Content uploaded by Pradipta Banerjee
Author content
All content in this area was uploaded by Pradipta Banerjee on Feb 24, 2016
Content may be subject to copyright.
Isolation and Identification of Cryptic Bioactive Regions in Bovine
Achilles Tendon Collagen
Pradipta Banerjee •G. Suseela •C. Shanthi
Published online: 5 May 2012
ÓSpringer Science+Business Media, LLC 2012
Abstract Several proteins are known to host specific
regions within their sequence, that when exposed or
excised out proteolytically can display a range of physio-
logical activities quite different from that of the parent
protein. Collagen, a class of structural biopolymers and an
important component of the extracellular matrix, is now
known to harbor several such bioactive peptides which can
act as physiological regulators. This study was undertaken
to identify such cryptic sites from bovine Achilles tendon
collagen and an antioxidative assay was used to screen for
bioactivity. Bacterial crude protease was used to hydrolyze
collagen and the hydrolysate was subjected to separation
through ion-exchange column chromatography. Fractions
were screened using conventional antioxidative assays and
further purified by gel permeation chromatography. Two
biologically active cryptic peptides were obtained dis-
playing high antioxidative properties, E1 and F3. At low
concentrations, both peptides displayed higher chelating
ability than EDTA and were able to reduce the auto-oxi-
dation of unsaturated fatty acid. The molecular weights of
the peptides were found out through column chromatog-
raphy and Tricine SDS PAGE; both displayed molecular
mass below 4 kDa. Overall E1 displayed a comparatively
better antioxidative ability than the others and was further
characterized by circular dichroism studies and sequencing.
A BLAST search of the active peptide sequence revealed
that an almost similar peptide also resides in human col-
lagen Type I.
Keywords Bovine Achilles tendon Chromatography
Collagen Cryptic peptides Antioxidative activity
screening
Abbreviation
ROS Reactive oxygen species
TNBS 2, 4, 6-trinitrobenzene sulfonic acid
BHT Butylated hydroxy toluene
DPPH 1, 1-diphenyl -2-picrylhydrazil
EDTA Ethylenediaaminetetraacetic acid
PP-II Polyproline-II
CD Circular dichroism
1 Introduction
The ECM is an intricate three dimensional network sur-
rounding the cells in a tissue. It is composed of fibrous
proteins and proteoglycans that are secreted locally and
arranged into a vast ordered meshwork of several macro-
molecules [17]. Collagen, a component of the ECM, is a
commonly occurring bio-polymer that is essential for the
structural integrity of most body organs [12]. It represents
the major structural protein accounting for approximately
30 % (dry weight) of all vertebrate body protein. More than
90 % of the extracellular protein in the tendon and bone
and more than 50 % in the skin consist of collagen [9].
Collagen, once considered only as an architectural
support for cells, has been recently found to possess other
P. Banerjee C. Shanthi (&)
School of Bio Science and Technology, VIT University, Vellore
632014, Tamil Nadu, India
e-mail: cshanthi@vit.ac.in
P. Banerjee
e-mail: roon17779@gmail.com
G. Suseela
Bacteriology Laboratory, Central Leather Research Institute,
Chennai 600020, Tamil Nadu, India
e-mail: suseelarajkumar@hotmail.com
123
Protein J (2012) 31:374–386
DOI 10.1007/s10930-012-9415-8
physiological activities [29], including antioxidative prop-
erties. Such kind of bioactivity is probably restricted to
small non-repetitive portions of the entire chain, which
generally remains hidden until structural changes in the
molecule expose them or when they are proteolytically
cleaved. More than one such fragment can be derived from
a single protein precursor and these peptides often show
biological activities that are different from the parent
molecule. Such peptides have been reported to as cryptic
bioactive peptides [4,47]. Screening for antioxidative or
free radical scavenging activity is a plausible way of
identifying these cryptic regions in collagen.
Free radicals are a highly reactive short-lived species
generated in vivo or in vitro through physiological reac-
tions, particularly in vertebrates [13]. Normally, they serve
in signaling routes and provide defense against infections
[9,13]. Any excessive amount of these highly reactive
species can result in cellular damage, which, in turn can
initiate several diseases including atherosclerosis, arthritis,
diabetes and cancer [8,10]. Under normal circumstances,
the antioxidant defense system of the body, comprising of
an array of enzymatic and non-enzymatic antioxidative
components are enough to counter this free radical load
[20]. However, in certain circumstances, like an excessive
free radical load or a faulty antioxidative component, this
defense system fails. This gives rise to oxidative stress, a
condition in which the generation of highly reactive mol-
ecules like hydroxide, superoxide and peroxide radicals
(collectively termed as reactive oxygen species or ROS)
exceed their quenching, resulting in uncontrolled reactions
with the cell macromolecules [28,44].
Moreover free radicals play a part in food storage.
Oxidation of fats and oils during processing and storage of
food products increase the chances of free radical pro-
duction, which can have a deleterious effect on the nutri-
tive value of the product.
Bovine Achilles tendon collagen, a good source of sol-
uble collagen type I was chosen for our study and obtained
from slaughter house waste. Although antioxidative pep-
tides have been isolated from the skin of bovine, porcine
and fish by utilization of a cocktail of protease enzymes or
sequential cleaving of the protein by the individual prote-
ases [21,24,30,33], less data exists on the isolation of
cryptic antioxidative peptides from bovine tendon.
This study was undertaken to identify bioactive regions
in Bovine Achilles tendon collagen by hydrolysis followed
by chromatography and antioxidative screening of the
resultant peptides. Whereas synthetic antioxidants like
BHT are cost-effective and efficient but may display some
toxic effects in higher amounts [18], use of such cryptic
bioactive peptides as antioxidants offer several advantages.
They are specific, of low molecular weight, with minimum
toxicity, stable over a broad range of environments and
present nutritional benefits besides their antioxidative
properties [14,54]. Moreover, the low antigenicity and
high biocompatibility of collagen [23,26] makes its use as
natural antioxidant justifiable.
2 Materials and Methods
TNBS, BHT, DPPH, linoleic acid, chromatographic matri-
ces (Carboxy methyl Sepharose, Sephadex G25, G75, G100
and G200) and molecular weight markers were obtained
from Sigma-Aldrich Chemicals Private Limited, Bangalore,
India. Ferrozine {3-(2-pyridyl)-5, 6-bis (4-phenylsulfonic
acid)-1, 2, 4-triazine, monosodium salt} was obtained from
Himedia, India. All other chemicals used were of analytical
grade. Bovine Achilles tendon collagen was isolated and
purified according to earlier procedures [40]. The crude
collagenase used in this study was harvested from Alcalig-
enes odorans (designated BL-11 and identified at Institute of
Microbial Technology, Chandigarh, India [49]).
2.1 Preparation of the Hydrolysate
A25mgml
-1
solution of collagen was prepared in 0.2 M
phosphate buffer, pH 7.5 and incubated at 37 °C for
30 min in a screw-cap culture vial. 2 ml of the crude
enzyme was added to the buffered collagen under sterile
conditions and the mixture incubated at 37 °C for 24 h. As
controls, vials containing only collagen and only enzyme
were used. After the incubation time, the reaction was
arrested by addition of 0.1 M acetic acid.
2.2 Degree of Hydrolysis
Degree of hydrolysis was obtained by the Hermanson
method [15] using TNBS. Every 3 h, 0.5 ml of aliquot was
aspirated from the reaction mixture under sterile conditions
till 24 h. The aliquots were dialyzed in buffer and lyoph-
ilized. 5 mg of the powdered samples were dissolved in a
reaction buffer of 0.1 M sodium bicarbonate, pH 8.5. To
0.25 ml of each sample, 0.25 ml of 0.01 % TNBS was
added and incubated at 28 °C for 2 h. The reaction was
arrested by addition of 0.25 ml of 10 % SDS followed by
0.125 ml of 1 N HCl. The absorbance was measured at
335 nm in a UV-1800 spectrophotometer (Shimadzu,
Japan). All samples were done in triplicates. For complete
hydrolysis, a separate sample of buffered collagen was
digested with 6 N HCl for 3 h and its degree of hydrolysis
determined. The remaining part of the samples was sub-
jected to antioxidative assay to determine the time at which
the bioactivity was most prominent.
Identification of Cryptic Sites in Collagen Type I 375
123
2.3 Purification of the Peptides by Chromatographic
Techniques
2.5 ml of the hydrolysate was aspirated under aseptic
conditions and mixed with 2 ml of binding buffer (acetate
buffer, 0.1 M, pH 4.5). The resultant solution was heated in
a water bath at 65 °C for 3 min and centrifuged at
2,000 rpm, for 5 min to remove the insoluble debris.
Separation of the peptides was performed in two stages. In
the first stage, the hydrolysate was subjected to ion-
exchange chromatography and the fractions obtained
screened for bioactivity. Active fractions were further
separated by gel-permeation chromatography.
For ion-exchange chromatography, 2 ml of the prepared
hydrolysate sample was dissolved in 2 ml of the binding
buffer and applied to a 1.4 915 cm column of CM
Sepharose equilibrated in the same buffer. Elution was
done using an increasing NaCl gradient varying from 0.025
to 0.5 M. The flow rate was maintained at 1.2 ml min
-1
.
3 ml fractions were collected and absorbance monitored at
210 and 220 nm. Active fractions were further separated by
automated gel-permeation chromatography using an
A
¨KTAprime plus unit (GE Healthcare, United Kingdom).
Fractions having high absorbance values were pooled,
desalted and lyophilized in a Micromodulyo freeze dryer
(Thermo Scientific, USA). The lyophilized fractions were
dissolved in acetate buffer at a concentration of 5 mg ml
-1
and applied to a 1.4 925 cm column of Sephadex G100
equilibrated with 50 mM acetate buffer, pH 4.5. The col-
umn effluent flow rate was maintained at 1.5 ml min
-1
;
absorbance was monitored at 214 nm (using a Zinc lamp)
and fractions of 3 ml were collected. Fractions representing
the major portion of a given peptide peak were pooled,
dialyzed in acetate buffer and lyophilized.
Following separation each of the fractions were
screened for bioactivity. The active fractions were dis-
solved in the same acetate buffer and applied to a
1.4 925 cm Sephadex G25 column for further separation.
The column effluent flow rate was maintained at
0.8 ml min
-1
, absorbance was monitored at 214 nm and
fractions of 3 ml were collected. Fractions showing high
absorbance values were pooled, dialyzed in acetate buffer,
lyophilized and re-subjected to antioxidative assays.
2.4 Determination of Antioxidative Properties
2.4.1 Determination of Radical Scavenging Activity
The scavenging effect of the peptides on DPPH free radical
was measured according to the method of Mensor et al.
[31] with some modifications. Known amounts of the test
peptides were dissolved in 1 ml methanol, diluted to 3 ml
with deionized water and 1 ml of 0.3 mM DPPH in
methanol was added to it. Samples were kept in the dark
for 30 min and the absorbance measured at 517 nm. The
radical scavenging activity (%) was calculated according to
the following equation:
Scavenging activity ð%)
¼Control þBlank SampleðÞ
fg
=Control½100:
Blank was set with methanol. 1 ml DPPH, 2 ml deion-
ized water along with 1 ml methanol was used for control.
For peptides insoluble in methanol, acetic acid buffered
methanol, as suggested by Sharma and Bhatt [50], was
used for dissolution. The controls and the blank were
suitably modified. The synthetic antioxidant, BHT was
used as positive control.
2.4.2 Determination of Chelating Ability
Ferrous ion chelating activity was determined according to
the method of Dinis et al.[6] with some modifications. The
test samples were dissolved in dilute acetic acid (0.05 M)
in increasing concentrations. 1 ml was aspirated from each
sample, diluted with 1 ml deionized water and added to a
0.5 ml solution of 2 mM FeSO
4
. The reaction was initiated
by addition of 0.1 ml of a 5 mM ferrozine solution, the
mixture shaken vigorously and incubated at 28 °C. After
10 min, absorbance of the solutions was measured at
562 nm. EDTA was used as positive control. BSA and
collagen were also checked for chelating ability. All tests
were conducted in triplicates. The chelating effect (%) was
calculated according to the following equation:
Chelation ð%)¼Abscontrol Abstest
ðÞ=Abscontrol
½100:
Since the ferrous-dye complex is stable over a pH range
from 4 to 8 [36], the effect of change in pH of the reaction
medium on the chelating ability of the peptides were also
measured. Different reaction systems with varying pH were
set up with suitable controls for the assay.
2.4.3 Measurement of the Reductive Ability
For the measurement of reductive ability, the ferric-ferrous
transformations in the presence of the peptides were mea-
sured according to the method of Oyaizu et al.[35] with
some modifications. Known amounts of the test peptides
were dissolved in 3 ml of 0.2 M phosphate buffer, pH 6.6.
2 ml of 1 % potassium ferricyanide solution was added and
the reaction mixture incubated at 50 °C for 20 min fol-
lowed by addition of 1 ml 20 % trichloroacetic acid.
2.5 ml of the test solutions was aspirated out, mixed with
2.5 ml deionized water, 0.5 ml of 0.1 % ferric chloride
solution and incubated at 28 °C for 10 min. The absor-
bance of the resulting solutions was measured at 700 nm.
376 P. Banerjee et al.
123
Triplicate tests were conducted for each sample. Increase
in absorbance was taken as an indication of greater
reductive abilities.
2.4.4 Measurement of Linoleic Acid Peroxidation
Inhibition Assay
The method of Li et al.[24] was followed to measure the
extent of lipid auto-oxidation. 1.5 ml of 50 mM linoleic
acid in 99 % ethanol was mixed with an equal volume of
phosphate buffer of pH 7 and 2 ml of the test peptide in
phosphate buffer was added. BHT was used as the positive
control. The final solution in a screw-cap glass tube was
incubated in the dark at 60 °C. Triplicate glass tubes were
used for each sample. The degree of oxidation was mea-
sured at 24 h intervals using ferric thiocyanate for color
development. To 100 ll of the reaction mixture, 4.5 ml of
75 % ethanol, 100 ll of 30 % ammonium thiocyanate,
200 ll of 1 N HCl and 100 ll of 20 mM ferrous chloride
solution in 3.5 % HCl was added sequentially and the color
developed measured at 500 nm.
2.5 Molecular Weight Distribution Profile
Molecular weight distribution of the peptides was done by
using gel permeation chromatography. Four different
Sephadex matrices ranging from Sephadex G25 to G200
were calibrated according to the standard protocol given by
Sigma. The fractions showing high bioactivity were
applied to the calibrated columns under identical condi-
tions. From the (elution volume/void volume) values
obtained, the molecular weights of the applied samples
were calculated.
2.6 Tricine SDS-PAGE of Purified Peptides
The hydrolysate, along with lyophilized purified peptides
E1 and F3 was run in a 16 %/6 M urea Tricine SDS-PAGE
according to Scha
¨gger’s protocol [46]. Ultra low range
molecular weight markers were also run under similar
conditions.
2.7 Sequence Determination
2.7.1 Trypsin Digestion
5 mg of the peptide E1 was dissolved in 50 mM ammo-
nium bicarbonate buffer, pH 8. Mass spectroscopy-grade
trypsin was added at an enzyme: substrate ratio of 1:50 and
the sample was digested for 15 h at 37 °C. The digestion
was stopped by freeze-drying the sample.
2.7.2 MALDI-TOF Mass Spectroscopy of Tryptic Peptides
For Matrix-assisted Laser Desorption/Ionization Time-
of-flight Mass Spectrometry (MALDI-TOF MS), the
Bruker Ultraflex TOF/TOF instrument was run in reflective
mode with delayed extraction and an acceleration voltage
of 25kvA to improve signal-to-noise ratio. 50–100 spectra
were summed. Flex Analysis 2.0 software was used to
analyze the mass spectra.
Trypsin autolysis products at 659.48, 803.024,
1,113.331, 1,431.608 values were removed from the final
spectra and the data was exported to Mascot Peptide Mass
Fingerprint (http://www.matrixscience.com/). The mass
values were matched to the UniProtKB/Swiss-Prot, a
curated protein sequence database (http://expasy.org/
sprot/).
For search, peptides were assumed monoisotopic and
oxidized at methionine residues. An ‘other mammalia’
taxonomy restriction was used, a maximum of three missed
cleavages were allowed, and a peptide mass tolerance of
0.6 kDa was used for peptide mass fingerprinting.
2.8 Circular Dichroism of E1
CD measurements were performed on a Jasco Model J-715
spectropolarimeter. (Jasco, Japan) using a quartz cylindri-
cal cuvette with a path length of 1 mm. The peptide was
dissolved in solutions varying pH at a concentration of
150 lgml
-1
. The solvents used for dissolution were
50 mM acetic acid (pH 3.4), 50 mM acetate buffer (pH
4.5), 50 mM phosphate buffer for pH 6 and 7.4. The cuv-
ette was pre-washed with the solvent before usage and
filled with 200 ll of the solution for each measurement.
The CD spectra were obtained by continuous wavelength
scans from 240 to 195 nm at a scan speed of 50 nm min
-1
at a temperature of 28 °C.
2.9 Statistical Analysis
EC
50
values were calculated using linear regression lines
from the data given in Fig. 6. Values of the activities are
reported as mean ±standard deviation. The data were
analyzed for statistical significance using one way ANOVA
and student’s ttest. pvalues less than 0.05 were considered
significant.
3 Results and Discussion
3.1 Degree of Hydrolysis
Enzymatic digestion of parental proteins is an effective
method of obtaining peptides with specific bioactivity.
Identification of Cryptic Sites in Collagen Type I 377
123
Hydrolysis can also result in the production of more free
N-terminal and C-terminal amino acid residues, which in
turn, effects the peptides hydrophobicity and as a result, its
availability in physiological systems.
Previous studies have reported that the degradation of
native collagen by prokaryotic and eukaryotic collagenases
follow distinctly different pathways. Bacterial collagenases
follow an unspecified mode of action, compared to the
strict specificity observed by eukaryotic collagenase [22].
The cleavage sites recognized by most bacterial collagenases
comprise of the Y-G bond in the G-X–Y alternating motif
of native collagen while mammalian collagenases are
specific for only one cleavage site in collagen Type I [48].
In order to obtain as many bioactive regions of collagen as
possible, we have used bacterial collagenase for hydrolysis,
resulting in more number of peptides. A curve was plotted
with dissociation (%) against time in hours as shown in
Fig. 1which gave an estimate of the time from which the
cryptic bioactive peptides actually started displaying their
activity. A positive correlation was found between the
degree of hydrolysis curve and the antioxidant activity
curve indicating that the activity increased with increase in
dissociation (%). Radical scavenging activity started
increasing only after 10 h and reached its maximum at
about 24 h. It was noted that there was a basal chelating
activity already present in collagen and this activity
increased rapidly from 10 h onwards till 24 h.
3.2 Purification of the Cryptic Bioactive Peptides
Chromatography has played a key role in the purification of
proteins and peptides for some decades. Collagen peptides are
known to be positively charged at acidicpH, and therefore are
amenable to be separated by ion-exchange chromatography
[57]. CM Sepharose, a weak cationic exchanger, was chosen
as the ion-exchange matrix for primary separation of peptides
in this study. A buffer pH of 4.5 was chosen as it lies very
close to the pk
a
of the CM groups and would provide effective
ionization. For the lack of a substantial number of tyrosine
and tryptophan residues in collagen, a wavelength range of
210–220 nm, specific for the peptide bond, was used for
absorbance measurement during elution [55]. The eluted
fractions were subjected to antioxidative screening to select
for bioactive cryptic peptides.
Upon elution with a salt step-gradient, five clear peaks were
observed and labeled as A, B, C, D, E and F. Pure collagen
eluted out at higher ionic strengths (as shown in Fig. 2). The
Fig. 1 Hydrolysis curve obtained by treating buffered collagen with
microbial protease. Reaction pH 7.5, temperature 37 °C and time of
incubation 24 h. The secondary axis denotes the % chelating (filled
square) and scavenging activity (open circle) which increased in
accordance with the degree of hydrolysis (filled triangle). Values
given are the mean of triplicates
Fig. 2 CM-Sepharose elution pattern of the hydrolysate. A non-
linear gradient of NaCl ranging from 0 to 0.3 M in acetate buffer
(filled triangle) was used as the eluant. As control, pure collagen was
also run under similar conditions (dotted line) and it eluted at a higher
ionic strength of 0.5 M NaCl. The hydrolysate (solid line) was
separated into five different fractions (from left to right, marked A,B,
C,D,Eand F) which were dialyzed, lyophilized and subjected to
antioxidative screening. Fractions Eand Fwere found to possess
comparatively better chelation (stripped bar), reductive ability (open
bar) and scavenging activity (filled bar) as depicted in the bar
diagram
378 P. Banerjee et al.
123
fraction eluting out with the binding buffer most probably
contained neutral or acidic peptides and was termed as fraction
A. Based on the initial antioxidative screening assays as dis-
played in Fig. 2, E and F were identified to be comparatively
more active. The fraction E wasappliedtoaSephadexG100
column and it resolved into three peptide fractions E1, E2 and
E2b as displayed in Fig. 3a. E2b when run in a Sephadex G25
column, resolved into two peaks, E3 and E4 (Fig. 3c). To
assess purity, all four peptides, i.e., E1, E2, E3 and E4 were run
individually in a Sephadex G25 column where each displayed
a single peak. The peptide fractions were pooled, dialyzed,
lyophilized and stored in -70 °C.
The fraction F upon loading onto a Sephadex G75 col-
umn resolved into two fractions as shown in Fig. 4a. Both
fractions were pooled, dialyzed, lyophilized and individu-
ally applied to a Sephadex G25 column for further purifi-
cation. F1 was found to display a single peak while F1B
was separated into two distinct peaks, F2 and F3 (Fig. 4c).
3.3 Antioxidative Assay
3.3.1 Screening of the Peptides
An initial comparative chelating ability assay was per-
formed with the crude fractions obtained after ion-exchange
chromatography (labeled A, B, C, D, E and F). As displayed
in Fig. 2, overall, fraction E was found to possess the highest
antioxidative properties (36.8 ±0.4 % scavenging and
27.2 ±1.23 % chelation activity) followed by F (29.4 ±
0.93 % scavenging, 18.1 ±0.35 % chelation activity) and
C (19.8 ±1.42 % scavenging and 19.3 ±0.29 % chelation
activity) in comparison to collagen (0 % scavenging,
8±0.53 % chelation), the hydrolysate (14.6 ±0.75 %
scavenging, 12.3 ±0.49 % chelation) and BSA (values not
shown in Fig. 2). EDTA, used as a positive control was
found to possess lower chelating ability (9.4 ±0.75 %) at
the amounts used. Fractions E and F were chosen for further
purification. The peptides separated from fraction E and F
were dissolved in methanol and buffered methanol, as per
the solubility of the sub-fractions, with respective controls.
As displayed in Figs. 3c, d, 4c, d, the peptides E1 and F3
showed maximum radical scavenging activity (34.167 ±
1.167 %, 33.8 ±0.65 % respectively), followed by E3, F2,
F1 and E2. E1 and F3 were also found to have a compara-
tively higher chelating ability, 33.8 ±0.85 %, and 25.4 ±
0.92 % respectively followed by E3, E4, F1, F2 and E2. The
purified peptides E1 and F3 were found to have comparable
reductive ability and both, at higher amounts, were able to
match the reductive ability of BHT. They were also able to
inhibit the linoleic acid auto-oxidation significantly as
Fig. 3 a Elution profile of
active fraction E in Sephadex
G100. Three fractions were
obtained, marked E1, E2 and
E2b (From left to right) in order
of decreasing molecular weight.
bThe three fractions were
subjected to antioxidative
screening assay{Chelation
activity (stripped bar), reductive
ability (open bar) and
scavenging activity (filled
bar)}; E1 and E2b displayed
high activity. cThe E2b
fraction, when run in Sephadex
G25 separated into two uneven
peaks, marked E3 and E4.
dUpon being subjected to
screening assays, E3 showed a
higher response. The unit of
Y-axis in both chromatograms
is in milli-absorbance
Identification of Cryptic Sites in Collagen Type I 379
123
shown in Fig. 5; the results were found significant
(pB0.001) at 95 % confidence level based on one way
ANOVA.
3.3.2 Effect of Concentration on Antioxidative Activity
of Purified Peptides
As displayed in Fig. 6, for all the test peptides, there was
an almost hyperbolic increase in radical scavenging and
chelating activity with increasing concentration. The pep-
tides acted as potent antioxidative agents at lower con-
centrations but the activity asymptotically approached a
maximum of 70 % at higher concentrations, probably due
to the inherent recoiling nature of collagen type I peptides
[43]. Upon comparison of the EC
50
values the scavenging
activity was found to be in the given order, E1 followed by
F3, E3 and F2. The EC
50
values for chelating ability were
in the given order, F3 followed by E1, F2 and E3.
BHT was used as the standard for scavenging activity
and achieved its maximum 74 % with a concentration as
low as 0.01 mg ml
-1
. EDTA was used as the positive
control for chelation assay. At amounts lower than
0.06 mg ml
-1
, the chelating ability of the peptides were
four times greater than that of EDTA. However at amounts
more than 0.2 mg ml
-1
, the chelation ability of EDTA
Fig. 4 a Elution profile of F in
Sephadex G75. Two fractions
were obtained; marked F1 and
F1b, from left to right in order of
decreasing molecular weight.
bThe fractions were subjected to
antioxidative screening
assay{Chelation activity (stripped
bar), reductive ability (open bar)
and scavenging activity (filled
bar)}; F1b displayed considerably
high activity. cElution profile of
F1b in Sephadex G25; it was
resolved into two peptides, marked
F2 and F3, from left to right with
decreasing molecular weights.
dF3 was noted to be a
comparatively better antioxidant.
The unit of Y-axis in both
chromatograms is in milli-
absorbance
Fig. 5 Lipid peroxidation inhibitory activity of the peptides. Purified
peptides E1 (filled square) and F3 (open square) along with the
hydrolysate (filled triangle) was used for the assay. Linoleic acid was
used as the control (filled diamond) and BHT was used as the positive
control (open diamond). As evident from the figure, the peptides
reduced the auto-oxidation of linoleic acid, resulting in reduction of
absorbance. Values in the figure are mean of triplicates ±standard
deviation. The results were found significant (p\0.001) at 95 %
confidence level based on two-way ANOVA
380 P. Banerjee et al.
123
reached 100 % and was double the %chelation values for
E1 and F3 and 2.5 times that of E3 and F2.
3.3.3 Effect of pH Variation on Chelating Ability
of Purified Peptides
We found that a change in pH affects a peptide’s bioac-
tivity quite drastically. The results, as depicted in Fig. 7,
show a general decrease in activity of the four peptides
with increasing pH until nil activity. Further increase in pH
leads to a gradual increase in the activity. This was prob-
ably due to the fact that as the pH changed, the ionization
state of the side chain functional groups changed conse-
quently, leading to change in its chelating ability. E1 and
E3 were found to have higher chelating activity at acidic
pH while F2 and F3 had higher activity in slightly alkaline
pH. So although E1 and E3 gave better results with radical
scavenging assay and reductive ability assay, it is F2 and
F3 which could be more potent chelating agents in physi-
ological pH. The results were significant (pB0.001) at
95 % confidence level based on one way ANOVA.
3.4 Identification and Characterization of the Cryptic
Peptide
Among the several requirements for being a bioactive
peptide, possession of a low molecular weight is fairly
important. As displayed in Table 1, the molecular weights
of E1 and F3 were found to be 3 kDa and 1.4 kDa
respectively.
Figure 8shows the electrophoretic pattern of the
hydrolysate along with the purified peptides E1 and F3.
The hydrolysate appears as a long dark streak and the
peptides appear as distinct single low molecular weight
bands. F3 and E1 matched with the 1.4 kDa and 3.8 kDa
band respectively, in the marker lane. These data support
the fact that most of the peptides exhibiting bioactivity are
those with low molecular weight, as noted in previous
studies [38]. Since E1 displayed an overall better antioxi-
dative ability than F3, it was further characterized by
sequencing and CD spectral analysis.
Fig. 6 Effect of concentration upon the antioxidative properties of
the peptides. E1 (open square), E3 (open diamond), F2 (cross) and F3
(filled diamond) displayed an asymptotic increase in their activity.
EDTA was used as the positive control for chelation assay and BHT
was used as the positive control (filled triangle) for scavenging and
reductive ability assay. Values given are the mean of triplicates
Fig. 7 Effect of pH variation upon the chelating ability. The four
purified peptides, E1 (stripped bar), E3 (filled bar), F2 (checked bar)
and F3 (open bar) were assayed for chelating ability at a range of pH
from 3.4 to 7.4. It was observed that each peptide at a particular pH
displayed nil activity and regained the activity beyond a certain pH.
This loss in activity could be due to the fact that the pH involved had
reached the isoelectric point (pI) of the individual peptides, rendering
them with no charge and making them unavailable for chelation
Identification of Cryptic Sites in Collagen Type I 381
123
Biological activities of the peptides are dependent on the
amino acid composition and also their positioning in the pep-
tide, along with their molecular weight. The amino acid
sequence obtained after mass peak identification of E1 was
found to match significantly with position 1,066 to 1,101 of
collagen alpha-1(I) chain precursor P02453 (CO1A1_BOVIN).
(URL:http://www.uniprot.org/uniprot/P02453)(Fig.9).
GETGPAGPAGPIGPVGARGPAGPQGPRGDKGETGEQ
An analysis of the sequence enabled the peptide to be
theoretically divided into two parts; beginning from the N
terminal G to the 17th residue and from the 18th residue to
the C terminal. The N terminal half is primarily hydro-
phobic in nature with only one charged residue, E, at the
beginning and two hydrophobic residues; I and V. The
remaining C-terminal half comprises of comparatively
more number of charged residues and this half would be
primarily hydrophilic in nature. Since charged amino acid
residues are involved in antioxidative reactions of the
peptide, probably the C terminal portion is responsible for
the chelation and scavenging activity, while the N terminal
portion may have more of a structural role. The hydro-
phobicity of the residues suggests that the C-terminal
portion is indeed more accessible and would, therefore be
the key player responsible for its activities The peptide
possessed two strictly hydrophobic residues (V and I) and
four A residues, 3 of which were present in the N terminal
half. In all probability, it is this half which could have been
responsible for the lipid solubility and, consequently,
inhibition of lipid peroxidation activity of E1. It is pre-
supposed that the presence of certain amino acids such as
Y, F, W, H, M and C are crucial for the antioxidative
activity of peptides [16,41,42,51]. Most of these amino
acids are aromatic in nature and/or have electron pairs
which can be donated to free radicals or used to chelate
ions. In spite of lacking the above mentioned amino acids,
E1 was able to exhibit a strong antioxidative capacity. This
is quite possible, as recent data suggests the presence of
amino acids such as L, Q, E and D [1,41] and certain
sequences such as QG and GP [5,15] can render a peptide
with antioxidative properties. An analysis of the amino
acids in E1 showed it to have six GP sequences, one QG
sequence along with charged amino acids K, D, E and R,
which could actively participate in charge/electron transfer
reactions, resulting in antioxidant properties. The presence
of pin the sequence imposes conformational restraints on
the secondary structure and may have an indirect effect on
the bioactivity. More than their presence, the order and
sequence in which these amino acids are arranged can
affect the structure, reactivity and solvent accessibility of
E1, which in turn, dictates its inherent bioactivity [7].
A comparison with other bioactive collagen peptides
showed E1 to have a moderate antioxidant activity. For rad-
ical scavenging activity, the EC
50
of the peptide E1 was found
to be the lowest (338 lgml
-1
) among the four bioactive
peptides obtained. It was lower than the peptides DPALA-
TEPDPMPF (EC
50
=660 lgml
-1
)obtainedfromNile
tilapia scale gelatin [34], and RSGH-Pc, a cobia gelatin
peptide (scavenging activity of 60.7 % at a concentration of
10 mg ml
-1
)[56]. At a concentration of 11 mg ml
-1
,pepsin
hydrolysate of porcine skin collagen had a radical scavenging
Table 1 Details of the four cryptic bioactive peptides including their molecular weight
Fraction Purified
Peptides
Elution pattern in IEC
a
Molecular
Weight (kDa)
Chelating ability Radical scavenging activity
EC
50
(mg ml
-1
)C
max
(%) EC
50
(mg ml
-1
)RS
max
(%)
E E1 Elutes out with 0.2 M NaCl 3 0.163 80.4 ±1.22 0.338 71.6 ±0.7
E3 Elutes out with 0.2 M NaCl 0.95 0.232 71.5 ±1.0 0.482 70.4 ±0.18
F F3 Elutes out with 0.3 M NaCl 1.4 0.147 73.9 ±2.15 0.382 70.0 ±0.47
F2 Elutes out with 0.3 M NaCl 2.4 0.211 63.6 ±1.56 0.458 58.4 ±0.8
EC
50
values and their maximum antioxidative activity (C
max
and RS
max
) are displayed. EC
50
values are calculated from the regression lines
obtained from the figures
a
Stands for ion-exchange chromatography
Fig. 8 Electrophoretic pattern of peptides E1 and F3 obtained by
tricine-SDS-PAGE. From the left;Lane 1 Ultra low range molecular
weight markers. lane 2 Hydrolysate. Lane 3 F3 purified and Lane 4 E1
purified
382 P. Banerjee et al.
123
activity of 87.18 ±1.84 % [24]. The EC
50
, however was
more than that of the peptide LEELEEELEGCE, isolated
from bullfrog skin (EC
50
=16.1 lM) [39]. The papain
hydrolysate of tuna backbone protein also showed a DPPH
scavenging activity of 36.72 % [19].
For chelating activity, the peptide E1 had an EC
50
of
163 lgml
-1
, which was quite low when compared to
collagen peptides from other sources. Peptides from por-
cine skin and squid collagen displayed a chelating activity
of 37.4 ±1.5 % at a concentration of 11 mg ml
-1
[24]
and 80 % activity at 0.2 mg ml
-1
[11] respectively. Che-
lating ability was also compared with peptides obtained
from other helical fibrous proteins; Pacific hake muscle
hydrolysate peptides possessed chelating ability ranging
from approximately 7–46 % at a concentration of
5mgml
-1
[45] and round scad protein muscle hydrolysate
possessed an activity of about 60 % [52].
The molecular weight of E1 was found to be
3.2805 kDa, which matched closely with the results
obtained from gel permeation chromatography and tricine-
SDS-PAGE. The peptide displayed the regular G-X–Y
arrangement of amino acids, so typical of collagen, as is
evident from the sequence. CD spectroscopy was done to
determine the secondary structural characteristics and
whether the conformation of the peptide affected its bio-
activity. The CD data of E1, as displayed in Fig. 10,
comprises of a large negative peak at 196–205 nm and a
small positive peak around 220 nm, similar to collagen
[37]. Both peaks responded to a change in pH. The nega-
tive band minimum at 196 nm was seen to shift towards
higher wavelengths with increase in pH. At physiological
pH, the negative minimum reached a wavelength of
205 nm, indicating a possible random coil configuration
[25]. The positive peak around 220 nm, on the other hand,
decreased with the pH, indicating a reduction in polypro-
line-II (PP-II) helical content and a subsequent increase in
random coil conformation [37]. Although native collagen is
triple helical, earlier reports have stated that single a-chain
of collagen takes up a PP-II conformation in solution [32,
53]. Overall, the CD data of the peptide supported the fact
Fig. 9 MALDI-TOF spectrum
of the Trypsin generated
fragments of E1
Identification of Cryptic Sites in Collagen Type I 383
123
that at lower pH, it attains a PP-II helix conformation and
changes to random coil at alkaline pH. This is strongly
indicative of the fact that the conformation of the peptide
does indeed affect the bioactivity. Since the PP-II
conformation is more solvent accessible [27], the func-
tional groups of the amino acids would be able to interact
better with the reactive radicals. In alkaline conditions, the
peptide shifts to the random coil conformation, leading to
probable structural changes in the reacting groups. In
addition, increasing pH brings about a change in the charge
state of the reactive functional groups. Both factors acting
together, might explain why the activity of E1 varies
markedly over a broad range of pH.
A BLAST search [2,3] of the sequence against Uni-
ProtKB/Swiss-Prot protein sequences revealed that other
than the RGD motif, the peptide did not contain any
putative conserved domains (Query ID: 11468). However,
the sequence, with minor variations was found to be
present in other mammals including humans (94 % iden-
tical) as shown in Table 2. It was interesting to note that in
most of the sequences, the peptide was present in at almost
the same position; in the vicinity of residue number
1,060–1,100, lying near the telopeptide region. The mam-
malian collagenases cleave collagen into two fragments
and most probably, wouldn’t be able to release the active
peptide immediately. But once this first stage of collagen
degradation is complete, other proteases can cleave further
Table 2 pBLAST results
The sequence of E1 compared
with closely matching collagen
sequences from other species. It
clearly depicts the near-
homogeneity in amino acid
sequence of collagen across
diverse species
Species Type Sequence Identity Score
1Bos taurus I
α1
1066 1101 100 159
GETGPAGPAGPIGPVGARGPAGPQGPRGDKGETGEQ
2Canis
lupus
familiaris
I
α1
1063 1098 100 159
GETGPAGPAGPIGPVGARGPAGPQGPRGDKGETGEQ
3Rattus
norvegicus
I
α1
1056 1091 97 154
GETGPAGPAGPIGPAGARGPAGPQGPRGDKGETGEQ
4Mus
musculus
I
α1
1056 1091 97 154
GETGPAGPAGPIGPAGARGPAGPQGPRGDKGETGEQ
5Homo
sapiens
I
α1
1067 1102 94 153
GETGPAGPTGPVGPVGARGPAGPQGPRGDKGETGEQ
6Rattus
norvegicus
III
α1
1064 1099 81 120
GETGPAGPSGAPGPAGARGAPGPQGPRGDKGETGER
7Mus
musculus
III
α1
1064 1099 81 120
GETGPAGPSGAPGPAGARGAPGPQGPRGDKGETGER
8Xenopus
laevis
II
α1
1091 1125 80 117
GESGPQGPLGPSGPAGARGLAGPQGPRGDKGEAGE
9Mus
musculus
II
α1
1089 1124 78 116
GEAGAQGPMGPSGPAGARGIAGPQGPRGDKGESGEQ
10 Homo
sapiens
I
α2
979 1014 75 111
GETGPSGPVGPAGAVGPRGPSGPQGIRGDKGEPGEK
11 Xenopus
(Silurana)
tropicalis
II
α1
1094 1128 77 111
GESGPQGPLGPSGPAGARGLPGPQGPRGDKGEAGE
12 Bos taurus II
α1
1089 1123 77 108
GEAGAQGPMGPAGPAGARGMPGPQGPRGDKGETGE
13 Rattus
norvegicus
II
α1
1021 1055 77 108
GEAGAQGPMGPSGPAGARGIAGPQGPRGDKGEAGE
14 Rana
catesbeiana
I
α2
970 1005 75 107
GEGGPSGPAGITGPSGPRGPAGPQGVRGDKGEAGER
Fig. 10 CD spectra of purified peptide E1 at pH 3.4 (solid line), pH
4.5 (dotted line), at pH 6 (dashed line) and at pH 7.4 (dot with dashed
line). The curve progressively shifts towards random coil conforma-
tion as pH increases. The four spectra were recorded at 28 °C
384 P. Banerjee et al.
123
until such bioactive peptides are either ‘‘exposed’’ or
released and can act as in vivo antioxidant in case of a free
radical overload.
4 Conclusions
Cryptic bioactive peptides are considered as short segments
of proteins that, when proteolytically cleaved or exposed
by conformational change, can exert various biological
functions. In this study, an antioxidative screening assay
was used to identify such a specific region in Collagen
Type I displaying comparatively high bioactivity. It was
interesting to note that even in the absence of the con-
ventional amino acids required for antioxidative activity;
the peptide could exhibit moderately high activity, proba-
bly due the presence of charged amino acids and the PP-II
conformation. Similar type of sequences was also found in
collagen from other species. The results of our study sub-
stantiate the potential bioactivity hidden in certain regions
of collagen, emphasizing its possible role as an active
participant in countering oxidative stress in vitro. Further
work is being carried out to identify other bioactive prop-
erties of the peptide.
Acknowledgments This study was supported by research grant
from DST, Govt. of India (SERC SL NO 1328) and VIT University,
Vellore. We also would like to thank Mr. Sathish and Mrs. Sunita,
Proteomics division, Molecular biophysics unit, Indian Institute of
Science, Bangalore who have helped enormously in determining the
peptide sequence.
References
1. Alema
´n A, Gime
´nez B, Pe
´rez E, Go
´mez-Guille
´n MC, Montero P
(2011) Food Chem 125:334–341
2. Altschul SF, Madden TL, Scha
¨ffer AA, Zhang J, Zhang Z, Miller
V, Lipman DJ (1997) Nucleic Acids Res 25:3389–3402
3. Altschul SF, Wootton JC, Gertz EM, Agarwala R, Morgulis A,
Scha
¨ffer AA, Yu Y-K (2005) FEBS J 272:5101–5109
4. Autelitano DJ, Rajic A, Smith AI, Berndt MC, Ilag LL (2006) M
Vadas. Drug Discov Today 11:306–314
5. Byun H-G, Lee JK, Park HG, Jeon J-K, Kim S-K (2009) Process
Biochem 44:842–846
6. Dinis TCP, Madeira VMC, Almeida LM (1994) Arch Biochem
Biophys 315:161–169
7. Elias RJ, Kellerby SS, Decker EA (2008) Crit Rev Food Sci Nutr
48:430–441
8. Evans JL, Goldfine ID, Maddux BA, Grodsky GM (2003) Dia-
betes 52:1–8
9. Friess W (1998) Eur J Pharm Biopharm 45:113–136
10. Gelse K, Po
¨schl E, Aigner T (2005) Adv Drug Delivery Rev
55:1531–1546
11. Gime
´nez B, Alema
´n A, Montero P, Go
´mez-Guille
´n MC (2009)
Food Chem 114:976–983
12. Halliwell B (1994) Lancet 344:721–724
13. Hancock JT, Desikan R, Neill SJ (2001) Biochem Soc Trans
29:345–350
14. Hattori M, Yamaji-Tsukamoto KA, Kumagai H, Feng Y, Ta-
kahashi K (1998) J Agric Food Chem 46:2167–2170
15. Hermanson G (1996) Bioconjugate techniques. Academic press,
San Diego
16. Herna
´ndez-Ledesma B, Da
´valos A, Bartolome
´B, Amigo L
(2005) J Agric Food Chem 53:588–593
17. Huxley-Jones J, Robertson DL, Boot-Handford R (2007) Matrix
Biol 26:2–11
18. Ito N, Fukushima S, Tsuda H (1985) CRC Crit Rev Toxicol
15:109–150
19. Je J-Y, Qian Z-J, Byun H-G, Kim S-K (2007) Process Biochem
42:840–846
20. Johansen JS, Harris AK, Rychly DJ, Ergul A (2005) Cardiovasc
Diabetol 4:5
21. Kim SK, Kim YT, Byun HG, Nam KS, Shahidi F (2001) J Agric
Food Chem 49:1984–1989
22. Lecroisey A, Keil B (1979) Biochem J 179:53–58
23. Lee CH, Singla A, Lee Y (2001) Int J Pharm 221:1–22
24. Li B, Chen F, Wang X, Ji B, Wu Y (2007) Food Chem
102:1135–1143
25. Machling DE, Gambee JE, Morris NP, Sakai LY, Keene DR,
Mayne R, Bachinger HP (1996) J Biol Chem 271:13781–13785
26. Maeda M, Tani S, Sano A, Fujioka K (1999) J Control Release
62:313–324
27. Mansiaux Y, Joseph AP, Gelly J-C, de Brevern AG (2011) PLoS
One 6:3
28. Maritim AC, Sanders RA, Watkins JB III (2003) J Biochem Mol
Toxicol 17:24–38
29. Maquart FX, Pasco S, Ramont L, Hornebeck W, Monboisse JC
(2004) Crit Rev Oncol Hematol 49:199–202
30. Megias C, Pedroche J, Yust MM, Giron-Calle J, Alaiz M, Millan
F, Vioque J (2008) Food Sci Technol 41:1973–1977
31. Mensor LL, Menezes FS, Leitao GG, Reis AS, Santos TC, Coube
CS, Leitao SG (2001) Phytother Res 15:127–130
32. Mezei M, Fleming PJ, Srinivasan R, Rose GD (2004) Proteins:
Struct Funct Bioif 55:502–507
33. Nagai T, Suzuki N, Tanoue Y, Kai N, Nagashima T (2007) J
Food Agric Environ 5:64–68
34. Ngo D-H, Qian ZJ, Ryu B, Park JW, Kim S-K (2010) J Func
Foods 2:107–117
35. Oyaizu M (1986) Jpn J Nutr 37:311–319
36. Pascual-Reguera MI, Ortega-Carmona I, Molina-Diaz A (1997)
Talanta 44:1793–1801
37. Pe
ˆcher J, Pires V, Djaafri I, Da nascimento S, Fauvel-Lafe
`ve F,
Legrand C, Sonnet P (2008) Eur J Med Chem 3:1–8
38. Pihlanto A (2006) Int Dairy J 16:1306–1314
39. Qian ZJ, Jung WK, Kim SK (2008) Bioresour Technol
99:1690–1698
40. Radhika M, Sehgal PK (1997) J Biomed Mater Res 35:497–503
41. Rajapakse N, Mendis E, Jung W-K, Je J-Y, Kim S-K (2005) Food
Res Int 38:175–182
42. Ren J, Zhao M, Shi J, Wang J, Jiang Y, Cui C, Kakuda Y, Xue SJ
(2008) Food Chem 108:727–736
43. Saeidi N, Sander EA, Ruberti JW (2009) Biomaterials
30:6581–6592
44. Sarmadi BH (2010) Ismail A 31:1949–1956
45. Samaranayaka AGP, Li-Chan ECY (2008) Food Chem
107:768–776
46. Scha
¨gger H (2006) Nat Protocols 1:16–22
47. Schenk S (2003) Quaranta V 13:366–375
48. Seifter S, Harper E (1971) The collagenases. The enzymes.
Elsevier BV, Amsterdam
49. Shanthi C, Suseela G (1999) Ind Patent 1575DEL99
50. Sharma OP, Bhat TK (2009) Food Chem 113:1202–1205
51. Sheih I-C, Wu T-K, Fang TJ (2009) Bioresour Technol
100:3419–3425
Identification of Cryptic Sites in Collagen Type I 385
123
52. Thiansilakul Y, Benjakul S, Shahidi F (2007) Food Chem
103:1385–1394
53. Vila JA, Baldoni HA, Ripoll DR, Ghosh A, Scheraga A (2004)
Biophys J 86:731–742
54. Xie Z, Huang J, Xu X, Jin Z (2008) Food Chem 111:370–376
55. Yan M, Li B, Zhao X, Ren G, Zhuang Y, Hou H, Zhang X, Chen
L, Fan Y (2008) Food Chem 107:1581–1586
56. Yang J-I, Ho H-Y, Chu Y-J, Chow C-J (2008) Food Chem
110:128–136
57. Zhuang Y, Sun L, Zhao X, Wang J, Hou H, Bafang L (2009) J Sci
Food Agric 89:1722–1727
386 P. Banerjee et al.
123