Fractionation of Gliadin

Hydrolysates in Water-Ethanol by
Ultrafiltration with Modified or
Unmodified Membranes
Serge Brot,1 Bernard Chaufer,2 Yannick Basso,1 Ccile
Legay,1 Yves Popineau1
1
Institut National de la Recherche Agronomique, Laboratoire de Biochimie et
Technologie des Prote ines, Rue de la Ge raudie` re, BP 71627, 44316 Nantes
Cedex 3, France; telephone: (33) 240-67-51-30; fax: (33)
240-67-50-25; e-mail: berot@nantes.inra.fr
2
Universite de Rennes 1, Laboratoire des Proce de s de Se paration, 85 Rue
de St Brieuc, 35000 Rennes, France
Received 24 April 1998; accepted 18 August 1998

Abstract: Ultrafiltration was applied to the

fractionation of neutral vs. charged peptides of similar
size. The pep- tides, produced from gliadins, a major
fraction of wheat storage proteins, were obtained by
limited hydrolysis with a-chymotrypsin in water-
ethanol 80/20 (v/v). Pep- tides, according to their
elution by RP-HPLC, were quasi- neutral (repetitive
peptides) irrespective of pH, or posi- tively charged
(nonrepetitive peptides) at pH below 5. The
transmission through the membranes of the nonre-
petitive peptides was less (until sevenfold) than that
of the repetitive ones, because of the role of
electrostatic repulsion involved in the retention of
charged solutes. The difference of transmission was
more efficient at acidic pH (3) and low ionic strength
with inorganic mem- branes and in a wider range of
pH and ionic strength with membranes modified by
coating of positively charged polymers
(polyvinylimidazole PVI, polyethyleneimine PEI). A
continuous diafiltration process using an inor- ganic
membrane of low molecular cut-off permitted the
selective enrichment of the retentate in nonrepetitive
peptides (up to 80%) and of the permeate in repetitive
peptides (up to 88%) from hydrolysate feed
containing about 60/40% of repetitive and
nonrepetitive peptides, respectively, with a diafiltration
volume of 4. 1999 John Wiley & Sons, Inc. Biotechnol
Bioeng 62: 649658, 1999.
Keywords: ultrafiltration; selectivity; functional
peptides; wheat; gliadins; fractionation
INTRODUCTION
Enzymatic hydrolysis is applied to plant proteins (legumi-
nous, cereals, etc.) for two purposes. When almost com-
plete, this leads to small peptides or even peptones or
amino acids; when limited, it results in polypeptides
with func- tional properties different from those of
proteins. By an appropriate choice of conditions including
substrate, en- zyme, hydrolysis medium, time, and
concentrations, hydro- lysates may be composed of
different polypeptides, or of a mixture of small peptides
and larger polypeptides (Adler-
Correspondence to: S. Berot
Contract grand sponsors: INRA; CNRS
Nissen, 1986). The hydrolysates are advantageously sepa-
rated to prepare fractions exhibiting diverse functional
prop- erties. The type of separation depends on the
characteristics of the polypeptides. For example,
chromatographic methods are very convenient when the
hydrolysates are diverse in hydrophobicity, charge or
adsorption behavior. However, these expensive techniques
ought to be reserved for high- value products. On the
contrary, membrane techniques, es- pecially ultrafiltration,
are used to separate small peptides from native proteins or
larger polypeptides. Their relative inexpensiveness enables
their application to most food products. Besides, it is
common knowledge that ultrafiltra- tion of charged solutes
is not only based on size differences, but also on protein
protein and protein membrane interac- tions (Chaufer et
al., 1991; Zeman, 1983). Unfortunately, fouling of the
membranes, by adsorption of proteins or by pore blocking,
usually causes an increase in retention (Nils- son, 1990).
Consequently, one may consider that molecules cannot be
fractionated by ultrafiltration unless their molecu- lar
weight ratios differ by an order of magnitude (Nelson,
1977).
In order to reduce fouling, low-binding membranes can
be used (Brink and Romijn, 1990). Thus, fractionations
based on size differences could be considered. On the other
hand, interactions between solutes and ultrafiltration mem-
branes could be advantageously increased to resolve the
fractionation of proteins. The retentate could be enriched in
a target protein when its retention is increased by means of
charge effects and vice versa. Nakao et al. (1988) have
shown that a protein was retained to a greater degree when
it had the same charge as a sulfonated or aminated polysul-
fone membrane; adjustment of pH to the isoelectric point of
the target protein allows its recovery in the permeate
whereas the charged protein(s) are retained. Zaksena and
Zydney (1994) with neutral membranes and Zhang and
Spencer (1993) with formed-in-place membranes used
elec- trostatic interactions to improve fractionation of
mixture of immunoglobulins and bovine serum albumin;
optimum
 1999 John Wiley & Sons, Inc.
separation was achieved at low ionic strength (when elec-
trostatic repulsion forces are the highest) and at the isoelec-
tric point of the target protein in the permeate. Thomas et
al. (1992) showed the selective enrichment of
immunoglobulin G from 8% to 20% in cheese whey by
adjustement of the pH to 5, close to the isoelectric point of
main whey proteins in a diafiltration process. Another
way for protein fraction- ation to reach purer protein in
the permeate is to use charged membranes of a charge
opposite to that of the target protein in the retentate.
Accordingly, both ionic and hydrophobic interactions of
proteins and a polyvinylimidazole coating on an inorganic
membrane were optimized for fractionation of a model
mixture (lysozyme and bovine serum albumine) (Mille
sime et al., 1996a) and whey proteins (Chaufer et al.,
1991). Therefore, the aim of the present study was to test
the ability of ultrafiltration membranes either
commercially available or custom-made for the
fractionation of polypep- tides of different charge but of
similar size.
Previously, it had been shown that the limited achymo-
tryptic hydrolysis of sulfur-rich gliadins, major wheat stor-
age proteins, depended on the type of gliadins (Legay et al.,
1997). From -gliadins, it released two types of polypep-
tides, corresponding to two different moieties of the
protein, whereas a/-gliadins were more resistant to
hydrolysis, in spite of strong sequence homologies.
Nevertheless, hydro- lysates of fractions containing about
40% of a/-gliadins had RP-HPLC patterns very similar to
those observed with purified -gliadins. The two moieties
of -gliadins are called repetitive and nonrepetitive with
reference to their amino acid sequences. The repetitive
structure is made of one or two motifs including
glutamine, proline, and aro- matic amino acids:
phenylalanine or tyrosine (Shewry et al., 1994). Contrary to
this, nonrepetitive sequences contain all sulfur amino acids
and most of the basic residues. More- over, nonrepetitive
domains are more hydrophobic than the repetitive ones
(Popineau et al., 1990). A nonrepetitive pep- tide purified
by RP-HPLC stabilized emulsions better than gliadins or
hydrolysates (Popineau and Pineau, 1993). Hence, the
fractionation of the hydrolysate is a way to ob- tain a
highly active peptide fraction. Here, we describe the
fractionation, by ultrafiltration with commercial and modi-
fied inorganic membranes, of chymotryptic hydrolysates
from a fraction enriched in -gliadins.
MATERIALS AND METHODS
Materials
Enriched -gliadin feed solution was prepared by suspend-
ing a wheat gluten fraction enriched in gliadins (Brot et
al., 1994) in ethanol / water 85/15 (v/v) for 1 h at room
tem- perature. After centrifugation the supernatant was
dialyzed and dried (Legay et al., 1997). The bovine a-
chymotrypsin (SIGMA ref. C4129) had an activity of 51
U/mg of dried matter. One unit corresponds to the
hydrolysis of 1 mole of N-Acetyl-L-Tyrosine-Ethyl Ester
per minute at pH 7.8 and 25C.
CCC 0006-3592/99/060649-10
Hydrolysis Conditions
Gliadins (25 g) were continuously stirred with5L of
buffer/ ethanol mixture (80/20 v/v) in a stirred batch reactor
(Eu- romelange Satelmix RD5, VMI-Rayneri,
Montaigu, France). The buffer was 0.0625M ammonium
bicarbonate pH 8, containing 0.0125M CaCl2. Enzymatic
hydrolysis was performed at 20C with a
chymotrypsin/protein ratio of 1: 500 (w/w). The reaction
was stopped after 24 h by a thermal treatment at 100C for
3 min (Popineau et al., 1990). After cooling, the
hydrolysates were acidified to pH 3 with 1M HCl and
centrifuged; supernatants were collected for ultra-
filtration.
Ultrafiltration
Ultrafiltration Rig
The ultrafiltration rig was comprised of a thermostated 2-L
feed tank, a volumetric feed pump (PFM 10 VRM, P.C.M.,
Vanves, France) equipped with a frequency variator (VLT,
Danfoss, Trappes, France), a single tubular membrane
(length 0.60 m, inner diameter 6 mm, membrane area A =
1.09 102 m2), three pressure gauges (P21, Philips, Bo-
bigny, France) for inlet P1 and outlet P2 of the retentate
and for the outlet P3 of the permeate, respectively, an
electro- magnetic flow meter (Picomag DN4, Endress
Hauser, Huningue, France) for measuring the retentate
tangential velocity (set at 4 m/sec to minimize the
thickness of the polarization layer), a temperature probe
(Pt 100) and a bal- ance (MC1, Sartorius, Palaiseau,
France) for measuring the permeate flow rate. The
retentate was recycled at the base of the feed tank to avoid
foam. The rig was monitored in real time using signal
transducers (6B, Analog Devices, Nor- wood, MA) and
software (Labtech Notebook, Laboratory Technologies
Corp., Wilmington, MA). The rig volume was measured as
0.35 L.
Ultrafiltration Parameters
The transmembrane pressure (TMP in Pa) is the difference
between the average pressure of the retentate and the pres-
sure of the permeate:
TMP [(P1 P2)/2] P3
From Darcys law,
Jo TMP/o Rm
where Jo = permeate flux when ultrafiltering a solvent (m 3
s1 m2), o = solvent viscosity (Pa.s) and Rm = hydraulic
resistance of the clean membrane (m1). By analogy,
J TMP/ Rt
where J = permeate flux when ultrafiltering a solute (m3 s-1
m2), = permeate viscosity (Pa.s) and Rt = hydraulic
resistance of the membrane (m1). The fouling index (FI)
expresses the decline of flux due to the solute:
650 BIOTECHNOLOGY AND BIOENGINEERING, VOL. 62, NO. 6, MARCH 20, 1999
Thank you for using www.freepdfconvert.com service!

Only two pages are converted. Please Sign Up to convert all pages.

https://www.freepdfconvert.com/membership