
futile. The use of molecular biology techniques to silence the
A number of studies have evaluated the effectiveness expression of PR proteins would probably lead to pro-
of ultrafiltration (UF) with membranes of different tein-free wines, but would certainly make the vines
nominal molecular mass cut-offs (MMCOs) as a bento- much more susceptible to fungal attack or to stress in
nite substitute for protein stabilizing grape juice and general.
wine. A progressive increase in membrane retention of It is interesting to note that the addition of protein to
soluble protein is observed with decreasing MMCO wines may, under some circumstances, be beneficial.
(from 100 to 10 kDa), up to 99% of the wine protein Fining is a common practice in many wineries to reduce
being retained with membranes of 10 kDa MMCO—3– levels of certain components from wine. In particular,
20 mg l1 protein are frequently left in ultrafiltered wine softening and reducing wine astringency is the major
permeates (Hsu et al., 1987). Therefore, ultrafiltration purpose of fining with protein. The proteins used are
with membranes of 10 kDa cut-off allow a good and generally casein, albumin, or gelatin.
efficient elimination of the proteins from wines (Hsu et
al.). However, this technique leads also to great losses in Conclusions
important organoleptic compounds (Feuillat, Peyron, & There is increasing evidence suggesting that the ther-
Jousset-Drouhin, 1987). mal stability of the wine proteins is linked to their col-
An interesting alternative to achieve protein stabili- loidal environment, making the wine proteins more or
zation in wines involves the addition of certain poly- less susceptible to precipitation by heat in the presence
saccharides to wines. Two active polysaccharides in of certain macromolecules such as, respectively, certain
promoting the stability of wines have been isolated, and polysaccharides or polyphenols. Therefore, the design
characterized from the total colloidal fraction of a red of new clarification technologies requires a better
wine (Waters et al., 1993, 1994a): (1) An arabinoga- knowledge of proteins, polysaccharides and other wine
lactan protein, representing 0.1% of the total wine constituents and of the interactions established among
polysaccharides, which derives from the grapes; protein them.
comprises 13% of the polymer. (2) A high molecular Bentonite treatment, with a number of important
mass mannoprotein, present in very low concentration drawbacks, remains the only effective methods to stabi-
in wine (0.007% of total polysaccharides), which derives lize wines. The recent observation that the major wine
from fermenting yeasts; protein constitutes 30% of the proteins are PR proteins, polymers of known resistance
mannoprotein. A protective effect on the wine proteins to proteolytic attack, suggests that strategies to remove
has also been observed with an arabinogalactan protein these proteins from wine based on proteolysis under
from apple and with a gum from a Senegal acacia (Pel- traditional winemaking conditions are probably futile.
 R.B. Ferreira et al. / Trends in Food Science & Technology 12 (2002) 230–239 237

In an era where the applications of molecular biology Cordonnier, R., & Dugal, A. (1968). Les activités protéolytiques du
are expanding at an enormous rate, the problems asso- raisin. Ann. Technol. Agric., 17, 189–206.
Davies, C., & Robinson, S. P. (2000). Differential screening indicates
ciated with wine turbidity could be tackled by geneti- a dramatic change in mRNA profiles during grape berry ripening.
cally modifying the patterns of gene expression in the Cloning and characterization of cDNAs encoding putative
grapes. Thus, lowering the expression of the genes cell wall and stress response proteins. Plant Physiology, 12,
encoding for PR proteins would most probably decrease 803–812.
or eliminate the problems associated to wine protein Dawes, H., Boyes, S., Keene, J., & Heatherbell, D. (1994). Protein
instability in wines: influence of protein isoelectric point.
instability. However, given the physiological role of PR American Journal of Enology and Viticulture, 45, 319–326.
proteins, the genetically modified vines would certainly Derckel, J.-P., Legendre, L., Audran, J.-C., Haye, B., & Lambert, B.
become more susceptible to the attack of fungal patho- (1996). Chitinases of the grapevine (Vitis vinifera L.): five isoforms
gens. On the contrary, the actual trend seems to be to go induced in leaves by salicylic acid are constitutively expressed in
in the opposite direction, i.e. to overexpress PR proteins other tissues. Plant Science, 119, 31–37.
Dizy, M., & Bisson, L. F. (1999). White wine protein analysis by
in order to obtain vines with enhanced resistance to capillary zone electrophoresis. American Journal of Enology and
fungal attack. For example, a rice chitinase gene was Viticulture, 50, 120–127.
recently introduced into the somatic embryos of grape- Dorrestein, E., Ferreira, R. B., Laureano, O., & Teixeira, A. R. (1995).
vine by Agrobacterium infection. Some of the transfor- Electrophoretic and FPLC analysis of soluble proteins in four
mants obtained exhibited increased resistance against portuguese wines. American Journal of Enology and Viticulture,
46, 235–242.
powdery mildew caused by Uncinula necator (Yama- Edens, L., Heslinga, L., Klok, R., Ledeboer, A. M., Maat, J., Tooner,
moto et al., 2000). This tendency will certainly augment M. Y., Visser, C., & Verrips, C. T. (1982). Cloning of cDNA
the problems associated with protein instability in encoding the sweet-tasting plant protein thaumatin and its
wines. expression in Escherichia coli. Gene, 18, 1–12.
Identification of factors that cause accumulation of Ferreira, R. B., Monteiro, S., Piçarra-Pereira, M. A., Tanganho, M. C.,
Loureiro, V. B., & Teixeira, A. R. (2000). Characterization of the
PR proteins in berries could allow the development of proteins from grapes and wines by immunological methods.
tactics to minimize their expression, without compro- American Journal of Enology and Viticulture, 51, 2–28.
mising the disease resistance of grapevines, Therefore, Feuillat, M., Brillant, G., & Rochard, J. (1980). Mise en evidence
appropriate viticultural practices, rather than post har- d’une production de proteases exocellulaires par les levures au
vesting processing, may hold the ultimate key to con- cours de la fermentation alcoolique du moût de raisin. Connaiss.
Vigne Vin, 14, 37–52.
trolling the level of protein in wine. Feuillat, M., Peyron, D., & Jousset-Drouhin, V. (1987). Influence de
la filtration tangentielle des vins sur leur composition physico-
Acknowledgements chimique et leur caractères sensoriels. Application aux vins de
This work was supported by the Fundação para a Bourgogne. Bull. OIV, 60, 227–240.
Ciência e a Tecnologia, under project no. PRAXIS Heatherbell, D., Ngaba, P., Fombin, J., Watson, B., Garcia, Z., Flores,
J., & Hsu, J. (1984). Recent developments in the application of
XXI/3/3.2/AGR/2180/95. ultrafiltration and protease enzymes to grape juice and wine
processing. In D. A. Heatherbell, P. B. Lombard, F. W. Bodyfelt, &
S. F. Price, eds., Proceedings of the International Symposium on
References Cool Climate Viticulture and Enology, Oregon State University,
Corvallis, OR (pp. 418–445).
Amerine, M. A., Berg, H. W., Kunkee, R. E., Ough, C. S., Singleton, Hsu, J.-C., & Heatherbell, D. A. (1987a). Heat-unstable proteins in
V. L, & Webb, A. D. (1980). The technology of wine making. wine. I. Characterization and removal by bentonite fining and
Westport, CT: AVI Publishing Co. heat treatment. American Journal of Enology and Viticulture, 38,
Amerine, M. A., & Joslyn, M. A. Table wines: the technology of their 11–16.
production. Berkeley: University of California Press. Hsu, J. C., & Heatherbell, D. A. (1987). Isolation and character-
Bayly, F. C., & Berg, H. W. (1967). Grape and wine proteins of white ization of soluble proteins in grapes, grape juice, and wine.
wine varietals. American Journal of Enology and Viticulture, 24, American Journal of Enology and Viticulture, 38, 6–10.
18–32. Hsu, J. C., Heatherbell, D. A., Flores, J. H., & Watson, B. T. (1987).
Belleville, M.-P., Brillouet, J.-M., Fuente, B. T., & Moutounet, M. Heat-unstable proteins in grape juice and wine. II. Characteriza-
(1990). Polysaccharide effects on cross-flow microfiltration of two tion and removal by ultrafiltration. American Journal of Enology
red wines with a microporous alumina membrane. Journal of and Viticulture, 38, 17–22.
Food Science, 55, 1598–1602. Jacobs, A. K., Dry, I. B., & Robinson, S. P. (1999). Induction of
Berg, H. W., & Akiyoshi, M. (1961). Determination of protein stabi- different pathogenesis-related cDNAs in grapevine infected with
lity in wine. American Journal of Enology and Viticulture, 12, 107– powdery mildew and treated with ethephon. Plant Pathology, 48,
110. 325–336.
Brissonnet, F., & Maujean, A. (1991). Identification of some foam- Jongedijk, E., Tigelaar, H., van Roeckel, J. S. C., Bres-Vloemans, S. A,
active compounds in champagne base wines. American Journal Dekker, I., van den Elzen, P. J. M., Cornelissen, B. J. C., & Mel-
of Enology and Viticulture, 42, 97–102. chers, L. S. (1995). Synergistic activity of chitinases and b-1,3-
Brissonnet, F., & Maujean, A. (1993). Characterization of foaming glucanases enhances fungal resistance in transgenic tomato
proteins in a champagne base wine. American Journal of Enology plants. Euphytica, 85, 173–180.
and Viticulture, 44, 297–301. Koch, J., & Brethauer, G. (1957). Zur kenntnis der eiweissstoffe des
Colby, G. E. (1896). Report of viticultural work, 1887—1893. weines. I. Chemische zusammensetzung des wärmetrubes
University of California: College of Agriculture. kurzzeiterhitzter weissweine und seine beziehung zur
238 R.B. Ferreira et al. / Trends in Food Science & Technology 12 (2002) 230–239

eiweisstrübung und zum weineweiss. Z. Lebensm. Untersuch. Paetzold, M., Dulau, L., & Dubourdieu, D. (1990). Fractionnement et
Forsch., 106, 272–280. caractérisation des glycoproteines dans les moûts de raisins
Koch, J., & Sajak, E. (1959). A review and some studies on grape blancs. J. Int. Sci. Vigne Vin, 24, 13–28.
protein. American Journal of Enology and Viticulture, 10, 114– Pellerin, P., Waters, E., & Brillouet, J.-M. (1993). Characterization of
123. two arabinogalactan-proteins from red wine. Carbohydrate
Krug, K. (1968). Die ursachen der eiweissnachtrubungen bei wein Polymers, 22, 187–192.
des ausfall der eiweisschönungen. Wein-Wiss, 23, 8–29. Pellerin, P., Waters, E., Brillouet, J.-M., & Moutounet, M. (1994). Effet
Lagace, L. S., & Bisson, L. F. (1990). Survey of yeast acid proteases de polysaccharides sur la formation de trouble protéique dans
for effectiveness of wine haze reduction. American Journal of un vin blanc. J. Int. Sci. Vigne Vin, 28, 213–225.
Enology and Viticulture, 41, 147–155. Powers, J. R., Nagel, C. W., & Weller, K. (1988). Protein removal
Lamikanra, O., & Inyang, I. D. (1988). Temperature influence on from a wine by immobilized grape proanthocyanidins. American
Muscadine wine protein characteristics. American Journal of Journal of Enology and Viticulture, 39, 117–120.
Enology and Viticulture, 39, 113–116. Pueyo, E., Dizy, M., & Polo, M. C. (1993). Varietal differentiation of
Ledoux, V., Dulau, L., & Dubourdieu, D. (1992). Interprétation de must and wines by means of protein fraction. American Journal
l’amélioration de la stabilité protéique des vins au cours de of Enology and Viticulture, 44, 255–260.
l’élevage sur lies. J. Intern. Sci. Vigne Vin, 26, 239–251. Robinson, S. P., & Davies, C. (2000). Molecular biology of grape
Lee, T. (1985). Protein instability: nature, characterization and berry ripening. Australian Journal of Grape Wine Research, 6,
removal by bentonite. In T. H. Lee, ed., Physical stability of wine. 175–188.
Proceedings of the Seminar by Australian Society of Viticulture Robinson, S. P., Jacobs, A. K., & Dry, I. B. (1997). A class IV chitinase
and Oenology (pp. 23–40). is highly expressed in grape berries during ripening. Plant
Linthorst, H. J. M. (1991). Pathogenesis-related proteins of plants. Physiology, 114, 771–778.
Critical Reviews in Plant Science, 10, 123–150. Samant, S. K., Singhal, R. S., Kulkaml, P. R., & Rege, D. V. (1993).
Luis, E. S. (1983). A study of proteins during grape maturation, juice Protein-polysaccharide interactions: a new approach in food
preparation and wine processing. Dissertation, University New formulation. International Journal of Food Science and Technol-
S. Wales, Australia. ogy, 28, 547–562.
Mauch, F., Mauch-Mani, B., & Boller, T. (1988). Antifungal hydro- Santoro, M. (1995). Fractionation and characterization of must and
lases in pea tissue. II. Inhibition of fungal growth by combina- wine proteins. American Journal of Enology and Viticulture, 46,
tions of chitinase and b-1,3-glucanase. Plant Physiology, 88, 936– 250–254.
942. Segarra, I., Lao, C., López-Tamames, E., & Torre-Boronat, M. C.
Mesquita, P. R., Piçarra-Pereira, M. A., Monteiro, S., Loureiro, V. B., (1995). Spectrophotometric methods for the analysis of poly-
Teixeira, A. R., & Ferreira, R. B. (in press). The importance of the saccharide levels in winemaking products. American Journal of
major characteristics of wines in determining their protein Enology and Viticulture, 46, 564–570.
instability. American Journal of Enology and Viticulture. Singh, N. K., Bracker, C. A., Hasegawa, P. M., Handa, A. K., Buckel, S.,
Mesrob, B., Gorinova, N., & Tzakov, D. (1983). Characterization of Hermodson, M. A., Pkankoch, E., Regnier, F. E., & Bressan, R. A.
the electrical properties and of the molecular weights of the (1987). Characterization of osmotin. Plant Physiology, 85, 529–536.
proteins of white wines. Die Nahrung, 27, 727–733. Somers, T. C., & Ziemelis, G. (1973). Direct determination of wine
Millies, K. (1975). Protein stabilization of wine using silica sol/gelatin proteins. American Journal of Enology and Viticulture, 24,
fining. Mitteilungsblatt der GDCh-Fachgruppe. Lebensm. Gerich. 47–50.
Chemie, 29, 50–53. Tattersall, D. B., Heeswijck, R., & Hoj, P. B. (1997). Identification and
Modra, E. J. (1989). In IV Symposium International Oenologie, characterization of a fruit-specific, thaumatin-like protein that
Bordeaux, France. accumulates at very high levels in conjunction with the onset
Modra, E. J., & Williams, P. J. (1988). Are proteases active in wines of sugar accumulation and berry softening in grapes. Plant
and juices? Australian Grapegrower and Winemaker, 292, Physiology, 114, 759–769.
42–46. Tyson, P. J., Luis, E. S., Day, W. R., Walker, B., & Lee, T. H. (1981).
Monteiro, S., Piçarra-Pereira, M. A., Mesquita, P. R., Loureiro, V. B., Research note on estimation of soluble protein in must and
Teixeira, A., & Ferreira, R. B. (2001). The wide diversity of the wine by high-performance liquid chromatography. American
structurally similar wine proteins. Journal of Agriculture and Food Journal of Enology and Viticulture, 32, 241–243.
Chemistry, 49, 3999–4010. Van Loon, L. C. (1985). Pathogenesis-related proteins. Plant
Monteiro, S., Piçarra-Pereira, M. A., Tanganho, M. C., Rente, J. P., Molecular Biology, 4, 111–116.
Loureiro, V. B., Teixeira, A. R., & Ferreira, R. B. (1999). Preparation Vernhet, A., Pellerin, P., Prieur, C., Osmianski, J., & Moutounet, M.
of polyclonal antibodies specific for wine proteins. Journal of (1996). Charge properties of some grape and wine poly-
Science and Food Agriculture, 79, 772–778. saccharide and polyphenolic fractions. American Journal of
Moretti, R. H., & Berg, H. W. (1965). Variability among wines to Enology and Viticulture, 47, 25–29.
protein clouding. American Journal of Enology and Viticulture, 16, Voilley, A., Lamer, C., Dubois, P., & Feuillat, M. (1990). Influence of
69–78. macromolecules and treatments on the behavior of aroma
Murphey, J. M., Spayd, S. E., & Powers, J. R. (1989). Effect of grape compounds in a model wine. Journal of Agricultural and Food
maturation on soluble protein characteristics of Gewürztraminer Chemistry, 38, 248–251.
and White Riesling juice and wine. American Journal of Enology Waters, E. J. (1991). Heat unstable wine proteins and their interac-
and Viticulture, 40, 199–207. tions with wine polysaccharides. PhD thesis, Adelaide University,
Ngaba, P. R., & Heatherbell, D. A. (1981). Protein instability in wine. South Australia.
I. Characterization of unstable proteins. II. The application of Waters, E. J., Hayasaka, Y., Tattersall, D. B., Adams, K. S., & Wil-
protease enzymes as bentonite substitutes in protein stabiliza- liams, P. J. (1998). Sequence analysis of grape (Vitis vinifera) berry
tion wine. In Proceedings of the 32nd Annual Meeting of the chitinases that cause haze formation in wines. Journal of
American Society of Enologists, San Diego, CA. Agricultural and Food Chemistry, 46, 4950–4957.
Oh, I. H., Hoff, J. E., & Hoff, L. A. (1985). Immobilized condensed Waters, E. J., Pellerin, P., & Brillouet, J.-M. (1994). A wine arabino-
tannins and their interactions with proteins. Journal of Food galactan-protein that reduce heat-induced wine protein haze.
Science, 50, 1652–1654. Bioscience, Biotechnology and Biochemistry, 58, 43–48.
 R.B. Ferreira et al. / Trends in Food Science & Technology 12 (2002) 230–239 239

Waters, E. J., Pellerin, P., & Brillouet, J. M. (1994). A Saccharomyces Isolation and partial characterization of a natural haze protec-
mannoprotein that protects wine from protein haze. tive factor from wine. Journal of Agricultural and Food Chemistry,
Carbohydrate Polymers, 23, 185–191. 41, 724–730.
Waters, E. J., Peng, Z., Pocock, K. F., & Williams, P. J. (1995). Waters, E. J., Wallace, W., & Williams, P. J. (1991). Heat haze
Proteins in white wine, I: procyanidin occurrence in soluble characteristics of fractionated wine proteins. American Journal
proteins and insoluble protein hazes and its relationship to of Enology and Viticulture, 42, 123–127.
protein instability. Australian Journal of Grape Wine Research, Waters, E. J., Wallace, W., & Williams, P. J. (1992). The identification
1, 86–93. of heat-unstable wine proteins and their resistance to pepti-
Waters, E. J., Peng, Z., Pocock, K. F., & Williams, P. J. (1995). Proteins dases. Journal of Agricultural and Food Chemistry, 40, 1514–1519.
in white wine, II: Their resistance to proteolysis is not due to Yamamoto, T., Iketani, H., Ieki, H., Nishizawa, Y., Notsuka, K., Hibi,
either phenolic association or glycosylation. Australian Journal of T., Hayashi, T., & Matsuta, N. (2000). Transgenic grapevine plants
Grape Wine Research, 1, 94–99. expressing a rice chitinase with enhanced resistance to fungal
Waters, E. J., Shirley, N. J., & Williams, P. J. (1996). Nuisance proteins pathogens. Plant Cell Reports, 19, 639–646.
of wine are grape pathogenesis-related proteins. Journal of Yokotsuka, K., Ebihara, T., & Sato, T. (1991). Comparison of soluble
Agricultural and Food Chemistry, 44, 3–5. proteins in juice and wine from Koshu juice. Journal of Fermen-
Waters, E. J., Wallace, W., Tate, M. E., & Williams, P. J. (1993). tation Technology, 71, 248–253.