Milk-clotting activity of enzyme extracts from ... - Ainfo - Embrapa

International Dairy Journal 17 (2007) 816–825
Milk-clotting activity of enzyme extracts from sunflower and albizia
seeds and specific hydrolysis of bovine k-casein
Abstract
A.S. Egito a , J.-M. Girardet c , L.E. Laguna a , C. Poirson c , D. Molle´ b , L. Miclo c ,
G. Humbert c , J.-L. Gaillard c,
a Laboratório de Tecnologia de Leite, Embrapa Caprinos, Estrada Sobral-Groaíras, Km 04—Fazenda Três Lagoas—Caixa postal D-10,
CEP 62011970, Sobral, Ceará, Brazil
b Laboratoire de Science et Technologie du Lait et de l’Œuf, Institut National de la Recherche Agronomique,
65 rue de Saint Brieuc, 35042 Rennes Cedex, France
c Unité de Recherche sur l’Animal et Fonctionnalités des Produits Animaux (URAFPA), Nancy-Université, U.C. INRA 340,
Faculté des Sciences et Techniques, UHP-Nancy 1, B.P. 239, 54506 Vandœuvre-lès-Nancy Cedex, France
Received 22 May 2006; accepted 26 September 2006
Milk-clotting activity found in ammonium sulfate-precipitated protein extracts from Albizia lebbeck and Helianthus annuus seeds was
studied. Specific clotting activity of albizia seed extract was 15 times higher than that of sunflower seed extract. Zymogram analysis
revealed several proteolytic bands in albizia seed extract and one diffuse proteolytic band for sunflower seed extract. Whole bovine casein
was incubated with the plant seed extracts or chymosin and some breakdown products were characterized by reversed-phase highperformance
liquid chromatography and electrophoresis. Similar to chymosin, the two seed extracts exhibited proteolytic activity toward
k-casein, as-casein and b-casein, with the highest activity observed for the albizia seed extract. Mass spectrometry analysis showed
that the sunflower extract hydrolyzed k-casein at the Phe 105–Met 106 bond, as does chymosin. The albizia extract also displayed activity
on k-casein, but the Lys116–Thr117 bond was its preferred target.
r 2006 Elsevier Ltd. All rights reserved.
Keywords: Plant rennet; Milk-clotting activity; Bovine casein; k-casein; Albizia; Sunflower
1. Introduction
Among the vast number of proteases with applications in
the food industry, aspartic proteases such as chymosin (EC
3.4.23.4) are used for milk clotting in cheese-making. The
primary cleavage occurs at Phe105–Met106 bond of bovine
k-casein (k-CN; Jolle` s, Alais, & Jolle` s, 1963) and causes
destabilization of the casein micelles, resulting in milk
coagulation to form the cheese curd.
Milk clotting can be achieved by a number of proteolytic
enzymes from various sources, such as different animal
(pig, cow, and chicken pepsins) and microbial species
(Rhizomucor miehei, R. pusillus and Cryphonectria parasitica).
Plant coagulants are of growing interest, as the use
Corresponding author. Tel.: +33 383 68 42 66; fax: +33 383 68 42 74.
E-mail address: jean-luc.gaillard@scbiol.uhp-nancy.fr (J.-L. Gaillard).
0958-6946/$ - see front matter r 2006 Elsevier Ltd. All rights reserved.
doi:10.1016/j.idairyj.2006.09.012
ARTICLE IN PRESS
www.elsevier.com/locate/idairyj
of animal rennet may be limited for religious reasons (e.g.,
Judaism and Islam), diet (vegetarianism), or consumer
concern regarding genetically engineered foods (e.g.,
Germany, Netherlands and France forbid the use of
recombinant calf rennet). More recently, the incidence of
bovine spongiform encephalopathy has reduced both
supply and demand for bovine rennet (Roseiro, Barbosa,
Ames, & Wilbey, 2003).
Plant sources for milk-clotting enzymes have been
identified from Ananas comosus (Cattaneo, Nigro, Messina,
& Giangiacomo, 1994), Calotropis procera (Sanni,
Onilude, & Momoh, 1999), Opuntia phylloclades, Cereus
triangularis, Euphorbia caducifolia, Ficus bengalensis,
F. elastica, E. hista (Umar Dahot, Yakoub Khan, &
Memon, 1990), Lactuca sativa (Lo Piero, Puglisi, &
Petrone, 2002), seven papilionoideae species (Eriosema
shirense, E. ellipticum, E. pauciflorum, E. gossweilleri,

E. psoraleoides, Adenolichos anchietae and Droogmansia
megalantha; Lopes, Teixeira, Liberato, Pais, & Clemente,
1998), the cardoons Cynara scolymus (Sidrach, Garcia-
Canovas, Tudela, & Rodriguez-Lopez, 2005) and
C. cardunculus (Sousa & Malcata, 2002), and Helianthus
annuus (Park, Yamanaka, Mikkonen, Kusakabe, &
Kobayashi, 2000). Unfortunately, most of these plant
rennets have been found to be unsuitable because they
produce extremely bitter cheeses. An exception to this
general rule is the aqueous extracts of cardoons. Extracts of
Cynara are used chiefly in the making of various Spanish
cheeses, e.g., Torta del Casar, La Serena, Los Pedroches,
Los Ibores, Flor de Guı´a, and Portuguese cheeses from
sheep’s milk, e.g., Serra da Estrela, Serpa, Azeita˜o, Nisa,
Castelo Branco, Évora (Roseiro et al., 2003).
The flower of C. cardunculus contains aspartic proteases,
cardosin A (GenBank accession No. CAB40134; Faro et al.,
1999), the most abundant; and cardosin B (No. CAB40349;
Vieira et al., 2001). Cardosin A has been studied in detail
(Verissimo, Esteves, Faro, & Pires, 1995) and was shown to
cleave bovine k-CN at the same peptide bond, Phe 105–
Met 106, as chymosin. Cardosin B, in comparison, is similar to
pepsin, in terms of specificity and activity. Aspartic proteases
have been found in the flower cells of C. cardunculus and of
C. scolymus by other authors and named cyprosins A and B
and cynarases A, B, and C, respectively (Cordeiro, Xue,
Pietrzak, Pais, & Brodelius, 1994; Sidrach et al., 2005).
An aspartic protease from sunflower seeds displaying a
milk-clotting activity has been identified and its primary
structure has been deduced from gene sequence as No.
AB025359 (Park et al., 2000); a comparison with the
sequence of a cynarase of C. cardunculus (No. X69193;
Cordeiro et al., 1994) shows 78% identity with the
sunflower aspartic protease (Park et al., 2000). However,
the sunflower enzyme displays a negligible value of milkclotting
activity, whereas the cynarase has high milkclotting
activity (Park et al., 2000). To the best of our
knowledge, the specific action of sunflower proteolytic
enzymes toward caseins is not known.
Similarly, the seeds of the tree Albizia julibrissin have been
shown to possess proteolytic enzymes which clotted milk
readily, without developing any bitterness in cheese after 3
months of ripening (Otani, Matsumori, & Hosono, 1991).
Surprisingly, no other work has been performed to study
more extensively the clotting activity of any Albizia species.
The aim of the present work was to study the potential
ability of protein extracts from A. lebbeck and H. annuus
seed to coagulate milk and to determine the action of these
milk-clotting plant extracts on bovine whole casein and, in
particular, k-CN.
2. Materials and methods
2.1. Preparation of crude and protein extracts
Dried seeds of A. lebbeck and H. annuus (variety
EMBRAPA 122-V2000) were obtained in experimental
ARTICLE IN PRESS
A.S. Egito et al. / International Dairy Journal 17 (2007) 816–825 817
farms of the Brazilian Agricultural Research Co. (EM-
BRAPA) located in the regions of Sobral and Londrina,
respectively. Ten grams of peeled sunflower seeds and of
whole albizia seeds were ground in a coffee grinder, and
aqueous extracts were prepared by soaking the seed
powders in 100 mL of distilled water containing 1% (w/v)
NaCl and 0.02% (w/v) sodium azide. The aqueous
mixtures were maintained for 24 h at 4 1C with agitation,
and then the samples were filtered to give crude extracts.
Proteins were precipitated from the crude extracts by
using ammonium sulfate at 40% saturation, and the
mixture was kept at 4 1C for 45 min before centrifugation
(15,000 g at 4 1C for 10 min). The pellets were discarded,
and ammonium sulfate was added to the supernatants to
reach 60% saturation in the case of sunflower and 70% in
the case of albizia. After 45 min of incubation at 4 1C, the
mixtures were again centrifuged (15,000 g at 4 1C for
10 min). The pellets were dissolved in 20 mL of pure water,
dialyzed for 48 h at 4 1C to remove salts, and finally freezedried
to give protein extracts of sunflower and albizia
seeds, respectively.
2.2. Milk-clotting experiments
The clotting activities of plant extracts were determined
according to the method of Berridge (1952). Crude and
protein extracts were dissolved at 20 mg mL 1 in 10 mM
CaCl2, and the clotting time was measured using 100 mL of
each solution mixed with 1 mL of reconstituted milk (12%,
w/v, commercial low-heat skim milk powder at pH 6.5
dissolved in 10 mM CaCl2; Re´gilait, Saint-Martin-
Belle-Roche, France) and incubated at 37 1C until milk
clotting occurred. One unit (1 U) was defined as being the
quantity (mg) of crude or protein extract needed to coagulate
1 mL of reconstituted skim milk powder in 1 min at 37 1C.
2.3. Zymogram analysis
Enzyme activities of plant extracts were detected by
zymography, adapted from the method of Dib, Chobert,
Dalgalarrondo, Barbier, and Haertle´ (1998). A quantity of
3 mg of each of plant crude extract or protein extract or
chymosin was added to 1 mL of 0.125 M Tris-HCl buffer,
pH 6.8, containing 5% (w/v) SDS, 1% (w/v) sucrose, and
0.05% (w/v) bromophenol blue. A volume of 10 mL of each
solution was loaded onto SDS-PAGE gels containing 0.1%
(w/v) gelatin. Electrophoresis (SDS-PAGE) was performed
with a 4.9% (w/v) polyacrylamide stacking gel in 0.125 M
Tris-HCl buffer, pH 6.8 and with a 15.4% (w/v)
polyacrylamide resolving gel in 0.38 M Tris-HCl buffer,
pH 8.8 containing 0.1% (w/v) SDS, at 4 1C for 150 min at
500 V, 60 mA, and 30 W (Laemmli & Favre, 1973). After
electrophoretic migration, the gel was washed two times
with 2% (v/v) Triton X-100 for 30 min. The hydrolysis
reaction then proceeded inside the gel during incubation at
37 1C for 48 h in a bath of 0.05 M Tris-HCl buffer, pH 7.5,
containing 15 mM CaCl2. The active enzymes were revealed

818
as translucent bands after incubation of the gel, first in a
mixture of 40% (v/v) ethanol, 10% (v/v) acetic acid, and
0.1% (w/v) R-250 Coomassie blue for 60 min, and second
in a destaining solution containing 30% (w/v) ethanol and
7.5% (v/v) acetic acid with several washings.
2.4. Preparation of bovine whole casein
Raw milk was obtained from a local dairy herd of
Prim’Holstein cows (experimental farm of La Bouzule,
Institut National Polytechnique de Lorraine, Vandœuvrele`
s-Nancy, France) and immediately stored at 20 1C until
used. The milk was skimmed by centrifugation (2100 g at
32 1C for 30 min) and the whole casein was prepared by
isoelectric precipitation at pH 4.6 with 1 M HCl. The
precipitate was washed three times with pure water,
solubilized at pH 7.0 by addition of 1 M NaOH, and the
precipitation–solubilization cycle was repeated twice.
Finally, the whole casein was solubilized at pH 7.0 with
1 M NaOH, dialyzed against pure water at 5 1C and freezedried.
2.5. Hydrolysis of casein by chymosin, albizia or sunflower
seed protein extracts
Whole casein and commercial k-CN purchased from
Sigma Chemical Co. (St. Louis, MO, USA) were dissolved
at 10 mg mL 1 in 100 mM sodium phosphate buffer, pH 6.5,
containing 0.02% (w/v) sodium azide. Commercial chymosin
from calf stomach (Sigma Chemical Co., St. Louis,
MO, USA), albizia seed protein extract, or sunflower seed
protein extract were added (2.5 10 3 UmL 1 ) to each
protein solution and hydrolyzes were carried out at 37 1C,
with aliquots being removed at times from 1 min to 24 h.
For SDS-PAGE analysis, 2.5 10 2 UmL 1 instead of
2.5 10 3 UmL 1 were used in the case of the sunflower
protein extract to stain efficiently the electrophoretic bands
of the breakdown products. For the electrophoretic
analysis, 300 mL of 0.125 M Tris-HCl buffer, pH 6.8,
containing 0.1% (w/v) SDS, 5% (v/v) 2-mercaptoethanol,
10% (v/v) glycerol, and 0.01% (w/v) bromophenol blue
were added to 100 mL of each hydrolysate solution. The
latter was then boiled at 100 1C for 3 min before electrophoretic
analysis. For reversed-phase HPLC analysis, the
hydrolysate solutions were heated at 100 1C for 10 min to
stop the proteolytic reaction, and 400 mL of 5% (v/v)
acetonitrile in the presence of 0.4% (v/v) trifluoroacetic
acid were added to 100 mL of each hydrolysate solution to
lower the pH to 2.0. The solutions were either directly
stored at 20 1C for HPLC analysis or freeze-dried and
stored at 20 1C for mass spectrometry analysis.
2.6. HPLC and SDS-PAGE analysis of casein hydrolysis by
vegetable seed extracts and chymosin
Reversed-phase HPLC was carried out using a Lichro-
Cart C18 column (250 4 mm internal diameter, 5-mm
ARTICLE IN PRESS
A.S. Egito et al. / International Dairy Journal 17 (2007) 816–825
particle size, 10-nm porosity; Merck, Darmstadt, Germany)
connected to a HPLC model Alliance 2690 (Waters,
Milford, MA, USA). Bovine glycomacropeptide (GMP),
k-CN (both purchased from Sigma) and whole casein, used
as standards, were dissolved at 10 mg mL 1 in 100 mM
sodium phosphate buffer, pH 6.5, containing 0.02% (w/v)
sodium azide, and four volumes of 5% (v/v) acetonitrile
containing 0.4% (v/v) trifluoroacetic acid were added to
one volume of each protein solution. Volumes of 500 mL of
protein or hydrolysate solutions (corresponding to 1 mg
protein) were loaded onto the C18 column. A linear gradient
from 5% to 50% (v/v) acetonitrile in the presence of 0.1%
(v/v) trifluoroacetic acid was applied. Detection was
performed between 200 and 310 nm with a photodiode
array detector model 996 (Waters).
The SDS-PAGE gel was prepared as described above.
Bovine GMP, used as standard, was dissolved at
2.5 mg mL 1 in 0.125 M Tris-HCl buffer, pH 6.8, containing
0.1% (w/v) SDS and 5% (v/v) 2-mercaptoethanol, 10%
(v/v) glycerol, and 0.01% (w/v) bromophenol blue and
boiled at 100 1C for 3 min. Whole casein and k-CN, used as
standards, were dissolved at 10 mg mL 1 in 100 mM sodium
phosphate buffer, pH 6.5, containing 0.02% (w/v) sodium
azide, and three volumes of 0.125 M Tris-HCl buffer, pH
6.8, containing 0.1% (w/v) SDS, 5% (v/v) 2-mercaptoethanol,
10% (v/v) glycerol, and 0.01% (w/v) bromophenol
blue were added to one volume of both casein solutions,
followed by boiling at 100 1C for 3 min. Volumes of
10 mL of protein and hydrolysate solutions were loaded
onto the gel. Electrophoresis was performed at 4 1C for
150 min at 500 V, 60 mA, and 30 W. The molecular mass
standards (Bio-Rad, Hercules, CA, USA) were myosin
(200.0 kDa), b-galactosidase (116.2 kDa), phosphorylase b
(97.4 kDa), bovine serum albumin (66.2 kDa), ovalbumin
(45.0 kDa), carbonic anhydrase (31.0 kDa), trypsin inhibitor
(21.5 kDa), lysozyme (14.5 kDa, but apparent molecular
mass of 15.5 kDa), and aprotinin (6.5 kDa). After
migration, proteins or peptides were fixed with 12% (w/v)
trichloroacetic acid for 30 min and then stained for 60 min
by 0.5% (w/v) R-250 Coomassie blue dissolved in a
mixture of 50% (v/v) ethanol and 12% (w/v) trichloroacetic
acid, followed by overnight destaining in solution of
30% (v/v) ethanol, 7.5% (v/v) acetic acid and 5% (w/v)
trichloroacetic acid.
2.7. Determination of molecular mass of peptides
The molecular mass determination of bovine commercial
k-CN and of k-CN hydrolysate peptides was performed by
on-line liquid chromatography onto a Vydac C4 column
(150 2.1 mm, 5 mm particle size; 30 nm porosity; Cluzeau,
Sainte Foy La Grande, France) coupled to electrospray
source ionization mass spectrometry (LC/ESI-MS), as
previously described (Gagnaire, Pierre, Molle´, &Le´onil,
1996). The whole k-CN and its peptide hydrolysates (1 mg)
were dissolved in 500 mL of Tris-HCl 10 mM, pH 8.0,
containing 8 M urea, 40 mM trisodium citrate, and 20 mM

dithiotreitol and the solutions were incubated at 37 1C for
2 h. Volumes of 25 mL of each sample were then loaded
onto the C4 column. Analytical reversed-phase HPLC was
carried out using Agilent HP1100 chromatographic system
(Agilent Technologies, Massy, France) and elution was
obtained with a 12–64% gradient of acetonitrile in 0.1%
trifluoroacetic acid for 20 min at flow rate of 0.25 mL min 1
at 40 1C. Mass spectra were recorded on a PE-Sciex API
III + triple quadrupole mass spectrometer (Sciex, Thornhill,
Ont., Canada) equipped with an API electrospray
source and Q3 quadrupole and scanned in mass-resolving
mode over a m/z 700–2400 Da with a step size of 0.3 Da
and a dwell time of 0.5 ms.
3. Results and discussion
Albizia and sunflower seed crude extracts and their
corresponding ammonium sulfate-precipitated protein extracts
exhibited milk-clotting activity, suggesting that the
two plants possessed one or more enzymes with rennet-like
activity (Table 1). In the case of the sunflower seed crude
and protein extracts, low specific milk-clotting activities
(5.8 10 3 and 39 10 3 Umg 1 , respectively; Table 1)
were found. Indeed, Park et al. (2000) reported that
the purified aspartic protease from H. annuus, variety
IS-3311, showed almost negligible milk-clotting activity
(20 10 3 Umg 1 , determined at 35 1C and pH 6.0). The
albizia seed crude and protein extracts displayed specific
milk-clotting activities of 156 and 591 10 3 Umg 1 ,
respectively (Table 1), which were much higher than those
of the corresponding sunflower seed extracts. This result
might be related to enzyme activity of the plant seed
extracts toward gelatin (Fig. 1), which was greater for the
albizia seed protein extract than for the sunflower seed
extract. Chymosin, used as the standard, exhibited one
single proteolytic band at ca. 48 kDa, a value corresponding
to the apparent molecular mass of this protease. The
apparent molecular mass of the sunflower proteolytic band
ARTICLE IN PRESS
(ca. 110 kDa) was high, compared with that of the
sunflower aspartic protease cloned and identified by Park
et al. (2000). The latter displayed a molecular mass of
47.6 kDa (mass of the mature enzyme consisting of 440
amino acid residues). The difference between the apparent
molecular masses of the two forms might be due to the fact
that the proteolytic band present might correspond to a
dimeric state, which the non-denaturing conditions used
for the zymogram analysis did not dissociate.
Proteolysis of caseins by the albizia and sunflower seed
protein extracts and by chymosin during incubation was
Table 1
Milk-clotting activity (mean7standard deviation, n ¼ 3) of crude extracts and ammonium sulfate-precipitated protein extracts of sunflower (Helianthus
annuus) and albizia (Albizia lebbeck) seeds
Plant Total protein (mg) Total activity (U) a
Specific milk-clotting
activity 10 3 (U mg 1 )
Yield b (%) Purification factor c
Albizia seed
Crude extract 2022.5 313713 15676 — —
Protein extract d
263.9 15577 591730 49.5 3.8
Sunflower seed
Crude extract 978.7 5.770.1 5.870.1 — —
Protein extract e
100.8 3.970.2 3972 68.4 6.7
a
A unit (U) equals the amount (mg) of crude or protein extract needed to coagulate 1 mL of reconstituted skim milk in 1 min at 37 1C and pH 6.5.
b
(Total activity of the protein extract/total activity of the crude extract) 100.
c
Specific milk-clotting activity of the protein extract/specific milk-clotting activity of the crude extract.
d Obtained with 40–70% ammonium sulfate.
e Obtained with 40–60% ammonium sulfate.
A.S. Egito et al. / International Dairy Journal 17 (2007) 816–825 819
Fig. 1. Gelatin zymogram analysis performed by sodium dodecyl sulfate
polyacrylamide gel electrophoresis of chymosin (Chy), albizia and
sunflower seed crude extracts (Alb1 and Sun1), and albizia and sunflower
seed protein extracts (Alb2 and Sun2).

820
ARTICLE IN PRESS
A.S. Egito et al. / International Dairy Journal 17 (2007) 816–825
Fig. 2. Sodium dodecyl sulfate polyacrylamide gel electrophoresis of bovine sodium caseinate (10 mg mL 1 ) hydrolyzed as a function of time at pH 6.5
and 37 1C by chymosin (A; 2.5 10 3 UmL 1 ), sunflower seed protein extract (B; 2.5 10 2 UmL 1 ), or albizia seed protein extract
(C; 2.5 10 3 UmL 1 ). CN: bovine sodium caseinate; as-CN: as1-+as2-caseins; b-CN: b-casein; k-CN: k-casein; para-k-CN: para-k-casein;
Std: molecular mass standards.

studied by SDS-PAGE (Fig. 2) and reversed-phase HPLC
(Fig. 3). It was noteworthy that the three main casein
components, as-CN, b-CN, and k-CN, were more sensitive
to the action of the albizia seed protein extract than toward
ARTICLE IN PRESS
A.S. Egito et al. / International Dairy Journal 17 (2007) 816–825 821
the action of the sunflower seed protein extract and of
chymosin (Fig. 3). The order of hydrolysis obtained with
the albizia seed protein extract was the following: k-CN,
as-CN, and b-CN (Figs. 2C and 3). Most of the k-CN and
Fig. 3. Reversed-phase high performance liquid chromatography (C 18 column) of bovine sodium caseinate (10 mg mL 1 ) and its hydrolysates generated
by albizia seed protein extract (2.5 10 3 UmL 1 ), sunflower seed protein extract (2.5 10 3 UmL 1 ), or chymosin (2.5 10 3 UmL 1 ) at pH 6.5, 37 1C
for different times. CN: bovine sodium caseinate; as: as1-+as2-caseins; b: b-casein; k: k-casein.

822
as-CN components disappeared in 40 min of hydrolysis
(albizia seed protein extract at 2.5 10 3 UmL 1 ),
whereas b-CN was still present after 6 h (Figs. 2C and 3).
This order of hydrolysis was similar to the order of
susceptibility of the different casein components in whole
bovine, caprine, and ovine caseins toward the action of
animal rennet (Trujillo, Guamis, & Carretero, 1997;
Pintado et al., 2001). The casein components were more
resistant to proteolysis caused by the sunflower seed
protein extract than to that caused by albizia seed protein
extract (Figs. 2B and 3). All the caseins were still present
after 6 h hydrolysis by chymosin (2.5 10 3 UmL 1 ) and
the sunflower seed protein extract (2.5 10 2 UmL 1 ),
whereas only b-CN was found when the albizia seed
protein extract (2.5 10 3 UmL 1 ) was used. Trace
amounts of as-CN and b-CN were still detected after 24 h
hydrolysis by the sunflower seed protein extract (Fig. 2B).
Chymosin mainly hydrolyzed k-CN, as expected (Fig. 2A).
For identical enzyme units (2.5 10 3 UmL 1 in all cases),
as-CN and b-CN seemed to be less susceptible to the action
of chymosin than to that of the sunflower and albizia seed
protein extracts (Fig. 3). In contrast, the plant rennet from
C. cardunculus was less proteolytic on ovine b-CN and
as-CN than the animal rennet (Sousa & Malcata, 1997).
In the case of action of chymosin on k-CN, para-k-CN
was generated as early as 1 min and it was located on
the SDS-PAGE gel as described by Trujillo et al. (1997) at
Fig. 4. Sodium dodecyl sulfate polyacrylamide gel electrophoresis of
commercial bovine k-casein (k-CN at 10 mg mL 1 ) hydrolyzed by chymosin
(C; 2.5 10 3 UmL 1 ), albizia seed protein extract (A; 2.5 10 3 UmL 1 ),
or sunflower seed protein extract (S; 2.5 10 2 UmL 1 )atpH6.5,371Cfor
1 h. CN: bovine sodium caseinate; a s-CN: a s1-+a s2-caseins; b-CN: b-casein;
k-CN: k-casein; para-k-CN: para-k-casein; GMP: glycomacropeptide;
aGMP: aglycomacropeptide; Std: molecular mass standards.
ARTICLE IN PRESS
A.S. Egito et al. / International Dairy Journal 17 (2007) 816–825
ca. 16 kDa (Fig. 2A). This breakdown product was still
present after 24 h of hydrolysis. A similar band at ca.
16 kDa was present in the case of hydrolysis of casein by
the sunflower seed protein extract, and might correspond
to a para-k-CN-like component (Fig. 2B). With the albizia
seed protein extract, a similar band was detected only after
20 min of hydrolysis and remained at low amount even
after 24 h of hydrolysis (Fig. 2C). However, another band
with higher apparent molecular mass of ca. 17 kDa readily
appeared from 1-min hydrolysis and was relatively
resistant toward the enzyme action, as it only partly
disappeared following a 24-h hydrolysis (Fig. 2C). The two
other major breakdown products of apparent molecular
masses of ca. 6 and 22 kDa were only found in the case of
the albizia seed protein extract, but not in the two other
cases; these two bands partly disappeared after 24 h of
hydrolysis.
Fig. 5. LC/ESI-MS chromatography profiles obtained with bovine
k-casein (k-CN at 2 mg mL 1 ) before and after hydrolysis at pH 6.5,
37 1C for 1 h by albizia seed protein extract (2.5 10 3 UmL 1 ),
sunflower seed protein extract (2.5 10 2 UmL 1 ), or chymosin
(2.5 10 3 UmL 1 ). k-CN A: k-casein variant A; para-k-CN: para-kcasein;
aGMP: aglycomacropeptide. Nomenclature of peptides was
according to Farrell et al. (2004).

To determine which major bands were the breakdown
products generated from k-CN, the latter was hydrolyzed
for 1 h at 37 1C and pH 6.5 by the three enzyme systems.
The breakdown products were characterized by SDS-
PAGE (Fig. 4) and reversed-phase HPLC onto a C4
column (Fig. 5). The 6 and 22 kDa electrophoretic bands
(and also the 25-kDa band in the case of the sunflower
extract; Fig. 2B) were not recovered when purified k-CN
instead of whole casein was used as the substrate,
suggesting that these bands were generated from casein
components other than k-CN (Fig. 4). Only the bands at
16 kDa and, in the case of Albizia, the bands at 16 and
17 kDa were found, showing that these bands actually
corresponded to k-CN peptides. Bovine GMP was detected
with difficulty, as it displays poor stainability and
abnormal, diffuse electrophoretic migration (apparent
molecular mass located between ca. 22 and 28 kDa), due
to its highly acidic polyhydroxylic nature interacting with
the gel matrix (Coolbear, Elgar, Coolbear, & Ayers, 1996).
The glycan-free form of GMP, aglycomacropeptide or
aGMP, was located as described by Coolbear et al. (1996).
The k-CN component and its hydrolysates produced by
action of the two plant seed protein extracts and by
chymosin were submitted to LC/ESI-MS analysis. In the
case of chymosin, aGMP-1P (i.e., with Ser 149 phosphorylated;
Jolle` s, Schoentgen, Alais, Fiat, & Jolle` s, 1972) and
para-k-CN were identified in HPLC fractions eluted at 15.3
and 16.5 min, respectively (Table 2 and Fig. 5). Para-k-CN
and residual k-CN were co-eluted from the C4 column. The
presence of a 16-kDa band and a k-CN band had been
observed by SDS-PAGE analysis of the fraction eluted at
16.5 min (data not shown). The present study shows that
the sunflower seed protein extract also cleaved bovine
k1-CN (i.e., the glycan-free form of k-CN) at the
Phe 105–Met 106 peptide bond, to generate aGMP-1P and
para-k-CN (Fig. 6A and Table 2). The 16-kDa band, which
was present on the electrophoretic profile (Fig. 4), probably
corresponds to para-k-CN.
ARTICLE IN PRESS
Table 2
Identification by LC/ESI-MS of the main products generated from bovine k-casein hydrolysis for 1 h by chymosin (2.5 10 3 UmL 1 ), albizia
(2.5 10 3 UmL 1 ) or sunflower (2.5 10 2 UmL 1 ) seed protein extracts
Main product a
M r found (Da) Identification b
Chymosin
para-k-CN 12,270 k-CN A (f1-105) 12,268.00
aGMP 6788 k-CN A-1P (f106-169) 6787.43
Albizia seed protein extract
para-k-CN-like peptide 13,522 k-CN A (f1-116) 13,519.51
aGMP-like peptide 5536 k-CN A-1P (f117-169) 5535.92
Sunflower seed protein extract
para-k-CN 12,270 k-CN A (f1-105) 12,268.00
aGMP 6788 k-CN A-1P (f106-169) 6787.43
a CN, casein; aGMP, aglycomacropeptide.
b Nomenclature according to Farrell et al. (2004).
c Average mass (Da).
A.S. Egito et al. / International Dairy Journal 17 (2007) 816–825 823
Theoretical M r c (Da)
The albizia seed protein extract displayed a different
behavior toward k-CN, as two bands were detected in the
16–17-kDa region, the major one at ca. 17 kDa and the
minor one at ca. 16 kDa (Fig. 4). The LC/ESI-MS analysis
of the k-CN hydrolysate generated by the albizia seed
protein extract revealed the presence of two main
fragments in fractions eluting at 15.6 and 16.2 min,
respectively. According to the primary structure of k-CN,
the first was identified as the carboxy-terminal 117–169
fragment of k-CN, carrying one phosphate group and
called k-CN-1P (f117-169), while the second was identified
as the amino-terminal f1-116 fragment, named k-CN
(f1-116). The respective MS spectra of these two k-CN
fragments are shown on Fig. 6B. The k-CN (f1-116)
peptide corresponds to the 17-kDa band. Para-k-CN
would correspond to the 16-kDa band but, unfortunately,
the LC/ESI-MS did not revealed the presence of para-k-
CN or aGMP, probably due to their very low levels.
The LC/ESI-MS analysis of k-CN did not detect the
presence of para-k-CN or GMP/aGMP or of any other
amino- and carboxy-terminal fragments.
In several countries, the use of calf rennet substitutes for
cheese-making, such as porcine pepsin A, porcine pepsine
C, Mucor miehei protease, and Endothia parasitica protease,
is common (Macedo, Faro, & Pires, 1993). Specificity
studies of these enzymes toward bovine k-CN show that
only the Phe 105–Met 106 bond was cleaved by all the
enzymes, except that of E. parasitica which only cleaved
the Ser104–Phe105 bond. This difference in cleavage site
does not seem to affect clotting (Drohse & Foltmann,
1989). The specificity of plant milk-clotting enzymes on
bovine k-CN is poorly investigated. Lettucine, a serine-like
protease from L. sativa, cleaves peptide bonds other than
the Phe 105–Met 106 bond, the Arg 97–His 98, Lys 111–Lys 112,or
Lys 112–Asn 113 bonds being putative target sites (Lo Piero
et al., 2002). On the other hand, an enzyme extract from
C. cardunculus that contains a mixture of cardosins A and
B, hydrolyzes the Phe105–Met106 bond of k-CN of bovine

824
and ovine milk; however, it cleaves preferentially caprine k-
CN at Lys116–Thr117 (Sousa & Malcata, 1998). Thus, in
addition to the Phe105–Met106 bond, the Lys116–Thr117
bond seemed to be another preferential target site for some
plant rennets such as albizia seed protease and cardosins.
4. Conclusions
Albizia seed protein extract might be a potentially
suitable substitute for animal rennet, being more active
than sunflower seed protein extract and exhibiting both
good milk-clotting and caseinolytic activity required for
cheese-ripening. As many plant rennets generate bitter
peptides, experimental cheese-making needs to be carried
out with A. lebbeck to ensure that its seed extract can lead
to cheese without bitterness, as has been already noted with
A. julibrissin.
ARTICLE IN PRESS
A.S. Egito et al. / International Dairy Journal 17 (2007) 816–825
Fig. 6. Reconstructed mass from electrospray ionization mass spectrometry (ESI-MS) of the main breakdown products generated from bovine k-casein
hydrolyzed for 1 h by (A) sunflower seed protein extract (2.5 10 3 UmL 1 ) and (B) albizia seed protein extract (2.5 10 2 UmL 1 ). k-CN A: k-casein
variant A; 1P: one phosphate residue; cps: counts per second; M: molecular mass in Da. Nomenclature of peptides was according to Farrell et al. (2004).
Acknowledgements
We thank Dr. Cla´udio Guilherme Portela de Carvalho
for the generous supply of sunflower seeds and Raphae¨l
Marenzoni, student of UHP-Nancy 1, France, for technical
assistance in the preparation of plant extracts. This work
was supported by grants of the Brazilian Agricultural
Research Co. (EMBRAPA) and of the Conseil Re´gional de
Lorraine, France.
References
Berridge, N. J. (1952). An improved method of observing the clotting of
milk containing rennin. Journal of Dairy Research, 9, 328–329.
Cattaneo, T. M. P., Nigro, F., Messina, G., & Giangiacomo, R. (1994).
Effect of an enzymatic complex from pineapple pulp on the primary
clotting phase. Milchwissenschaft, 49, 269–272.

Coolbear, K. P., Elgar, D. F., Coolbear, T., & Ayers, J. (1996).
Comparative study of methods for the isolation and purification of
bovine k-casein and its hydrolysis by chymosin. Journal of Dairy
Research, 63, 61–71.
Cordeiro, M. C., Xue, Z. T., Pietrzak, M., Pais, M. S., & Brodelius, P. E.
(1994). Isolation and characterization of an cDNA from flowers of
Cynara cardunculus encoding cyprosin (an aspartic proteinase) and its
use to study the organ-specific expression of cyprosin. Plant Molecular
Biology, 24, 733–741.
Dib, R., Chobert, J.-M., Dalgalarrondo, M., Barbier, G., & Haertle´, T.
(1998). Purification, molecular properties and specificity of a thermoactive
and thermostable proteinase from Pyrococcus abyssi, strain st
549, hyperthermophilic archaea from deep-sea hydrothermal ecosystem.
FEBS Letters, 431, 279–284.
Drohse, H. B., & Foltmann, B. (1989). Specificity of milk-clotting enzymes
towards bovine k-casein. Biochimica et Biophysica Acta, 995,
221–224.
Faro, C., Ramalho-Santos, M., Vieira, M., Mendes, A., Simo˜es, I.,
Andrade, R., et al. (1999). Cloning and characterization of cDNA
encoding cardosin A, an RGD-containing plant aspartic proteinase.
Journal of Biological Chemistry, 274, 28724–28729.
Farrell, H. M., Jr., Jimenez-Flores, R., Bleck, G. T., Brown, E. M., Butler,
J. E., Creamer, L. K., et al. (2004). Nomenclature of the proteins of
cows’ milk—Sixth revision. Journal of Dairy Science, 87, 1641–1674.
Gagnaire, V., Pierre, A., Mollé, D., & Léonil, J. (1996). Phosphopeptides
interacting with colloidal calcium phosphate isolated by tryptic
hydrolysis of bovine casein micelles. Journal of Dairy Research, 63,
405–422.
Jollès, P., Alais, C., & Jollès, J. (1963). Study of k-casein form cows.
Characterization of the linkage sensitive to the action of rennin.
Biochimica et Biophysica Acta, 69, 511–517.
Jollès, J., Schoentgen, F., Alais, C., Fiat, A. M., & Jollès, P. (1972).
Studies on the primary structure of cow k-casein. Structural features of
para-k-casein; N-terminal sequence of k-caseinoglycopeptide studied
with a sequencer. Helvetica Chimica Acta, 55, 2872–2883.
Laemmli, U. K., & Favre, M. (1973). Maturation of the head of
bacteriophage T4. I. DNA packaging events. Journal of Molecular
Biology, 80, 575–599.
Lo Piero, A. R., Puglisi, I., & Petrone, G. (2002). Characterization of
‘‘Lettucine’’, a serine-like protease from Lactuca sativa leaves, as a
novel enzyme for milk clotting. Journal of Agricultural and Food
Chemistry, 50, 2439–2443.
Lopes, A., Teixeira, G., Liberato, M. C., Pais, M. S., & Clemente, A.
(1998). New vegetal source for milk clotting enzymes. Journal of
Molecular Catalysis B: Enzymatic, 5, 63–68.
Macedo, I. Q., Faro, C. J., & Pires, E. M. (1993). Specificity and kinetics
of the milk-clotting enzyme from cardoon (Cynara cardunculus L.)
ARTICLE IN PRESS
A.S. Egito et al. / International Dairy Journal 17 (2007) 816–825 825
toward bovine k-casein. Journal of Agricultural and Food Chemistry,
41, 1537–1540.
Otani, H., Matsumori, M., & Hosono, A. (1991). Purification and some
properties of a milk clotting protease from the young seeds of Albizia
julibrissin. Animal Science and Technology, 62, 424–432.
Park, H., Yamanaka, N., Mikkonen, A., Kusakabe, I., & Kobayashi, H.
(2000). Purification and characterization of aspartic proteinase from
sunflower seeds. Bioscience, Biotechnology and Biochemistry, 64, 931–939.
Pintado, A. I., Macedo, A. C., Teixeira, G., Pais, M. S., Clemente, A., &
Malcata, F. X. (2001). Caseinolytic activity of fruit extract from
Opuntia ficus-indica on bovine, caprine, and ovine sodium caseinates.
Biotechnology Progress, 17, 643–646.
Roseiro, L. B., Barbosa, M., Ames, J. M., & Wilbey, A. (2003).
Cheesemaking with vegetable coagulants—The use of Cynara L. for
the production of ovine milk cheese. International Journal of Dairy
Technology, 56, 76–85.
Sanni, A. I., Onilude, A. A., & Momoh, M. O. (1999). Selection of starters
and a starter-mediated novel procedure for production of wara, a West
African soft cheese. International Journal of Food Science and
Technology, 34, 325–333.
Sidrach, L., Garcia-Canovas, F., Tudela, J., & Rodriguez-Lopez, J. N.
(2005). Purification of cynarase from artichoke (Cynara scolymus L.):
Enzymatic properties of cynarase A. Phytochemistry, 66, 41–49.
Sousa, M. J., & Malcata, F. X. (1997). Comparison of plant and animal
rennets in terms of microbiological, chemical, and proteolysis
characteristics of ovine cheese. Journal of Agricultural and Food
Chemistry, 45, 74–81.
Sousa, M. J., & Malcata, F. X. (1998). Proteolysis of ovine and caprine
caseins in solution by enzymatic extracts from flower of Cynara
cardunculus. Enzyme and Microbial Technology, 22, 305–314.
Sousa, M. J., & Malcata, F. X. (2002). Advances in the role of a plant
coagulant (Cynara cardunculus) in vitro and during ripening of cheeses
from several milk species. Le Lait, 82, 151–170.
Trujillo, A. J., Guamis, B., & Carretero, C. (1997). Proteolysis of goat
b-casein by calf rennet. International Dairy Journal, 7, 579–588.
Umar Dahot, M., Yakoub Khan, M., & Memon, A. N. (1990). Screening
of some Pakistani plants for milk clotting activity. Journal of Islamic
Academy of Sciences, 3, 284–286.
Verissimo, P., Esteves, C., Faro, C., & Pires, E. (1995). The vegetable
rennet of Cynara cardunculus L. contains two proteinases with
chymosin and pepsin-like specificities. Biotechnology Letters, 17,
621–626.
Vieira, M., Pissara, J., Verissimo, P., Castanheira, P., Costa, Y., Pires, E.,
et al. (2001). Molecular cloning and characterization of cDNA
encoding cardosin B, an aspartic proteinase accumulating extracellularly
in the transmitting tissue of Cynara cardunculus L. Plant
Molecular Biology, 45, 529–539.