Cell Biochem Biophys
DOI 10.1007/s12013-017-0817-2
ORIGINAL PAPER
Fish Collagen Hydrolysates Modulate Cartilage Metabolism
Kanchanit Boonmaleerat1 Orawan Wanachewin2 Thanyaluck Phitak1
● ●
Peraphan Pothacharoen1 Prachya Kongtawelert1
●
Received: 19 October 2016 / Accepted: 17 July 2017
© Springer Science+Business Media, LLC 2017
Abstract Osteoarthritis is a degenerative joint disease in
which interleukin-1β plays a major role in the inflammatory
process. Administration of collagen hydrolysate was an
optional treatment of osteoarthritis. Fish has become an
interesting source of collagen hydrolysate because of reli-
gious reason and there is no risk from mad cow disease.
However, the effects of different sizes of fish collagen
hydrolysate on cartilage and chondrocyte metabolism have
not been well studied yet. This study examined the effect of
different sizes of fish collagen hydrolysate on cartilage
metabolism. Three different sizes of fish collagen hydro-
lysate were prepared by size exclusion using centrifugation,
which composed of small fraction (<3 kDa), medium frac-
tion (3–10 kDa) and large fraction (>10 kDa). Using por-
cine cartilage explant, in physiological condition, all the
three fractions had no effect on cartilage metabolism, but
they could induce pro-MMP3 and pro-MMP13 secretions
through activation of p-ERK and p-p38. In pathological
* Prachya Kongtawelert
prachya.k@cmu.ac.th
Kanchanit Boonmaleerat
kanchanit_b@cmu.ac.th
Orawan Wanachewin
orawan.wan@crru.ac.th
Thanyaluck Phitak
thanyaluck.phitak@cmu.ac.th
Peraphan Pothacharoen
peraphan.p@cmu.ac.th
1 additive effect with IL-1β and Oncostatin M (OSM) on
Thailand Excellence Center for Tissue Engineering and Stem
Cells, Department of Biochemistry, Faculty of Medicine, Chiang
Mai University, Chiang Mai 50200, Thailand
2
Program in Biology, Faculty of Science and Technology, Chiang
Rai Rajabhat University, Chiang Rai 57100, Thailand
●
condition induced by interleukin-1β and oncostatin-M,
small and medium fractions showed additive effect with
interleukin-1β and oncostatin-M on cartilage degradation,
whereas large size had no effect. In addition, the effect of
small size occurred through further activation of p-p65,
which resulted in further induction of active-MMP13, while
medium size had a different mechanism. In conclusion, all
three fractions fish collagen hydrolysate had no effect on
cartilage metabolism in physiological condition, but small
and medium fractions had adverse effect on cartilage in
pathological condition. Taken together, various sizes of fish
collagen hydrolysate showed different effects on cartilage
metabolism. Therefore, different sizes of fish collagen
hydrolysates play different roles on cartilage metabolism,
especially in the pathological condition.
Keywords Fish collagen hydrolysate Cartilage explants
● ●
Chondrocyte IL-1β OSM
● ●
Introduction
The aim of this study is to examine the effects of different
sizes of fish collagen hydrolysate (FCH) on cartilage
mechanism in both physiological and pathological condi-
tions. Our results showed that all three different sizes of
FCHs had no effect on cartilage metabolism in physiolo-
gical condition. However, small and medium FCHs had an
cartilage degradation in pathological condition.
Collagen is a main constituent of cartilage, bone and
skin. It belongs to a family that comprises 28 members [1].
The distinguishing feature of the collagen molecule is a

right-hand triple helix of α-chains with a distinctive Gly-X-
Y repeating pattern (X and Y positions correspond to pro-
line or hydroxyproline which convert to 3-hydroxyproline
and 4-hydroxyproline respectively during procollagen
synthesis). Procollagen assembly occurs within endo-
plasmic reticulum, then the triple helical molecules are
transported to the golgi aparatus and secreted into extra-
cellular space, and assemble with other types of collagen
[2]. The roles of collagen molecule are structural scaffold,
cell organization and serving mechanonical property. It can
also bind to cells via receptors and induce different sig-
naling pathways, such as proliferation, differentiation and
migration [1]. Collagen turnover is regulated by endo-
peptidase, the matrix metalloproteinases that degrade
molecules, resulting in collagen fragments detected in both
normal and diseased cartilage [3].
Since collagen is a main extracellular matrix (ECM)
component in many tissues, the effects of degraded collagen
(collagen hydrolysate or CH) have been widely studied. To
prepare CH, collagen is extracted from different sources (e.g.,
bovine, porcine, marine or fresh water fish) and organs (e.g.,
skin, scale, fish-frame or waste byproducts) by acid extraction
and heat treatment to produce a gelatin and transformed by
enzymatic digestion into collagen peptide or CH, which is a
mixture of 3–10 kDa peptides [4]. CH has been used as a
supplementary food and a therapeutic agent due to many
beneficial biological activities. In the past, sources of CHs
were porcine and bovine, however, there were limitations
because of a religious reason and effects of a toxin from
bovine spongiform encephalopathy. Thus, recently a marine
or fresh water fish has become an attractive source.
Articular cartilage is a dense connective tissue that involved
in weight bearing and joint movement. It is composed of three
layers; superficial, middle and deep zone, where living
chondrocytes are distributed. Chondrocytes produce extra-
cellular molecules including collagen, non-collagen proteins,
degrading enzymes and cytokines [5]. When chondrocyte
catalytic processes exceed anabolic processes, it leads to
osteoarthritis (OA), a degenerative joint disease which causes
destruction of aggrecan molecules in the early stage of OA,
and collagen destruction in the late stage [6]. When inflam-
mation is present, many cytokines are produced by synovio-
cytes and chondrocytes, such as IL-6, TNF-α and IL-1β [7]. In
particular, numerous studies of IL-1β have reported that IL-1
induced MMP13 promoter activity [8]. Osteoarthritic chon-
drocytes responded to IL-1 activation via JNK, p38 and NFκB
pathways. Normal chondrocytes can also respond to exogen-
ous IL-1 activation via the same pathways [9].
Many studies have reported about different effects of
CH, focusing on cartilage and its disease (OA). In many
clinical research reports, it shows that 10 g/day CH inges-
tion can relieve pain in OA patients [10]. Steffen Oesser and
colleagues were the first group to report about the
Cell Biochem Biophys
absorption, accumulation and action of CH. Bovine CH
(3.1–3.3 kDa, FORTIGAL, Gelita AG) can stimulate col-
lagen and proteoglycan synthesis in chondrocytes [11–13].
Kenneth and colleagues discovered that 3 kDa of bovine
CH (Gelita) elevated sulfated glycosaminoglycan (sGAG)
and collagen contents in chondrocyte-seeded agarose
hydrogels [14]. On the other hand, fish collagen peptide (3
kDa) showed the chondroprotective effect in combination
with glucosamine in a rat ACLT model [15]. Fish CH
induced the proteoglycan production in OA mice model
better than porcine CH [16]. Moreover, Steinmeyer and
colleagues compared the effects of different sources of CH
with a similar molecular weight distribution (approximately
2.9–3.7 kDa) on human OA cartilage. They found that none
of the CH stimulated collagen synthesis, but the porcine CH
increased collagen degradation while bovine and fish CH
had no effect. However, MMP1, MMP3 and MMP13 were
up-regulated by both bovine and fish CH [17]. Therefore,
size of CH may effect on chondroprotection and ECM
production or degradation. However, the effects of different
sizes of fish CH on cartilage metabolism and their molecular
mechanisms have not previously been investigated. Thus,
this study investigated the effects of different sizes of fish
CH on both physiological and pathological conditions.
Materials and Methods
Reagent
Tissue culture medium and related reagents were purchased
from Gibco (Grand Island, NY, United States). Recombi-
nant human IL-1β was purchased from Peprotech (Rocky
Hill, NJ, United States) and recombinant human OSM was
purchased from R&D Systems (Minneapolis, MN, United
States). An RNA extraction kit was purchased from GE
Healthcare (Buckinghamshire, United Kingdom). A Tetro
cDNA™ synthesis kit and 2x Sensifast™ SYBR Lo-Rox
mix were purchased from Bioline (London, United King-
dom). The antibody against MMP3 (ab53015) antibody was
purchased from Abcam (Cambridge, United Kingdom). The
antibodies against type II collagen (MAB8887) and
MMP13 (MAB3321) were purchased from MERCK
(Temecula, CA, United States). Antibodies recognizing the
MAPK pathway and NF-κB signaling were purchased from
Cell Signaling (Danvers, MA, United States). Cocktail
proteinase inhibitor and phosphatase inhibitor were pur-
chased from Roche (Mannheim, Germany). The Super-
Signal ECL Chemiluminescent Kit was purchased from
Thermo Scientific (Rockford, IL, USA). FCH was pur-
chased from Hainan Huayan Biotech Co., Ltd. (Haikou,
China). Other reagents were purchased from Sigma-Aldrich
and Invitrogen.
Cell Biochem Biophys
Table 1 Characteristic of fish collagen hydrolysate
Items Percentage
Protein 94.31
Ash 0.19
Moisture 5.53
Fat 0.00
Collagen/protein 99.87
Preparation of Different Sizes of FCH
FCH was purchased from Hainan Huayan Biotech Co., Ltd.
The FCH was extracted from the skin and scales of tilapia
fish (Oreochromis niloticus) and manufactured through
enzymatic engineering and spray drying [18]. The char-
acteristic data of FCH was obtained from the supplier as
shown in Table 1.
FCH (10 g) was dissolved in 100 ml of ultrapure water
and separated by continual centrifugation with Amicon™
(10 kDa and 3 kDa cut-off membranes). The solution on top
of the 10 kDa cut-off membrane was designated as fraction3
(F3; MW > 10 kDa). The solution which penetrated this
membrane was re-centrifuged with a 3 kDa cut-off mem-
brane. The solution retained on the 3 kDa cut-off membrane
was collected and designated fraction2 (F2; MW 3–10
kDa). A final fraction was collected from the solution which
passed through the 3 kDa cut-off membrane and designated
fraction1 (F1; MW < 3 kDa). For characterization, all frac-
tions were loaded on 4–12% Bris-Tris minigel (Invitrogen)
under reducing conditions, the separated protein on the gel
was stained with Page-Blue reagent. The hydroxyproline
content in all three fractions were also investigated.
Measurement of hydroxyproline in separated FCH
Pyrrole oxidation of the samples was done with chloramine-
T at pH 6. This complex produced a pink color with 4-
dimethylaminobenzaldehyde. Samples were added with a
diluent solution (67% propanol-2-ol), an oxidant solution
(50 mM chloramine-T) and a color reagent (7.5% dime-
thylamino benzaldehyde in propan-2-ol) [19]. The reaction
was performed at 60 °C for 45 min and light absorbance
was measured at 540 nm (pink color) using a microtiter
plate reader (MULTISKAN Ex, Thermo Scientific).
Porcine Cartilage Explants and Treatment
The metacarpo-phalangeal and metatarso-phalangeal joints
from 20–24 week-old pigs were dissected for the articular
cartilage explants model [20]. After dissection, the porcine
explants were incubated with 100 Units/ml penicillin and
100 Units/ml streptomycin for 30 min, after that they were
cultured in DMEM at 37 °C and 5% CO2 overnight. Three
explant pieces, 10 mg each, were randomly selected and
cultured in a 24-well culture plate. The conditioned media
was collected and replaced at day 0, 7, 14, 21 and 28 to
detect sGAG release using 1,9-dimethylmethylene blue
(DMMB) colorimetric assay. All conditioned media were
kept at −20 °C. For evaluation of the physiological condi-
tion, DMEM was used as control and 100 μg/ml of each
FCH fraction (F1 or F2 or F3) was used as treatment. For
the pathological condition, DMEM was also used as a
control. Additionally, 10 ng/ml IL-1β and 10 ng/ml OSM
were used as combinative cytokines to induce cartilage
ECM molecule degradation. The effect of each fraction was
investigated by co-treatment of the cytokines with 100 μg/
ml of each FCH fraction (F1 or F2 or F3). The remaining
and degradation levels of ECM molecules including sGAG
and type II collagen were analyzed.
Measurement of sGAG released in conditioned media
To measure the level of sGAG degradation, the release of
sGAG from cartilage into the culture media was quantified.
In brief, sGAG was detected by DMMB reagent [21].
Chondroitin 6-sulfate (CS-C) in a range of 0–40 μg/ml was
used as a reference standard. Fifty micro litters of CS-C or
diluted samples were added to a 96 well-plate followed by
DMMB reagent. The reaction produced a purple to pink
color; light absorbance was measured at 540 nm by
microtiter plate reader (MULTISKAN Ex, Thermo-
Scientific). %sGAG release was calculated using the for-
mula: %sGAG release = ((sGAG Dsample−sGAG D0)/
sGAG D0)*100.
Histological study
To study the remaining sGAG and type II collagen, the
cartilage explants were used to prepare section for histology
and immunohistology, respectively. The explants were
constructed, embedded and sectioned into 5 μm paraffin-
embeded sections. Before histological or immunohistolo-
gical staining, the paraffin-embedded sections were depar-
affinized by xylene and serially dehydrated using absolute
ethanol, 95% ethanol and 70% ethanol. To detect sGAG,
the sections were submerged in 1% (w/v) safranin-O in
distilled water for 10 min. After that, excess dye was
eliminated using distilled water followed by rehydration and
mounting at room temperature. For immunohistochemistry
as previously described [22, 23], in brief, the sections were
prepared in a moist chamber then digested with 0.25%
trypsin for 15 min at 37 °C, 0.25 Units of chondroitinase
ABC for 15 min at 37 °C, 1.45 Units testicular hyalur-
onidase for 15 min at 37 °C and 1 mg/ml pepsin for 15 min
at 37 °C. Then they were incubated in 3% H2O2 in methanol

for 5 min and washed with phosphate buffered saline (PBS).
After that, they were blocked with 5% bovine serum albu-
min (BSA) in PBS for 30 min. Primary antibody, anti-
human type II collagen mouse monoclonal antibody
(MAB8887, MERCK) in a ratio of 1:100 was added cover
the section and incubated at 4 °C overnight. After the pri-
mary antibody washing step, horse radish peroxidase-
conjugated IgG anti-mouse antibody (1:25) was added to
cover the section at room temperature for 1 h. A brown
color was developed from using DAB reagent (Invitrogen).
The stained sections were photographed using a Zeiss
microscope camera at 5 and 20x magnification.
Chondrocyte Isolation, Culture and Treatment
Primary chondrocytes were isolated from non-inflammatory
human articular cartilage which had been obtained after
arthroscopic diagnosis of flat-pad syndrome patients at
Maharaj Nakorn Chiang Mai Hospital with fully informed
patient consent (ethics approval code 070CT111016). The
cartilage was digested with 0.5% type IA collagenase in
serum-free DMEM for 24 h at 37 °C with agitation. Cells
were then washed with PBS and cultured in DMEM con-
taining 10% fetal bovine serum (Gibco) at 37 °C in 5% CO2
[24]. When the cells reached 80% confluence, they were
stocked in liquid nitrogen or expanded to passage2 and 3.
At passage4, cells were used in further experiments.
Before treatment, primary chondrocytes were cultured
overnight in serum-free medium. For evaluations of gene
expression and protein production, cells were treated with 1,
10 and 100 μg/ml of each fraction for 24 h for the physiolo-
gical condition, cells were co-treated with 5 ng/ml IL-1β and
1, 10 and 100 μg/ml of each fraction for 24 h for the patho-
logical condition. For signaling experiment, cells were treated
with 1, 10 and 100 μg/ml of each fraction for 24 h for the
physiological condition, cells were pre-treated with 1, 10 and
100 μg/ml of each fraction for 2 h and followed by 5 ng/ml of
IL-1β induction for 15 min for the pathological condition.
Measurement of Gene Expression
The effects of FCHs on chondrocyte gene expressions in the
physiological and pathological conditions were investi-
gated. After treatment, cell lysates were collected with 1%
β-mercaptoethanol in lysis buffer. mRNA was extracted
using Illustra™ RNASpin Mini RNA Isolation Kits (GE
Healthcare). Five hundred nanaograms of mRNA were
converted to cDNA using a Tetro cDNA Synthesis Kit
(Bioline). 0.5 μg of cDNA was used to amplify interested
gene with 2x Sensifast™ SYBR Lo-Rox Mix (Bioline)
using a 7500 Fast Real-Time PCR System (Applied Bio-
systems). Reactions were performed for 40 cycles with 95 °C
for denaturation, 60 °C for annealing and 72 °C for
Cell Biochem Biophys
extension. The data were collected and the fold changes in
gene expression following stimulation in the presence or
absence of IL-1β or FCHs was calculated using the 2−ΔCt
method [25]. The sequences of primers for real time reverse
transcription polymerase chain reaction (RT-PCR) were;
GAPDH F: 5′-AGGGCTGCTTTTAACTCTCGT-3′, R: 5′-C
CCCACTTGATTTTGGAGGGA-3′. SOX9 F: 5′-CATGAG
CGAGGGCACTCC-3′, R: 5′-TCGCTTCAGGTCAGCCT
TG-3′. COL2a1 F: 5′-GGTGGCTTCCATTTCAGCTATG-
3′, R: 5′-TTGCAGTGGTAGGTGATGTTCTG-3′. COL1a1
F: 5′-CAGCCGCTTCACCTACAGC-3′, R: 5′-TTTTGT
ATTCAATCACTGTCTTGCC-3′. MMP3 F: 5′-TTTTGG
CCATCTCTTCCTC-3′, R: 5′- TGTGGATGCCTCTTGGG-
TAT-3′. MMP13 F: 5′-TCCCAGGAATTGGTGATAAAG
TAGA-3′, R: 5′- CTGGCATGACGCGAACAATA-3′.
Western Blot
To determine MMPs production in conditioned media, the
conditioned media were concentrated by centrifugation with
Amicon™ (10 kDa cut-off memebrane, MERCK) at 3500 g
for 20 min at 15 °C. To determine chondrocyte signaling,
cells were collected with 250 μl of radio-
immunoprecipitation assay buffer (1 M HEPES, 10% (v/v)
NP-40, 0.5% (w/v) sodium deoxycholate, 4 M NaCl, 0.5 M
NaF, 100 mM sodium orthovanadate, cocktail proteinase
inhibitor and phosphatase inhibitor) followed by standing
on ice and vertex every 15 min for four times after which
they were centrifuged at 14,000 g for 10 min at 4 °C. The
supernatants were collected and protein level was estimated
using the Bio-Rad protein assay based on the Bradford
method. Concentrated media or cell lysate supernatant (10
μg protein) was analyzed using 12% SDS-PAGE [26]. After
electrophoresis, samples were transferred to a nitrocellulose
membrane (MERCK) with a glycine buffer (20 mM Tris
base, 0.2 M glycine, 20% (v/v) methanol) by semi-dry
Western blotting. The conditioned media were blocked with
3% BSA in PBS while cell lysate was blocked with 5%
skim milk in PBS. Then the proteins of interest were probed
overnight at 4 °C with either specific antibody against
MMP3 (1:500) and MMP13 (1:500) in PBS-containing
0.05% tween-20 or with specific antibody against MAPK
(1:1000) or NFκB (1:1000) in 1% skim milk in PBS. After
primary antibody incubation, the membranes were washed
with PBS containing 0.05% tween-20 and incubated with
secondary antibody (horseradish peroxidase-conjugated
goat anti-mouse IgG (1:1000) or horseradish peroxidase-
conjugated goat anti-rabbit IgG (1:1000)) for 1 h. The
specific bands were detected with a SuperSignal ECL
Chemiluminescent Kit (Thermo Scientific) and exposed in a
Gel Doc Universal Hood II (BioRad). Band density was
calculated using TottalLab TL20 analysis software.
Cell Biochem Biophys
Data and Statistical Analysis
The results of all experiments were expressed as mean ± SEM
of triplicate independent experiments. One way ANOVA for
multiple comparisons; Turkey’s multiple comparison test or
LSD was used to determine differences which were con-
sidered to be significant at P values less than 0.05.
Results
Preparation and Characterization of Different Sizes of
FCH
Ten grams of FCH from Hainan Huayan Biotech Co., Ltd.
were dissolved in 100 ml ultrapure water and separated by
centrifugation with Amicon™ (a 10 kDa cut-off mem-
brane). The retained solution above the membrane was
collected for use as Fraction3 (F3, MW > 10 kDa). The
solution which passed through the 10 kDa cut-off mem-
brane was re-centrifuged with a 3 kDa cut-off membrane.
The solution retained above the membrane was collected as
Fraction2 (F2, MW 3–10 kDa) while the solution passed
through the membrane was collected as Fraction1 (F1, MW
Fig. 1 Characterization of different sizes of Fish Collagen Hydrolysate
(FCH). After separation of FCH by centrifugation, the sizes of all
fractions were characterized by electrophoresis and stained with Page-
Blue reagent. Lane1: low-range protein marker, Lane2: desalted FCHs,
Lane3: Fraction3 (MW > 10 kDa), Lane4: Fraction2 (MW 3–10 kDa)
and Lane 5: Fraction1 (MW < 3 kDa)
Table 2 Hydroxyproline content in each fraction of fish collagen hydrolysate
Fraction Dry weight of total (g)
Fraction1: F1 (MW < 3 kDa) 0.3 ± 0.0035
Fraction2: F2 (MW 3–10 kDa) 1.1 ± 0.0038
Fraction3: F3 (MW > 10 kDa) 8.0 ± 0.0128
Total 9.4% ± 0.02 (94% remaining)
< 3 kDa). The sizes of all three fractions were then char-
acterized using electrophoresis and stained with Page-Blue
reagent. The molecular weight pattern is shown in Fig. 1.
Fraction3 (Lane 3) contained large size of FCH with a MW
more than 12 kDa. Fraction2 (lane 4) contained middle size
of FCH with MW of 3.5–12 kDa. Fraction1 (lane 5) con-
tained small size of FCH with less than 3.5 kDa. All frac-
tions were examined for hydroxyproline content as shown
in Table 2. Using this molecular weight cut-off technique,
the amount of F3 had been highest obtained followed by F2
and F1, respectively. The hydroxyproline content per total
weight of all fraction was not different.
Effects of Different MWs of FCH on Porcine Cartilage
Explants
To examine the effect of FCHs on cartilage metabolism, a
porcine cartilage explants model was used. To investigate their
effects in physiological condition, the explants were treated
directly with FCH. For the pathological condition, they were
treated with 10 ng/ml IL-1β and 10 ng/ml OSM to induce an
OA condition, then the FCHs were co-treated. The explants
were incubated in the indicated conditions for 28 days. The
media were collected and replaced every 7 days for exam-
ination of sGAG release. Finally, these paraffin explants sec-
tions were histologically analyzed for aggrecan and collagen
remainings by safranin-O staining and immunohistochemistry
using anti-type II collagen antibody, respectively.
In the physiological condition, it was found that 100 μg/
ml of all three fractions did not affect sGAG release when
compared to the control (Fig. 2c). The results were con-
sistent with the remaining sGAG as shown by safranin-O
staining (Fig. 2a). Compatible with type II collagen
remaining, all three fractions had no significant effect on
type II collagen remaining compared to the control (Fig.
3a). It was suggested that all three fractions had no effect on
type II collagen and sGAG contents in physiological con-
dition. In the pathological condition, IL-1β and OSM were
used to induce the degradation of collagen and sGAG. The
cytokine treatment significantly increased sGAG release (2-
fold higher compared to the control on day 7 and continued
to induce release on subsequent days) (Fig. 2d). The sGAG
remaining in this condition had decreased as shown by
Percentage of total Total Hydroxyproline (mg) Hydroxyproline (mg/g)
3.2 ± 0.037 2.12 ± 0.0271 7.05 ± 0.0903
11.7 ± 0.040 8.29 ± 1.6123 7.54 ± 1.4658
85.6 ± 0.136 59.20 ± 15.1347 7.39 ± 1.8918
100% – –
 Cell Biochem Biophys

Fig. 2 sGAG release into media

and sGAG remaining in porcine
cartilage explants. After 28 days
of treatment, explants were
obtained for preparation of
sections. The sections were
stained with safranin-O in both
physiological condition (a), and
pathological condition (b)
(microscopic view of cartilage
explants at 5x magnification).
Levels of accumulation of
sGAG in physiological
condition (c) and pathological
condition (d). Bars on graph
showed mean ± SEM. The
statistic was analyzed by
ANOVA with multiple
comparisons Turkey that p <
0.05, * when compared to
control group in each day,
#
when compared to IL/OSM
group in each day

safranin-O staining (Fig. 2b). F1 showed an additive effect
with IL-1β and OSM, inducing sGAG release in all terms of
treatment, whereas F2 only induced release in the early
stage of treatment and F3 showed no effect (Fig. 2d). The
results of F1 and F2 effects were consistent with sGAG
remaining (Fig. 2b). For sGAG remaining, F1 and
F2 showed additive effect with IL-1β and OSM as shown
by paler stained color at the bottom of the cartilage piece
when compared to the color of IL-1β and OSM control (Fig.
2b). However, F3 could slightly increase sGAG remaining
when compared to IL-1β and OSM control. Since it could
not inhibit sGAG release induced by IL-1β and OSM (Fig.
2b), this result might due to its ability to increase sGAG
synthesis. For collagen, it was found that IL-1β and OSM
slightly decreased the amount of type II collagen remaining.
All three fractions did not show any effect on collagen
remaining when compared to the IL-1β and OSM treatment
(Fig. 3b).
Effects of Different MWs of FCH on Chondrocyte Gene
Expression
The porcine cartilage explant results indicated that different
sizes of FCH showed different effects on cartilage. Next, the
Cell Biochem Biophys
Fig. 3 Type II collagen
remaining in porcine cartilage
explants. The section of cartilage
explants in each condition was
stained with anti-collagen type II
antibody. a physiological
condition and b pathological
condition which induced by
10 ng/ml IL-1β and 10 ng/ml
OSM. Microscopic view of
cartilage explants at 20x
magnification
mechanism of each fraction on chondrocyte metabolism
was studied in human articular chondrocytes (HACs). To
investigate the physiological condition, HACs were treated
with different concentrations of each fraction for 24 h. In the
pathological condition, the cells were co-treated with 5 ng/ml
IL-1β and various concentrations of each fraction. The
expression of anabolic genes including SOX9 which is a
transcription factor for type II collagen, COL2a1, and
COL1a1 as well as catabolic genes including MMP3 and
MMP13 were investigated using real time RT-PCR. The
expressions of COL2a1 and COL1a1 was shown as
COL2a1/COL1a1 ratio which the value inversely indicated
dedifferentiation of chondrocytes.
Firstly, the toxicity of FCH (1–100 μg/ml) was examined
by MTT assay. The results showed that all conditions had
no toxic effect to HACs (data not shown). In the physio-
logical condition, 1–100 μg/ml of all 3 fractions had no
effect on the expression of SOX9, and COL2a1/COL1a1
ratio (Fig. 4a). In part of protease expressions, MMP3
expression was significantly increased by 1 μg/ml F1 (2.5-
fold), 10 μg/ml F3 (7.5-fold), and 100 μg/ml F3 (7.5-fold).
Similarly, MMP13 expression was significantly increased
by 100 μg/ml F3 (3-fold) and slightly up-regulated by F1
(1.5-fold). Nevertheless, F2 (1–100 μg/ml) had no effect on
either MMP3 or MMP13 expression (Fig. 4a).
In the pathological condition, the results are shown in Fig.
4b. IL-1β could significantly reduce both the expression of
SOX9 and the COL2a1/COL1a1 ratio. The reduction of
COL2a1/COL1a1 ratio indicated dedifferentiation of HACs.
All three fractions had no effect on both SOX9 and the
COL2a1/COL1a1 ratio when compared to IL-1β control. In
the case of MMP3 and MMP13, 5 ng/ml IL-1β could increase
expressions of MMP3 (about 75-fold) and MMP13 (about 25-
fold) when compared to the control. All three fractions had no
significant effect on MMP3 expression when compared to IL-
1β control. However, F2 and F3 showed an additive effect
with IL-1β on induction of MMP13 expression (2.5-fold and
2.5-fold compared to IL-1β control). Taken together, in the
physiological condition, both F1 and F3 could induce the
expression of MMP3 (2.5 and 7.5-fold, respectively) whereas
only F3 could induce the expression of MMP13 (2.5-fold).
However, these inductions were uncomparable with the effect
of IL-1β. In the pathological condition, IL-1β (5 ng/ml)
induced the expressions of MMP3 (75-fold) and MMP13 (25-
fold), and F2 and F3 showed an additive effect with IL-1β on
inducing expression of MMP13.
Effects of Different MWs of FCH on Catabolic Enzyme
Release by Primary HACs
None of the FCH fractions affected anabolic gene expression,
but they affected catabolic gene expression (MMP3 and
MMP13) in both the physiological and pathological condi-
tions. Therefore, this experiment focused on the secretions of
MMP3 and MMP13 in conditioned media.
In the physiological condition, MMP3 levels are shown in
Fig. 5a (upper panel). Only 54 kDa, indicated as pro-MMP3,
was detected. In comparison to the control, it was found that
1 μg/ml F1, 100 μg/ml F2 and 1–100 μg/ml F3 significantly
increased pro-MMP3 level. A maximum induction carried
out by F3 was 2.5-fold compared to the control as shown in
Fig. 5c. Results of MMP13 were shown in Fig. 5a (middle
panel). Two bands were detected, 60 kDa for pro-MMP13
and 48 kDa for active-MMP13. Most detected band was pro-
MMP13. Comparing to the control, all three fractions sig-
nificantly increased pro-MMP13 production.
In the pathological condition, it was found that 5 ng/ml
IL-1β significantly increased MMP3 production level (Fig. 5b
upper panel), especially pro-form (54 kDa). In the combi-
nation of each fraction with IL-1β, F1 slightly reduced the
effect of IL-1β on pro-MMP3 release, whereas 100 μg/ml
F2 and 10–100 μg/ml F3 significantly induced the release of
pro-MMP3 compared to IL-1β control (Fig. 5b upper
panel). For MMP13 (Fig. 5b middle panel), active-MMP13
(48 kDa) was obviously detected after IL-1β induction. F1
and F3 did not show any additive effect with IL-1β on
induction of active-MMP13 secretion. However, 10–100
µg/ml F2 could significantly increase the effect of IL-1β on
induction of active-MMP13 secretion.

Fig. 4 Chondrocyte gene expressions. a Cells were treated with
1–100 μg/ml of each fraction of FCH for 24 h in physiological con-
dition. b Cells were co-treated with 5 ng/ml IL-1β and 1–100 μg/ml of
each fraction of FCH for 24 h in pathological condition. SOX9 and
COL2a1/COL1a1 ratio are anabolic gene markers, while MMP3 and
Effects of Different MWs of FCH on Signal
Transduction in Primary HACs
Based on gene and protein expression, different sizes of
FCH had different effects on the expression and secretion of
MMP3 and MMP13. Since, IL-1β induces MAPK and
NFκB cascades in OA cartilage and in primary chondrocyte
[9] resulting in the up-regulation of MMP3 and MMP13
expression [8, 27]. Therefore, activations of MAPK,
Cell Biochem Biophys
MMP13 are catabolic gene markers. Bars on graph showed mean ±
SEM. The statistic was analyzed by ANOVA with multiple compar-
isons Turkey that p < 0.05, * when compared to control group, # when
compared to IL-1β group
including p38, JNK, ERK and NFκB signalings were
examined by Western blotting.
In the physiological condition, after 24-h incubation, it
was found that F2 could activate p-p38, whereas F3 acti-
vated p-ERK and p-p38. F1 could not activate any of those
signaling and all three fractions had no effect on JNK
activation (Fig. 6a, c). Moreover, NFκB signaling was not
changed by any of the treatments (Fig. 6b, c). In the
pathological condition, chondrocytes were pre-treated with
Cell Biochem Biophys
Fig. 5 MMP3 and MMP13 secretions into cultured media from pri-
mary chondrocytes. Conditioned media were collected after 24 h
incubation. a Cells were treated with 1–100 μg/ml of each fraction
of FCH in the physiological condition. b Cells were co-treated with
5 ng/ml IL-1β and 1–100 μg/ml of each fraction of FCH in the
each fraction for 2 h prior to induction with 5 ng/ml IL-1β
for 15 min. It was revealed that IL-1β treatment activated p-
p38 and p-JNK pathways but none of the FCH fractions
affected these effects of IL-1β (Fig. 6d, f). For NFκB, IL-1β
clearly stimulated p-p65 when compared to the control. In
addition, 100 μg/ml F1, F2 and 10, 100 μg/ml F3 showed an
additive effect with IL-1β on inducing p-p65 activation
(Fig. 6e, f). These data indicated that FCHs affected chon-
drocyte signaling in the different way from IL-1β did.
Moreover, all three fractions did not show any additive
effect with IL-1β on p-p38 and p-JNK activations, but they
did show an additive effect on p-p65 activation.
Discussion
OA is a degenerative joint disease which is related to aging,
trauma and inflammation. Regarding inflammation, IL-6,
pathological condition. c Bar graphs of band densities showed mean ±
SEM. The statistic was analyzed by ANOVA with multiple compar-
isons LSD that p < 0.05, * when compared to control group, # when
compared to IL-1β group
TNF-α and IL-1β are dominant cytokines in OA. IL-1β in
particular can activate MMP13 promoter [8], and can also
induce the JNK and p38 MAPK pathways as well as the
NFκB pathway [9] resulting in increased severity of OA
pathogenesis. One of the alternative medicines for OA is the
consumption of supplementary food. Many researches
indicated that 10 g oral administration of CH was a ther-
apeutic dose to help relieve the pain of joint disease [10].
CH is a mixture of collagen peptides with around 3–10 kDa
final product size. Over the past decade, the main sources of
CHs have been bovine and porcine. After an oral admin-
istration of bovine CH at 10 mg/g bodyweight (6.1–6.8
kDa), the accumulation of bovine CH was detected in
mouse cartilage [28], suggesting that administration of CH
could be a nutrient source to synthesize cartilage ECM.
Additionally, an in vitro study found that bovine and
chicken CH (3.3 and 3.1 kDa, respectively) could increase
collagen synthesis in bovine chondrocytes [29]. Bovine CH

(3–3.7 kDa) could also increase sGAG and collagen content
in normal condition of bovine chondrocyte monolayers
under reduced O2 condition [11–13]. On the contrary,
Fichter and colleagues reported that type II collagen frag-
ments from bovine cartilage, which was less than 10 kDa,
could induce MMP3 and MMP13 expression together with
Fig. 6 MAPK and NFκB signaling in human articular chondrocytes.
Cells were treated with 1–100 μg/ml of each fraction of FCH for 24 h
in the physiological condition. a MAPK and b NFκB signaling were
detected by Western blotting and c bar graphs of band densities. Cells
Cell Biochem Biophys
MMP2 and MMP9 activity in chondrocyte cultivation [3].
Moreover, a comparative study of different sources of
similar size CH (2.9–3.7 kDa) in human OA cartilage dis-
covered that porcine CH could increase collagen degrada-
tion whereas bovine CH inhibited collagen synthesis and
fish CH had no effect [17]. Thus, different effects on
were pre-treated with 1–100 μg/ml of each fraction of FCH for 2 h and
then co-treated with 5 ng/ml IL-1β for 15 min in the pathological
condition. d MAPK and e NFκB signaling were examined and f bar
graphs of band densities
Cell Biochem Biophys
Fig. 6 (Continued)
cartilage metabolism resulted from different sizes and
sources of CH and its precise effect was still unclear.
Recently, marine and fresh water fish have become alter-
native sources for CH production according to religious
restrictions and to avoid potential toxic poisoning from cow
and pig. However, the effects of different MWs of fish CHs
on cartilage and chondrocyte metabolisms have not been
investigated yet. Thus, in this study we examined those
aspects in both the physiological and the pathological
conditions.
To examine the effect of FCHs on cartilage metabolism,
a porcine cartilage explants model was used. Three different

sizes of FCH (<3 kDa, 3–10 kDa and >10 kDa) were
prepared by centrifugation with a cut-off membrane. In
the physiological condition, it was found that all three
sizes FCH did not affect sGAG and collagen content of
the explants. This agrees with Steinmeyer’s work that
small size of fish CH (2.9–3.6 kDa) had no effect on
collagen synthesis in normal human cartilage [17]. The
effect of other sizes of FCH, however, has not been
reported. Our investigation of the FCH effect on cartilage
metabolism in the physiological condition showed that it
was quite different from the effect of bovine CH. In 2013,
Schadow and colleagues reported that small size of bovine
CH (3.3–3.5 kDa) at high dose (0.5 and 2 mg/ml) could
inhibit collagen synthesis in human cartilage, but it did
not induce collagen degradation. In the pathological
condition, IL-1β and OSM were used and they sig-
nificantly reduced sGAG content and slightly reduced
collagen content. The small size (F1, MW < 3 kDa) and
medium size FCH (F2, MW 3–10 kDa) had an additive
effect with IL-1β and OSM in reducing sGAG content
whereas the large size (F3, MW > 10 kDa) did not.
However, all three sizes FCH did not show any effect on
collagen content. Similar to previous reports, these data
suggest that different sizes and sources of CHs show
different effects on cartilage metabolism.
Next, to address molecular metabolism of each FCH
fraction, HACs were used as a model. In the physiological
condition, it was discovered that all three sizes of FCH did
not affect all interested anabolic gene expressions (SOX9
and COL2a1/COL1a1 ratio). Interestingly, small and large
fractions had an effect on MMP3 and MMP13 gene
expression. For those MMPs secretion, all three sizes of
FCH could up-regulate pro-MMP3 and pro-MMP13
secretions but not active forms. Markedly, medium size
FCH did not show any effect on MMP3 or MMP13 gene
expression, but it induced both pro-MMP3 and pro-MMP13
secretion. These apparently contradictory results might be
due to the ability of medium size FCH to inhibit pro-MMP3
and pro-MMP13 degradation. To date, there have been only
a few studies reporting on the effect of FCH on the
expression of MMP3 and MMP13. However, there have
been numerous studies on the effect of bovine CH on
chondrocyte metabolism, particularly on small size. In
2007, Kenneth and colleagues found that 1 mg/ml of bovine
CH (3.3 kDa) could increase MMP3 and MMP13 gene
expression in chondrocyte-seeded agarose hydrogel model
[14]. This result corresponds to our findings even though
the CH source was different from our source. Altogether, in
the physiological condition, all three sizes of FCH could
induce both pro-MMP3 and pro-MMP13 secretions by
chondrocytes. However, these pro-MMPs were not con-
verted to their active-forms resulting in the lack of effect on
cartilage matrix contents.
Cell Biochem Biophys
For the pathological results, all three sizes FCH had no
effect on interested anabolic gene expressions when com-
pared to IL-1β control. Conversely, IL-1β could sig-
nificantly increase MMP3 and MMP13 gene expression.
The catabolic gene expressions induced by IL-1β were
much greater than that of FCH in the physiological condi-
tion. Regarding the effect of different sizes of FCH on
MMP3 gene expression, it was found that all three sizes did
not affect gene expression when compared to the IL-1β
control. The studies regarding MMP3 secretion showed that
pro-MMP3 was predominantly detected. Medium and large
size FCH significantly showed an additive effect with IL-1β
on induction of pro-MMP3 secretion. The explanation of
this apparently contradictory result could be the same as for
medium size FCH in the physiological condition. For
MMP13, medium and large size FCH showed an additive
effect with IL-1β on the induction of MMP13 gene. How-
ever, only medium size FCH showed an additive effect with
IL-1β on induction of active-MMP13 secretion but not large
size FCH. Consistent with the explant result, medium size
FCH significantly showed additive effect on cartilage
degradation with the used cytokines, whereas large size did
not show the effect. This might indicated the different roles
of different sizes of FCH on regulation of post-
transcriptional modification of MMP13. Moreover, small
size FCH did not show any additive effect with IL-1β on
both gene and active-MMP13 protein expression. However,
it showed an additive effect with IL-1β and OSM on car-
tilage degradation in the explants model. Thus, this small
size FCH might induce other degrading enzymes except for
MMP3 and MMP13.
The different sizes of FCH had different effects on gene
and protein expression in both physiological and patholo-
gical studies on the chondrocyte metabolism. Their ability
to induce cellular signaling is an interesting area for
investigation. Many studies [8, 27] reported that IL-1β
could induce MMP3 and MMP13 gene expression via ERK,
JNK and p38 and NFκB stimulation pathways in chon-
drocytes monolayer culture. In this study, the activation of
MAPK and NFκB signaling pathways were examined. In
the physiological condition, medium-sized FCH could
activate p-p38 whereas large-sized could activate both p-
ERK and p-p38. Small size FCH had no effect on MAPK
activation and all three sizes of FCH had no effect on NFκB
pathways. Investigation of p-ERK and p-p38 levels found
that the levels activated by each fraction were not equal.
Large-sized FCH had greater activation ability than
medium-sized FCH. They were consistent with gene and
protein levels. The results indicated that medium and large-
sized FCH might activate the same pathway, p-p38,
resulting in induction of catabolic gene expressions and
finally pro-MMP3 and pro-MMP13 secretions.
Nevertheless, small-sized FCH did not show an ability to
Cell Biochem Biophys
activate those signaling pathways. It is possible that this size
might activate signaling pathways other than MAPK and
NFκB.
In the pathological condition, the signaling results
showed that IL-1β could induce p-p38 and p-JNK but not p-
ERK, and that all the three sizes of FCH had no additive
effect with IL-1β on those inductions. Moreover, IL-1β
could activate p-p65 and all the three sizes of FCH showed
an additive effect with IL-1β on the activation of this signal.
Although all the three sizes FCH showed the effect on
activation of p-p65, the gene and protein expression they
controlled were different. These data might indicate that
FCH could activate other signaling pathways in chon-
drocytes, such as PI3K/Akt signaling through the α2 A
domain of integrin [30, 31].
In conclusion, different sizes of FCH showed different
effects on chondrocyte metabolism in both physiological
and pathological conditions. In the physiological condition,
none of the 3 sizes FCH had an effect on porcine cartilage
degradation. Medium and large sizes could activate p-p38
whereas small size might be able to activate others pathway.
These three sizes could induce both pro-MMP3 and pro-
MMP13 protein. However, these pro-MMPs were not
converted to their active-form. In the pathological condi-
tion, the three sizes FCH showed an additive effect with IL-
1β to activate the p-p65, but not the MAPK pathway.
However, they showed different effects on cartilage
degradation. Small and medium-sized FCH showed an
additive effect with IL-1β and OSM, and further induced
cartilage degradation, but large size did not. Medium-sized
FCH showed this effect by further induction of active-
MMP13 secretion, whereas small sized FCH might induce
other enzymes. Therefore, FCH might also activate other
signaling pathways. Altogether, this study is the first report
of the effect of different sizes of FCH on chondrocyte
metabolism. Different sizes of FCH might induce the same
or different signaling pathways at different levels, which
resulted in different patterns of gene and protein expressions
in chondrocytes. In addition, different sources of CH might
also give different results.
Acknowledgements This study was supported by Thailand Excel-
lence Center for Tissue Engineering and Stem Cells, Department of
Biochemistry, Faculty of Medicine, Chiang Mai University and Center
of Excellence for Innovation in Chemistry (PERCH-CIC) for the
financial support. Moreover, we would like to thank Dr. Robert G
Lamar for his manuscript proofreading and Chiang Mai Graduate
School, Chiang Mai University.
Compliance with Ethical Standards
Conflict of Interest The authors declare that they have no compet-
ing interests.
References
1. Ricard-Blum, S. (2011). The collagen family. Cold Spring Harbor
perspectives in biology, 3, a004978.
2. Mark, K. V. D. (2006). Structure, biosynthesis and gene regula-
tion of collagens in cartilage and bone. In M. J. Seibel, S. P.
Robins, & J. P. Bilezikian (Eds.), Dynamics of bone and cartilage
metabolism (pp. 3–26). London: Elsevier.
3. Fichter, M., Korner, U., Schomburg, J., Jennings, L., Cole, A. A.,
& Mollenhauer, J. (2006). Collagen degradation products mod-
ulate matrix metalloproteinase expression in cultured articular
chondrocytes. Journal of Orthopaedic Research: Official pub-
lication of the Orthopaedic Research Society, 24, 63–70.
4. Wang, W., Li, Z., Liu, J., Wang, Y., Liu, S., & Sun, M. (2013).
Comparison between thermal hydrolysis and enzymatic proteo-
lysis processes for the preparation of tilapia skin collagen
hydrolysates. Czech Journal Food Science, 31, 1–4.
5. Goldring, M. B., Tsuchimochi, K., & Ijiri, K. (2006). The control
of chondrogenesis. Journal of Cellular Biochemistry, 97, 33–44.
6. Aigner, T., Zien, A., Gehrsitz, A., Gebhard, P. M., & McKenna,
L. (2001). Anabolic and catabolic gene expression pattern analysis
in normal versus osteoarthritic cartilage using complementary
DNA-array technology. Arthritis and Rheumatism, 44,
2777–2789.
7. Florián, A. M. (2011). IL-1 and its role in osteoarthritis. Open
Journal of Medicine, 1, 1–6.
8. Mengshol, J. A., Vincenti, M. P., & Brinckerhoff, C. E. (2001).
IL-1 induces collagenase-3 (MMP-13) promoter activity in stably
transfected chondrocytic cells: Requirement for Runx-2 and acti-
vation by p38 MAPK and JNK pathways. Nucleic Acids
Research, 29, 4361–4372.
9. Fan, Z., Soder, S., Oehler, S., Fundel, K., & Aigner, T. (2007).
Activation of interleukin-1 signaling cascades in normal and
osteoarthritic articular cartilage. The American journal of
Pathology, 171, 938–946.
10. Bello, A. E., & Oesser, S. (2006). Collagen hydrolysate for the
treatment of osteoarthritis and other joint disorders: A review of
the literature. Current Medical Research and Opinion, 22,
2221–2232.
11. Schunck, M., Schulze, C. H., & Oesser, S. (2006). P199 Disparate
efficacy of collagen hydrolysate and glucosamine on the extrcel-
lular matrix metabolism of articular chondrocytes. Osteoarthritis
and Cartilage, 14, S114.
12. Schunck, M., Schulze, C. H., & Oesser, S. (2007). P189 Collagen
hydrolysate supplementation stimulates proteoglycan metabolism
and gene expression of articular chondrocytes. Osteoarthritis and
Cartilage, 15, B136.
13. Schunck, M., & Oesser, S. (2010). 547 A combination of a spe-
cific collagen hydrolysate with a refined rose hip extract has a
positive effect on the extra cellular matrix maintenance of carti-
lage cells. Osteoarthritis and Cartilage, 18, S245.
14. Ng, K. W., Saliman, J. D., Lin, E. Y., Statman, L. Y., Kugler, L.
E., Lo, S. B., et al. (2007). Culture duration modulates collagen
hydrolysate-induced tissue remodeling in chondrocyte-seeded
agarose hydrogels. Annals of Biomedical Engineering, 35,
1914–1923.
15. Ohnishi, A., Osaki, T., Matahira, Y., Tsuka, T., Imagawa, T., &
Okamoto, Y., et al. (2013). Evaluation of the chondroprotective
effects of glucosamine and fish collagen peptide on a rabbit ACLT
model using serum biomarkers. Journal of Veterinary Medical
Science, 75, 421–429.
16. Ohara, H., Iida, H., Ito, K., Takeuchi, Y., & Nomura, Y. (2010).
Effects of Pro-Hyp, a collagen hydrolysate-derived peptide, on
hyaluronic acid synthesis using in vitro cultured synovium cells
and oral ingestion of collagen hydrolysates in a guinea pig model

of osteoarthritis. Bioscience, Biotechnology and Biochemistry, 74,
2096–2099.
17. Steinmeyer, J., Schadow, S., Kordelle, J., & Lochnit, G. (2012).
Effect of collagen hydrolysates on human osteoarthritic cartilage
explants. http://www.ors.org/Transactions/58/0697.pdf.
18. Hainan-Huayan Biotech Co. Ltd. (2005). Company pro-
files. http://hainan-huayan-biotech-co-ltd.imexbb.com/.
19. Kolar, K. (1990). Colorimetric determination of hydroxyproline as
measure of collagen content in meat and meat products: NMKL
collaborative study. Journal of Association of Official Analytical
Chemists, 73, 54–57.
20. Phitak, T., Pothacharoen, P., Settakorn, J., Poompimol, W.,
Caterson, B., & Kongtawelert, P. (2012). Chondroprotective and
anti-inflammatory effects of sesamin. Phytochemistry, 80, 77–88.
21. Farndale, R. W., Buttle, D. J., & Barrett, A. J. (1986). Improved
quantitation and discrimination of sulphated glycosaminoglycans
by use of dimethylmethylene blue. Biochimica et Biophysica Acta,
883, 173–177.
22. Vankemmelbeke, M., Dekeyser, P. M., Hollander, A. P., Buttle,
D. J., & Demeester, J. (1998). Characterization of helical clea-
vages in type II collagen generated by matrixins. The Biochemical
Journal, 330(Pt 2), 633–640.
23. Liu, F. C., Hung, L. F., Wu, W. L., Chang, D. M., Huang, C. Y.,
Lai, J. H., et al. (2010). Chondroprotective effects and mechan-
isms of resveratrol in advanced glycation end products-stimulated
chondrocytes. Arthritis Research & Therapy, 12, R167.
24. Flannery, C. R., Little, C. B., Hughes, C. E., & Caterson, B.
(1998). Expression and activity of articular cartilage
Cell Biochem Biophys
hyaluronidases. Biochemical and Biophysical Research Commu-
nications, 251, 824–829.
25. Livak, K. J., & Schmittgen, T. D. (2001). Analysis of relative gene
expression data using real-time quantitative PCR and the 2(-Delta
Delta C(T)) Method. Methods (San Diego, Calif.), 25, 402–408.
26. Tew, S. R., & Hardingham, T. E. (2006). Regulation of SOX9
mRNA in human articular chondrocytes involving p38 MAPK
activation and mRNA stabilization. The Journal of Biological
Chemistry, 281, 39471–39479.
27. Fan, Z., Yang, H., Bau, B., Soder, S., & Aigner, T. (2006). Role of
mitogen-activated protein kinases and NFkappaB on IL-1beta-
induced effects on collagen type II, MMP-1 and 13 mRNA
expression in normal articular human chondrocytes. Rheumatol-
ogy International, 26, 900–903.
28. Oesser, S., Adam, M., Babel, W., & Seifert, J. (1999). Oral
administration of (14)C labeled gelatin hydrolysate leads to an
accumulation of radioactivity in cartilage of mice (C57/BL). The
Journal of Nutrition, 129, 1891–1895.
29. Oesser, S., & Seifert, J. (2003). Stimulation of type II collagen
biosynthesis and secretion in bovine chondrocytes cultured with
degraded collagen. Cell and Tissue Research, 311, 393–399.
30. Yasuda, T. (2014). Type II collagen peptide stimulates Akt
leading to nuclear factor-kappaB activation: its inhibition by
hyaluronan. Biomedical Research, 35, 193–199.
31. Siebert, H. C., Burg-Roderfeld, M., Eckert, T., Stotzel, S., Kirch,
U., & Diercks, T., et al. (2010). Interaction of the alpha2A domain
of integrin with small collagen fragments. Protein & Cell, 1,
393–405.