ARTICLE IN PRESS

LWT 39 (2006) 902910

www.elsevier.com/locate/lwt

Guava seed storage protein: Fractionation and characterization

Aurea Bernardino-Nicanora, Adriana A. Scilingob,
M. Cristina Anonb,, Gloria Davila-Ortza
a
Departamento de Graduados e Investigacion en Alimentos de la Escuela Nacional de Ciencias Biologicas (IPN).
Prolongacion de Carpio y Plan de Ayala. 11340. Mexico 17, DF
b
Centro de Investigacion y Desarrollo en Criotecnologa de Alimentos (CIDCA)Facultad de Ciencias Exactas, Universidad Nacional de La Plata y
Consejo Nacional de Investigaciones Cientficas y Tecnicascalle 47 y 116 1900, La Plata, Argentina
Received 14 January 2005; accepted 6 June 2005

Abstract

In the present work, guava seed storage proteins have been fractionated and characterized. Glutelins (8690 g/100 g) and globulins
(E10 g/100 g) are the main components of the protein extract. Albumins and prolamins are minor components (E2 g/100 g). Guava
seed glutelin extracts, like rice and amaranth glutelins, are legumine-like proteins that, due to their solubility properties, have to be
extracted using extreme pH (borate buffer, pH 10, Gt-Bo; NaOH pH 12 Gt-Na), denaturing (borate buffer plus sodium dodecyl
sulfate, Gt-BoSDS) or reducing conditions (borate buffer plus 2-mercaptoethanol, Gt-BoME; borate buffer plus sodium dodecyl
sulfate and 2-mercaptoethanol, Gt-BoSDSME). The highest yield was obtained with SDS extraction, suggesting that proteins in the
seeds form aggregates stabilized mainly by non-covalent interactions. Glutelins are mainly composed of 65 and 67 kD subunits, with
a lower proportion of 55 kD subunits. These subunits are formed by disulde bond-linked polypeptides with molecular masses
4045 kD, 2227 kD and 2325 kD, respectively. The guava seeds protein isolate (GSI) exhibited a polypeptide prole very similar to
that of the glutelin fraction.
The guava seed could be an alternative source of protein for human and animal consume, additional to this to solve at least in part
the pollution problem that fruit processing industry has for discarding this material.
r 2005 Swiss Society of Food Science and Technology. Published by Elsevier Ltd. All rights reserved.

Keywords: Guava seed; Protein fractions; Glutelins

1. Introduction 2001; Liadakis, Constantine, Vassiliki, & Christos, 1995;

One of the most common problems in food processing
industry is the disposal of the sub products generated.
This waste material produces ecological problems
related to the proliferation of insects and rodents, and
an economical burden because of transportation to
repositories. Therefore, strategies for the protable use
of these materials are needed. Several studies on the use
of waste products generated by the food industry
(Bernardino-Nicanor, Ortz, Martnez, & Davila-Ortz,
Corresponding author. Tel.: +54 221 4249287;
fax: +54 221 4254853.
E-mail address: mca@biol.unlp.edu.ar (M.C. Anon).
Ravindran & Sivakanesan, 1996), have shown that these
products are an alternative source of oil and protein for
human and animal feeding. Examples of such products
are tomato seeds (Liadakis et al., 1995) and sesame seeds
(Adawy, 1997). Guava fruit, usually consumed in
Mexico, belongs to a dicotyledoneous family. The pulp
(88 g/100 g of fruit weight) is used for juice production,
but seeds (12 g/100 g of fruit weight) are discarded. The
in vitro digestibility of storage proteins of guava seeds is
higher than that of the soybean isolate (94.8 g/100 g vs.
89.9 g/100 g). Except for lysine content, the essential
amino acid prole is above that recommended in the
FAO/WHO (1985) pattern for adults (Bernardino-
Nicanor et al., 2001).

0023-6438/$30.00 r 2005 Swiss Society of Food Science and Technology. Published by Elsevier Ltd. All rights reserved.
doi:10.1016/j.lwt.2005.06.006
 ARTICLE IN PRESS
A. Bernardino-Nicanor et al. / LWT 39 (2006) 902910
Seed storage proteins were initially classied accord-
ing to their solubility properties into albumins (water
soluble), globulins (saline soluble), prolamins (alcohol
soluble), and glutelins (residue) (Osborne, 1924). Based
on more recent and extensive molecular and biochemical
analysis of storage proteins and their genes, these
proteins have been classied into two major groups,
globulins and prolamins (Shewry & Casey, 1999, Chaps.
2, 6, 13, 22, 23, and 24). According to this classication,
the guava seed proteins could be considered globulins,
like most of dicotyledonous storage proteins.
The objective of the present work was to obtain
protein fractions of the guava seed and deter-
mine partially their chemical and molecular character-
istics. Another goal was to use these data to locate the
guava seed storage proteins within the proposed
classications.
2. Materials and methods
2.1. Materials
Guava pomace was obtained from a guava processing
plant (Boing industry located in Queretaro, Mexico). It
was sundried (2030 1C, 3 days). The skins were
removed using a 1 mm sieve. The seeds were pulverized
in a stone mill and passed through a 0.5 mm sieve,
producing guava seeds meal (GSM). Defatted GSM
(GSMd) was obtained by treatment with anhydrous
ether in a Soxhlet apparatus (AOAC, 1995).
2.2. Quantification of macrocomponents
The protein content of meals, isolate and different
fractions was determined by the Kjeldahl method
(AOAC, 1995). The protein/nitrogen coefcient used
was 6.25 (Method 955.04). The proximate analysis was
completed with crude ber (Method 962.09), crude fat
(Method 920.39), moisture (Method 934.01) and ash
(Method 923.03) (AOAC, 1995).
2.3. Protein isolate (GSI)
The method of Liadakis et al. (1995) (as modied by
Bernardino-Nicanor et al. (2001)), was used. The
meal:water ratio was 1:20. The pH of the suspension
(11.5) was kept constant during the extraction procedure
by the addition of 0.1 mol/l NaOH. The temperature
(40 1C) was regulated with a water bath. After 30 min,
the slurry was centrifuged at 2600g for 30 min at 4 1C.
The supernatant was collected and the pH was adjusted
to its isoelectric point (pH 5.0) using 0.1 mol/l HCl. The
protein precipitate was separated by centrifugation at
2600g for 30 min at 4 1C and freeze-dried.
903
2.4. Guava protein fractions
2.4.1. Obtained by Osbornes method (Osborne, 1924)
Albumins: Defatted guava meal, GSMd, was rst
extracted with distilled water (0.1 g/ml) by two stirring
steps of 1 h at 4 1C and then centrifuged at 10 000g for
30 min at 4 1C. Supernatant was freeze-dried and called
AlbO.
Globulins: The residue from albumins extraction was
extracted under magnetic stirring for 1 h with NaCl
(10 g/100 g) at 4 1C and centrifuged at 10 000g for 30 min
at 4 1C. Supernatant was dialysed against distilled water
for 5 days, changing water dialysis every day and freeze-
dried. It was called GlbO.
Prolamins: The residue resulting from globulin
extraction was extracted under magnetic stirring
for 1 h at 4 1C with 70 ml/ 100 ml aqueous 2-propanol
and centrifuged at 10 000g for 30 min at 4 1C. Super-
natant was dialysed against acetic acid (1 ml/ 100 ml) for
5 days, with daily changes of solution, and freeze-dried
(ProO).
Glutelins: After prolamins were obtained by the
method of Osborne (Osborne, 1924), the glutelin
fraction was extracted with one of the following
extracting agents: (a) Na2B4O7 (Gt-Bo), (b) Na2-
B4O7+SDS (sodium dodecyl sulfate) (Gt-BoSDS), (c)
Na2B4O7+2-ME (2-mercaptoethanol) (Gt-BoME), (d)
Na2B4O7 +SDS+2-ME (Gt-BoSDSME), all at pH 10,
and (e) NaOH (Gt-Na), pH 12. Samples were suspended
in the buffer solutions by magnetic stirring during 1 h
and centrifuged at 10 000g for 30 min at 4 1C. Super-
natants were dialyzed against acetic acid (1 ml/100 ml)
for 5 days, with daily changes of solution, and freeze-
dried. The Na2B4O7 and NaOH concentrations were
0.1 mol/l, SDS concentration was 1 g/100 ml and that of
2-ME was 0.6 ml/100 ml.
2.4.2. Obtained by Barba de la Rosas method (Barba De
La Rosa, Paredes-Lopez, & Gueguen, 1992)
Albumins+Globulins: A suspension of our in 0.1
mol/l Na2B4O7 pH 7.0 (0.1 g/ml) was stirred for 1 h
at room temperature and centrifuged at 10 000g for
30 min at 4 1C. Then, a second extraction was done
with the same reagent. Supernatants were collected
and dialysed at 4 1C against deionized water for
5 days; with daily changes of water. The content of
dialysis tubes was centrifuged at 10 000g for 30 min
at 4 1C. Supernatant was albumin fraction (AlbBR)
and pellet was globulin fraction (GlbBR). Both were
freeze-dried.
Prolamins: The residue of albumins and globulins
extraction was mixed with 70 ml/100 ml aqueous 2-
propanol for 1 h at 4 1C and centrifuged at 10 000g for
30 min at 4 1C. Supernatant was dialysed against acetic
acid (1 ml/100 ml) for 5 days, with daily changes of
solution, and freeze-dried.
 ARTICLE IN PRESS
904 A. Bernardino-Nicanor et al. / LWT 39 (2006) 902910
2.5. Differential scanning calorimetry (DSC)
Runs were performed in a Polymer Laboratories
(Rheometric Scientic Ltd., UK) calorimeter using Plus
V 5.41 software. Calibration was carried out at a heating
rate of 10 1C/min using indium.
For runs, 20 g/100 ml suspensions of protein (glutelin
fractions) were prepared in distilled water. DSC samples
consisted of hermetically sealed aluminum pans lled
with 1214 mg of suspensions. They were run at a rate of
10 1C/min from 20 to 150 1C, and a double empty pan
was used as a reference. The denaturation parameters
were calculated with the equipment software. The cell
constant and temperature calibrations were performed
according to ASTM Norm E 967-83 (ASTM, 1984) and
E 968-83 (ASTM, 1984), respectively, using indium
thermograms.
2.6. Monodimensional and bidimensional electrophoresis
SDSpolyacrylamide gel electrophoresis
(SDSPAGE) was performed in 10 g/100 ml separating
gels with 4 g/100 ml stacking gels according to the
method of Laemmli (1970), using the Mini Protean 3
Cell (Bio-Rad Laboratories, Hercules, CA 94547 USA)
vertical unit. Molecular masses of the polypeptides were
calculated using the following standard proteins (Bio-
Rad Laboratories Hercules, CA 94547, USA): phos-
phorylase b (94 kD), bovine serum albumin (67 kD),
ovalbumin (45 kD), carbonic anhydrase (30 kD), trypsin
inhibitor (20.1 kD), a-lactalbumin (14.4 kD). Protein
samples were dissolved in sample buffer (0.1 mol/l
TrisHCl, pH 6.8, 20 ml/100 ml glycerol, 2 g/100 ml
SDS, and 0.05 g/100 ml bromophenol blue). For redu-
cing conditions, 5 ml/ 100 ml 2-mercaptoethanol (2-ME)
was added, and samples were heated (100 1C, 5 min).
Gels were xed and stained with Coomassie Brillant
Blue.
In order to run bidimensional electrophoresis, the rst
electrophoresis dimension was performed under dena-
turing conditions in a 10 g/100 ml polyacrylamide gel.
Samples were prepared in the same way as for
monodimensional electrophoresis. After the run, a lane
cut from the rst dimension gel was treated with
reducing buffer (0.0625 mol/l TrisHCl, pH 6.8, 1 g/
100 ml SDS, 20 g/100 ml sucrose, 0.2 mol/l 2-ME) at
55 1C for 30 min, with at least two solution changes. The
second dimension run was performed in a 12 g/100 ml
polyacrylamide gel with a stacking gel of 4 g/100 ml
polyacrylamide.
2.7. Isoelectric focusing
Isoelectric focusing of glutelin fractions was per-
formed according to Bollag & Edelstein (1991, Chaps. 6
and 7) in a 5.1 g/100 ml polyacrylamide gel (2.4 g/100 ml
carrier ampholytes, pH 39, 0.16 ml/100 ml TEMED,
0.1 ml/100 ml glycerol). Ten milligrams of protein were
focused for 1.5 h/200 V and 2 h/400 V. Standards used
were phycocyanin (three bands of pI 4.45, 4.65 and
4.75), b lactoglobulin B (one band of pI 5.1), bovine
carbonic anhydrase (one band of pI 6.0), human
carbonic anhydrase pI 6.5, equine myoglobin (two
bands of pI 6.8 and 7.0), human hemoglobin A (one
band of pI 7.1), human hemoglobin C (one band of pI
7.5), lentil lectin (three bands of pI 7.8, 8.0 and 8.2) and
Cytochrome C (one band of pI 9.6).
Gels were stained with Coomassie Brilliant Blue R-
250.
3. Results and discussion
3.1. Proximate analysis
The percent composition of the GSM on a dry weight
basis was determined. The main component was raw
ber (72.170.1 g/100 g), followed by lipids (12.570.5 g/
100 g), proteins (7.270.1 g/100 g), carbohydrates
(6.870.1 g/100 g) and ashes (1.5070.05 g/100 g) (values
in parenthesis indicate the mean and standard deviation
of three replicates).
Since our goal was to obtain a protein extract, GSM
was defatted (GSMd) to lipid levels lower than 1 g/100 g
producing an increase of the percent content of the
remaining components.
The protein content of guava seeds (7.2 g/100 g) was
lower than that of legume seeds (1944 g/100 g) (Lam-
part-Szczapa, 2001, Chap. 14), but similar to that
reported for cereal seeds (717 g/100 g) (Segura-Nieto,
Barba De La Rosa, & Paredes-Lopez, 1994, Chap. 5).
3.2. Protein fractionation of the guava seed
Table 1 shows the proportion of guava seed protein
fractions, extracted by the Osbornes (1924) method or
following the procedure of Barba de la Rosa et al.
(1992). A large proportion of the protein content
(E8690 g/100 g), corresponding to the glutelin fraction,
is obtained as insoluble residue. The remaining proteins
(E14 g/100 g) are distributed into globulins (E10 g/
100 g) and albumins and prolamins (E2 g/100 g each
one). The percent distribution of protein fractions is
very similar to that of rice (Cheftel, Cuq, & Lorient,
1992, Chaps. 1, Introduccion, and 6). These results show
that the two extraction methods used in the present
study yield similar percentages of each protein fraction.
3.3. Glutelins extraction
The insoluble residue left after extraction by the
Osbornes (1924) method was subjected to extraction
 ARTICLE IN PRESS
A. Bernardino-Nicanor et al. / LWT 39 (2006) 902910 905

with ve buffer solutions: NaOH and Na2B4O7, with
and without SDS and 2-ME. The results obtained are
presented in Table 2. The highest glutelins extraction
was obtained with Na2B4O7+SDS (81.8 g/100 g) (Gt-
BoSDS), followed by Na2B4O7+SDS+2-ME (Gt-
BoSDSME) (72.6 g/100 g). In contrast, the yield of
guava seed glutelin extract was only 59.9 g/100 g (dry
basis) with 0.1 NaOH (Gt-Na), and the 0.1 mol/l
Na2B4O7 buffer solution (Gt-Bo) produced the lowest
yield (6.8 g/100 g, dry basis). On the other hand, these
results contradict those of Barba de la Rosa et al. (1992),
who reported similar yields of amaranth glutelins
extracted with either borate buffer or 0.1 mol/l NaOH.
The higher extraction yield of Gt-Na compared to Gt-
Bo can be related to denaturation and dissociation of
protein molecules by the rst buffer, which facilitates
their extraction. In contrast, protein molecules probably
present a more conserved structure in Gt-Bo, similar to
that observed with amaranth glutelins, making extrac-
tion more difcult (Abugoch, Martnez, & Anon, 2003).
The addition of 2-ME did not improve the yield. This
may indicate that disulde bridges are not importantly
involved in the solubility of the glutelin fraction (Barba
de la Rosa et al., 1992).
Glutelin fractions were analysed by DSC. No
endotherms were recorded in the studied samples, which
may be due to the protein denaturation produced by the
pH, and the use of denaturing and reducing agents
during glutelins extraction. Similar results were obtained
Table 1
Protein fractions of the guava seed
Fraction g of protein fraction/100 g of total protein
seed
Osbornes method Barba de la Rosas
(1924) method (1992)
Albumins 2.670.1 1.5070.05
Globulins 9.4770.04 6.170.1
Prolamins 1.970.4 1.970.3
Insoluble residue E86 E90
N  6.25. Values are expressed as mean7standard deviation of three
replicates.
by Abugoch et al. (2003) during the extraction of
amaranth glutelins.
3.4. Electrophoretic characteristics of guava seed
fractions
SDSPAGE runs of albumins, globulins, prolamins
and guava isolate are shown in Fig. 1A (non-reducing
conditions) and Fig. 1B (reducing conditions).
The electrophoretic patterns of the albumin fractions
obtained by the methods of Osborne (Fig. 1A, AlbO)
and Barba de la Rosa (Fig. 1A, AlbBR) shared
polypeptides of 73, 68, 2225 and 14 kD. An additional
polypeptide of 65 kD was observed in AlbBR.
The globulin fraction obtained by the Osbornes
method (Fig. 1A, GlbO) showed less bands than that
of the globulin fraction obtained by the method of
Barba de la Rosa (Fig. 1A, GlbBR). Both fractions
contained polypeptides of 58 kD (less intense in GlbO
than in GlbBR), peptides around 30 kD (33 kD and
2829 kD), and wide ill-dened bands in the 1420 kD
range. Additional polypeptides of higher molecular
mass (94 and 90 kD) and a less intense one at 45 kD
were observed in GlbBR.
Fig. 1A also shows the electrophoretic pattern of
prolamins obtained by Osbornes method (ProO).
Several bands of molecular mass higher than 94 kD, a
polypeptide of 94 kD (of lower intensity than that of
GlbBR) and two very well-dened bands of 75 and 73 kD
are clearly seen in this lane. Additional bands of 60, 38,
24 and 14 kD are also observed.
High molecular mass aggregates, as those described
for GlbBR and ProO, were observed in the guava seed
protein isolate (Fig. 1A, GSI), together with polypep-
tides of 68 and 65 kD and bands of 58, 38, 28 and 14 kD.
Under reducing conditions (Fig. 1B), no protein was
retained in the sample well in any lane.
The electrophoretic pattern of AlbO under reducing
conditions differed from that under non-reducing
conditions, especially regarding the high molecular mass
range, where several bands of low intensity were
observed. A polypeptide of 58 kD and two less intense
bands of lower molecular mass were especially notice-
able. Two tiny bands below the 30 kD marker were also

Table 2
Guava seed glutelins extraction

Buffer solution g of soluble protein/100 g of g of protein in the residue/100 g

total protein of total protein

0.1 mol/l NaOH pH 12 59.170.5 26.970.8

0.1 mol/l Na2B4O7, pH 10 6.870.4 79.170.5
0.1 mol/l Na2B4O7+1 g/100 ml SDS, pH 10 81.870.2 4.170.2
0.1 mol/l Na2B4O7+0.6 ml/100 ml 2-ME, pH 10 25.770.5 60.170.4
0.1 mol/l Na2B4O7+1 g/100 ml SDS +0.6 ml/100 ml 2-ME pH 10 72.770.9 13.170.3

N  6.25. Values are expressed as mean7standard deviation of three replicates.

 ARTICLE IN PRESS
906 A. Bernardino-Nicanor et al. / LWT 39 (2006) 902910

Alb Glb Pro A polypeptide of 67 kD was detected in GlbO under

reducing conditions but not under non-reducing condi-
O BR O BR O GSI
S tions. No important differences were observed between
kD the electrophoretic patterns of reduced and non-reduced
GloBR. Profusely stained zone in the run front were
observed in both samples of globulins.
Treatment of ProO (Fig. 1B) with reducing agents led
94 to the disappearance of the 7572 kD doublet and to the
appearance of a new doublet formed by polypeptides of
approximately 5058 kD. Two polypeptides were ob-
67 served below the 30 kD marker.
The reduced guava protein isolate (GSI, Fig. 1B)
showed an electrophoretic pattern different from those
45
described above. Some diffuse bands of high molecular
mass were observed, together with three sharp bands
30 between 50 and 40 kD (absent in the non-reduced
sample), two proteins of 29 and 27 kD, and polypeptides
of 25, 18 and 16 kD also detected in AlbO. It is evident
20.1 that the albumins, globulins and prolamins fractions are
(A) 14.4 not the main components of GSI.
Fig. 2 depicts the electrophoretic patterns of glutelins
fractions obtained with ve different solvents. NaOH
Alb Glb Pro (Gt-Na), borate buffer (Gt-Bo), borate buffer with SDS
GSI kD (Gt-BoSDS), borate buffer with 2-ME (Gt-BoME) and
O BR O BR O S borate buffer with SDS and 2-ME (Gt-BoSDSME). The
electrophoretic patterns of the ve extracts run under
non-reducing conditions (sample buffer without 2-ME,
Fig. 2A) differed markedly. For glutelins fractions
extracted under non-reducing conditions (Fig. 2A,
94
Lanes 1, 2, and 3), rather sharp bands of high molecular
mass (higher than 94 kD) were observed (fractions Gt-
67 Na, Gt-Bo and Gt-BoSDS). These bands were not
detected when the extraction buffer included 2-ME (Fig.
45 2A, Lanes 4, and 5). All the ve patterns included
polypeptides of 65 and 67 kD, albeit with different
30 relative intensity. These bands were more intense in the
glutelins fraction obtained with borate buffer containing
SDS (Fig. 2A, Lane 3, Gt-BoSDS), but were very faint
20.1 in the patterns of glutelins extracted under reducing
conditions. The Gt-Na extract (Fig. 2A, Lane 1)
14.4 exhibited ve polypeptides of molecular masses 43, 29,
(B) 25, 22 and 1816 kD, being the ones at 25 and 22 kD the
most important. The SDSPAGE pattern of glutelins
Fig. 1. SDSPAGE of guava seeds proteins in non-reducing (A) and
extracted with borate buffer (Fig. 2A, Lane 2, Gt-Bo)
reducing (B) conditions. Fractions: albumins (Alb), globulins (Glb) included only a polypeptide at 6567 kD and proteins in
and prolamins (Pro), obtained by the methods of Osborne (O) or the run front that could correspond to the light
Barba de la Rosa (BR). GSI: guava seed protein isolate. S: molecular polypeptides between 14 and 18 kD. Glutelins extracted
mass standards. with borate buffer containing SDS included besides the
high molecular mass bands and the polypeptides at 65
and 67 kD, a band at 55 kD, two additional sharp bands
observed, together with three smaller and better dened at 43 and 27 kD, similar to that described for Gt-Na
bands of 25, 18 and 16 kD. An ill-dened electrophoretic (Fig. 2A, Lanes 3, and 1), and the frontal band
pattern was obtained with AlbBR under reducing corresponding to light polypeptides (1418 kD). The
conditions (Fig. 1B), characterized by a large spot in electrophoretic pattern of glutelins extracted with borate
the run front, probably corresponding to low molecular buffer containing 2-ME was similar to that observed for
mass proteins. glutelins extracted with borate buffer containing 2-ME
 ARTICLE IN PRESS
A. Bernardino-Nicanor et al. / LWT 39 (2006) 902910
1 2 3 4 5 S
(A)
1 2 3 4 5 S
(B)
Fig. 2. SDSPAGE of guava seeds glutelins fraction (Gt) in non-
reducing conditions (A) and reducing conditions (B). 1: Gt-Na; 2: Gt-
Bo; 3: Gt-BoSDS; 4: Gt-BoME; 5: Gt-BoSDSME. S: molecular mass
standards.
and SDS (Fig. 2A, Lanes 4, and 5), although all bands
were of low intensity. In both cases, two polypeptides
between 40 and 50 kD were observed, together with
several bands of lower molecular mass, some of which
coincided with the patterns described above. Bands at
29, 25, 22 y 1418 kD were also observed in the Gt-Na
pattern (Lanes 5 and 1), and the band a 27 kD coincided
with the Gt-BoSDS pattern (Lanes 5 and 3).
The electrophoretic patterns of the ve glutelins
fractions run under reducing conditions (2-ME in the
sample buffer, Fig. 2B) were similar among them, but
differed from those obtained in the absence of reducing
agent. These ndings agree with those of Jood,
Schoeld, Tsiami, and Bollecker (2000), who found that
glutelins are formed by high MW polymers stabilized
through disulde bridges. The pattern of the reduced
907
kD samples includes three bands between 40 and 50 kD,
other three polypeptides at 29, 27 and 25 kD, and two
bands at 16 and 14 kD. The most marked differences
were observed in lane 1, which corresponds to glutelins
94 extracted with NaOH. This fraction is the only one
exhibiting a large amount of protein of molecular mass
67 higher than 94 kD. In addition, of the three sharp bands
between 40 and 50 kD, only the lighter one predomi-
45 nates.
The common electrophoretic pattern of glutelins
30
fractions run under reducing conditions (Fig. 2B, Lanes
2, 3, 4, and 5) agrees with that obtained for the GSI run
under reducing conditions (Fig. 1B, GSI). This result
20.1
conrms that the protein isolate (GSI) is composed
14.4 mainly of glutelins. In guava seeds, this fraction would
be highly aggregated, as it happens in other seeds. These
aggregates would be mainly stabilized by non-covalent
interactions. Notwithstanding, the results indicate that
kD disulde bridges also stabilize the glutelins subunits, but
are not involved in the aggregation process. Such
conclusion is drawn from the fact that the highest yield
of glutelins corresponds to the extraction with SDS-
94 containing solvent (E80 g/100 g, Table 2). The NaOH
buffer was less suitable, since the extraction yield was
67 lower (E60 g/100 g, Table 2), the solubilized protein was
still aggregated (high amount of high molecular mass
45 proteins) and partially dissociated (low molecular mass
polypeptides shared with reduced samples), and the
30 6567 kD and 55 kD subunits that seem to constitute
glutelins were absent.
The SDSPAGE patterns under non-reducing condi-
20.1 tions of glutelins extracted in the presence of reducing
agent (Fig. 2A, Lanes 4 and 5, Gt-BoME and Gt-
14.4 BoSDSME) differed from those obtained under redu-
cing conditions (Fig. 2B, Lanes 4 and 5, Gt-BoME and
Gt-BoSDSME). This suggests that the reducing effect
exerted by 2-ME during extraction would be reverted by
the lyophilization process, which promotes a partial
reoxidation (Hoshi & Yamauchi, 1983; Wolf, 1993).
Similar to rice glutelins, the guava seed glutelins are
the major fraction of seed proteins (86 and 7080 g/
100 g, respectively), and have subunits joined by
hydrophobic interactions and disulde bridges. A
similarity of guava seeds glutelins to 11S soybean
globulins (Katsube et al., 1999) and 11S amaranth
globulins (Abugoch et al., 2003) can be suggested. This
suggestion is supported by the more recent classication
of storage proteins in globulins and prolamins (Shewry
& Casey, 1999, Chaps. 2, 6, 13, 22, 23, and 24).
In order to analyse if some protein subunits from
guava seeds were constituted by polypeptides binded by
disulde bonds, bidimensional electrophoresis under
reducing conditions was carried out. The bidimensional
electrophoretic analysis of Gt-Na (Fig. 3A) showed that
the high molecular mass aggregates of the initial portion
of gel are constituted by monomeric polypeptides, and
 ARTICLE IN PRESS
908 A. Bernardino-Nicanor et al. / LWT 39 (2006) 902910

SDS-PAGE

20.1

20.1
14.4

14.4
94

67

45

30
94

67

45

30
kD kD

94 94
67 67

45 45

30 30

20.1 20.1
14.4
14.4
(A) (B)
SDS-PAGE + 2-ME

14.4
20.1
20.1

14.4
94

67

45

30

94

67

45

30
kD kD

94
94
67
67

45
45

30 30

20.1 20.1
14.4 14.4
(C) (D)

Fig. 3. Bidimensional (SDS-SDS+2-ME, non-reducing-reducing conditions) electrophoresis of: Gt-Na (A); Gt-Bo (B); Gt-BoSDS (C); Gt-
BoSDSME (D). S: molecular mass standards.

to a lesser extent by other polypeptides released under
reducing conditions. Low molecular mass polypeptides
of 25 and 22 kD are monomeric. The 6567 kD
polypeptide detected in Gt-Bo (Fig. 3B) is cleaved by
2-ME. It appears that two bands run virtually together
in the rst dimension, since two polypeptides of
molecular mass close to 45 kD are observed in the
second dimension. Although these polypeptides are
expected to be accompanied by lighter peptides (close
to 20 kD), only a faint spot is detected in the
corresponding area of the gel. Bidimensional analysis
of Gt-BoSDS allowed conrming that the 65 and 67 kD
bands of the rst dimension are constituted by more
than one polypeptide close to 45 kD and its accompany-
ing band close to 20 kD. The analysis also showed that
the 55 kD polypeptide described in this extract is

A. Bernardino-Nicanor et al. / LWT 39 (2006) 902910
1 2 4 5
Fig. 4. Isoelectric focusing of guava seed glutelins under reducing
conditions. S: isoelectric point standards; 1: Gt-Na; 2: Gt-Bo; 3: Gt-
BoSDS; 4: Gt-BoME; 5: Gt-BoSDSME.
constituted by two polypeptides of equal or similar
molecular mass bound by disulde bridges. Similar to
that observed with Gt-Na (Fig. 3A), the 43 kD
polypeptide was not cleaved by 2-ME as were the
polypeptides of lower molecular mass. In the Gt-
BoSDSME fraction, no spot was observed under the
451 diagonal. Since this sample differs from the former
by the presence of 2-ME in the extraction solvent, it can
be inferred that the absence of 6567 kD and 55 kD
polypeptides results from the reduction of the disulde
bridges present in these proteins, which yields peptides
of molecular mass close to 45 kD and lighter peptides
between 20 and 30 kD.
3.5. Guava seed glutelin isoelectric focusing
determination
Fig. 4 shows the isoelectric focusing patterns of
glutelin fractions obtained with different extraction
solvents. The pattern corresponding to the Gt-Na
fraction (Fig. 4, Lane 1) included polypeptides with
isoelectric points different from those observed in the
patterns of the other four extracts, which were very
similar among them. All the samples exhibited polypep-
tides with pI values between 8 and 9. A low intensity
band was observed at pI 8 in all samples, except for the
glutelin extracted with NaOH (Fig. 4, Lane 1). A
polypeptide of higher intensity was present at pI 7.5 in
samples extracted with solvents containing SDS. Small
differences in the pI of some polypeptides (6.26.4,
6.06.6, 6.36.4, and 5.96.5) were observed for samples
extracted with or without SDS. Less intense bands
were present in the pI range 4.05.9, especially in Gt-Bo
ARTICLE IN PRESS
909
S (Fig. 4, Lane 2). These were the most important bands in
pI the Gt-Na fraction.
9.6
8.2
8.0 Acknowledgments
7.8
7.5
This work was partly funded by Instituto Politecnico
7.1
6.8 Nacional (I.P.N.) through de CGPI20010892 and XI.17-
6.5 CYTED projects. Aurea Bernardino Nicanor acknowl-
6.0 edges a study grant from CONACYT Mexico and a
fellowship from PIFI-IPN, and is grateful to Dra. Mara
5.1 Cristina Anon from the Universidad Nacional de La
4.8 Plata.
4.7
4.4
References
Abugoch, L., Martnez, E. N., & Anon, M. C. (2003). Inuence of
extracting solvent upon the structural properties of amaranth
(Amaranthus hypochondriacus) glutelin. Journal of Agricultural and
Food Chemistry, 51, 40604065.
Adawy, T. (1997). Effect of sesame seed protein supplementation on
the nutritional, physical, chemical and sensory properties of wheat
our bread. Food Chemistry, 59, 714.
Association of Ofcial Analytical Chemists (AOAC). (1995). Official
methods of analysis (8th ed.). Virginia, USA.
ASTM E967-83. (1984). Standard practice for temperature calibration
of differential scanning calorimeters and differential thermal
analyzers. Annual Book of standards, 14(02), 782.
ASTM E968-83. (1984). Standard practice for heat ow calibration of
differential scanning calorimeters. Annual Book of Standards,
14(02), 788.
Barba De La Rosa, A., Paredes-Lopez, O., & Gueguen, J. (1992).
Characterization of amaranth globulins by ultracentrifugation and
chromatographic techniques. Journal of Agricultural and Food
Chemistry, 40, 937940.
Bernardino-Nicanor, A., Ortz, M., Martnez, A., & Davila-Ortz, G.
(2001). Guava seed protein isolate: Functional and nutritional
characterization. Journal of Food Biochemistry, 25, 7689.
Bollang, D.M., & Edelstein, S.J. (1991). Protein methods (pp. 365369).
New York: Wiley-Liss.
Cheftel, J., Cuq, J., & Lorient, D. (1992). Los principales sistemas
proteicos alimenticios. In Proteinas alimentarias (pp. 13, 235250).
Zaragoza, Espana: Acribia Ed.
FAO/WHO. (1985). Application of risk analysis to food standards
issues. Report FAO/WHO expert consultation, Geneva, Switzer-
land.
Hoshi, Y., & Yamauchi, F. (1983). Determination of sulfhydryl and
disulde contents of soybean 11S globulin and their change by
lyophylization. Agricultural and Biological Chemistry, 47,
24352440.
Jood, S., Schoeld, J., Tsiami, A., & Bollecker, S. (2000). Effect of
composition of glutenin subfractions on rheological properties of
wheat. Journal of Food Biochemistry, 24, 275298.
Katsube, T., Kurisaka, N., Ogawa, M., Maruyama, N., Ohtsuka, R.,
Utsumi, S., et al. (1999). Accumulation of soybean glycinin and its
assembly with the glutelins in rice. Plant Physiology, 120,
10631073.
Laemmli, U. (1970). Cleavage of structural proteins during the
assembly of the head of bacteriophage T4. Nature, 227, 680685.
Lampart-Szczapa, E. (2001). Legume and oilseed proteins. In Z. E.
Sikorski (Ed.), Chemical and Functional Properties of Food Proteins
(pp. 407434). New York: CRC Press.
 ARTICLE IN PRESS
910 A. Bernardino-Nicanor et al. / LWT 39 (2006) 902910
Liadakis, G., Constantine, T., Vassiliki, O., & Christos, D. (1995).
Protein isolation of tomato seed meal; extraction optimization.
Journal of Food Science, 60, 477482.
Osborne, T. B. (1924). The vegetable proteins (2nd ed). New York:
Longmans, Green.
Ravindran, V., & Sivakanesan, R. (1996). The nutritive value of
mango seed kernels. Journal of the Science and Food Agriculture,
71, 245250.
Segura-Nieto, M., Barba De La Rosa, P., & Paredes-Lopez, O.
(1994). Biochemistry of amaranth proteins. In Amaranth.
Biology, chemistry and technology (pp. 75106). CRC. Press.
Inc. Ed.
Shewry, R., & Casey, R. (1999). Seed proteins. Klower Academic.
Wolf, W. (1993). Sulfhydryl content of glycinin: effect of re-
ducing agent. Journal of Agricultural and Food Chemistry, 41,
168176.