Original Research

Collagen hydrolysate inhibits zymosan-induced inflammation

Anita Hartog1,2, Miranda Cozijnsen1,2, Gerrit de Vrij1 and Johan Garssen1,2

1
Danone Research, Centre for Specialised Nutrition, 6704 PH, Wageningen, the Netherlands; 2Department of Pharmacology &
Pathophysiology, Utrecht Institute for Pharmaceutical Sciences (UIPS), Utrecht University, 3584 CA, Utrecht, the Netherlands
Correponding author: Anita Hartog. Email: anita.hartog@danone.com

Abstract
During the past years, evidence accumulated showing that glycine comprises anti-inflammatory activities. These effects occur, at
least in part, via the activation of glycine-gated chloride channels (GlyR). Glycine is one of the major structural units of collagen,
making up about 30% of the amino acids. This study aims to investigate the anti-inflammatory potential of collagen hydrolysate
(CH) using the zymosan-induced ear-skin inflammation mouse model. After oral intake of 12.5, 25 or 50 mg CH the plasma levels of
glycine increased in a concentration-dependent manner. CH was able to counteract zymosan-induced ear-skin inflammation
locally (ear swelling) as well as systemically (IL-6 production by lipopolysaccharide (LPS)-stimulated whole blood cells). The LPS-
stimulated IL-6 production in whole blood correlated positively with the ear swelling response. This correlation was abolished by
strychnine (a glycine receptor antagonist), indicating the involvement of GlyR. Collectively, these data show that CH is able to
modulate inflammatory responses both locally as well as systemically. This effect might be constituted by inhibiting pro-inflam-
matory cytokine production via GlyR.

Keywords: Collagen hydrolysate, GlyR, Glycine, inflammation

Experimental Biology and Medicine 2013; 238: 798–802. DOI: 10.1177/1535370213480740

Introduction intestine, a decrease in pro-inflammatory cytokine expres-

sion in the fat tissue of lean and obese mice, the inhibition of
The simple nonessential amino acid glycine is found in
peptidoglycan polysaccharide-induced arthritis, and the
many different proteins. It acts as an inhibitory neurotrans-
inhibition of zymosan-induced inflammation.8–11
mitter in the central nervous system via glycine-gated chlor-
Glycine is one of the major structural units of collagen,
ide channels (GlyR). The existence of GlyR has been
making up about 30% of the amino acids. It has been indi-
demonstrated on a wide variety of cells including different cated that serum glycine levels increase after intake of col-
cell types involved in immune responses, such as macro- lagen hydrolysate (CH).12 Clinical studies point to a
phages, monocytes, neutrophils and T lymphocytes.1–3 In possible beneficial effect of CH in osteoarthritis by a reduc-
the past years, evidence accumulated showing that glycine tion in hip or knee pain, after daily intake of 10 g CH for 60
causes an anti-inflammatory response, at least in part via days to six months.13–16
the activation of GlyR.3 Activation of GlyR in immune cells This study aims to answer the question whether CH is
results in an influx of chloride ions, membrane hyperpolar- able to modulate inflammatory responses. The anti-inflam-
isation and inhibition of voltage-operated calcium channels matory potential of CH was examined using a mouse model
and cell activation. Glycine largely prevents endotoxin- for zymosan-induced ear-skin inflammation.
induced TNF-a production by Kupffer cells and alveolar
macrophages, and it reduces IL-1b and TNF-a expression Materials and methods
while simultaneously stimulating IL-10 expression in lipo-
polysaccharide (LPS)-activated monocytes.1,4,5 Moreover, All experimental procedures using laboratory animals were
glycine interferes with TNF-a-mediated NF-kB activation approved by an independent animal experiments commit-
in human coronary arterial endothelial cells and differen- tee (DEC Consult, Bilthoven, the Netherlands).
tiated 3T3-L1 adipocytes.6,7
The in vitro data of the effects of glycine are reinforced by Prompting zymosan-induced ear-skin inflammation
in vivo results derived from different animal models includ- Male Balb/C mice (Charles River, Maastricht, the
ing: the inhibition of pro-inflammatory cytokine expression Netherlands), aged 14 weeks at the start of the experiment
in surgical manipulation-induced inflammation in the were randomly assigned to groups of six animals and

ISSN: 1535-3702 Experimental Biology and Medicine 2013; 238: 798–802

Copyright ß 2013 by the Society for Experimental Biology and Medicine
Downloaded from ebm.sagepub.com at UCSF LIBRARY & CKM on March 15, 2015
 Hartog et al. Collagen hydrolysate inhibits zymosan-induced inflammation 799
..........................................................................................................................
Whole blood assay
acclimatized to the animal housing for one-week prior to
the start of the experiment. All animals had free access to a Heparinized murine blood (50 mL) was incubated for 20 h
standard rodent diet and tap water. Different amounts of in the presence of LPS (E. coli, B55:055, final concentration
CH (0, 12.5, 25 or 50 mg/mouse) were administered daily 1 mg/mL, Sigma-Aldrich BV, Zwijndrecht, the Netherlands)
for three constitutive days (day 1 till 3) by oral gavage. Tap with and without strychnine (a glycine receptor antagonist,
water (vehicle) was applied in the same volume, 200 mL, to final concentration 10 M, Sigma-Aldrich BV) in a total
the control mice. Inflammation was induced at day 3, 2 h volume of 200 mL in culture medium (RPMI-1640 containing
after administration of CH, by injecting 25 mL zymosan 25 mM HEPES and 2 mM L-glutamine, Life-Technologies,
(0.5% suspended in PBS) or PBS (sham), intradermally in Merelbeke, Belgium, enriched with 100 U/mL penicillin/
both ears.17,18 Ear thickness was measured prior to and at streptomycin 100 mg/mL, Life-Technologies). The plates
3 and 6 h after zymosan injection using an engineer’s were subsequently centrifuged and the supernatants were
micrometer (Mitutoyo Dinamic, Veenendaal, the harvested and stored at 80 C until cytokine detection.
Netherlands). After the final ear thickness measurement,
mice were bled under terminal anaesthesia (isoflurane/ Cytokine detection
N2O/O2) and sacrificed. Cytokine levels were detected using a commercial
In a second group of mice, blood was taken at different Multiplex Bead immunoassay (BioRad, Veenendaal, the
time points (1 and 3 h) after CH (50 mg/mouse) intake. Netherlands) including IL-6 and TNF-a, according to the
manufacturer’s protocol. The analysis was performed by
using the Bio-Plex system (BioRad). Results were calculated
Plasma isolation using Bio-Plex Manager Software 3.1 (BioRad).
Blood was collected in Lithium Heparin MiniCollect
tubes (Greiner Bio-One B.V, Alphen a/d Rijn, the Statistical analysis
Netherlands) and preserved at room temperature awaiting
All data are expressed as mean  standard error of the mean
whole blood analysis. After starting the whole blood cell
(SEM). All readout conditions were compared to the control
assay, the remaining blood was centrifuged for 5 min at
using the analysis of variance (one-way ANOVA). The over-
6000 rpm in an EppendorfÕ centrifuge, and the plasma all significance of differences for all calculations was tested
was then collected and stored at 20 C for glycine using the post hoc Dunnett’s test. Differences between LPS
detection. stimulated IL-6 production and the LPS stimulated IL-6
production in the presence of strychnine were calculated
Plasma glycine detection by a T-test. Correlations were calculated using the
Pearson’s linear regression model.
After precipitation of proteins and polypeptides with per-
chloric acid, the plasma samples were centrifuged. Glycine Results
concentrations were determined in the supernatant by fully
automated High Performance Liquid Chromatography Plasma glycine levels
(Figure 1, HPLC, Shimadzu, ‘s-Hertogenbosch, the Plasma glycine levels were detected in control conditions
Netherlands). and at 1 (50 mg/mouse), 3 (50 mg/mouse) and 8 h (12.5, 25

Detector A Ex:340nm,Em:455nm
ala

4000
tau

Glycine
3000
Intensity (mV)

gln

gly

IS

2000
val
ser

1000
thre

leu

lys
ile
meth
arg

pheala
citrul
asn
glu

tyr

trp
his
asp

0
0 1 2 3 4 5 6 7 8 9 10 11 12 13
Retention Time (min)

Figure 1 A typical chromatogram of the plasma free amino acid analysis by HPLC. (A color version of this figure is available in the online journal)

Downloaded from ebm.sagepub.com at UCSF LIBRARY & CKM on March 15, 2015
800 Experimental Biology and Medicine Volume 238 July 2013
..........................................................................................................................
and 50 mg/mouse) after CH intake. CH increased the collected. TNF-a and IL-6 levels were detected by a
plasma glycine level (detected at 8 h after intake) in a con- Multiplex Beat Immunoassay. CH was able to inhibit IL-6
centration-dependent manner (Figure 2A). Intake of 50 mg production significantly at 25 mg/mL (Figure 4A).
of CH more than doubled the glycine plasma levels at 1 and Strychnine counteracted this effect (Figure 4B). Strychnine
3 h after intake (Figure 2B). significantly increased IL-6 levels in all collagen-treated
conditions but not in the control (Figure 4C1, C2, C3 and
Ear-skin inflammation C4). There were no differences detectable in the TNF-a level
To study the effect of CH on zymosan-induced ear-skin at the different tested conditions (data not shown).
inflammation in mice, CH was administered orally for Possible correlations between LPS induced IL-6 levels in
three days. On the third day, zymosan or PBS were injected whole blood, plasma glycine levels and ear-swelling at 6 h
into both ears intradermally. Ear thickness was measured after zymosan injection were calculated. The IL-6 produc-
before injection and at 3 and 6 h after injection (5 and 8 h tion and ear-swelling at 6 h after zymosan injection showed
after CH intake, respectively). Basal ear thickness was sub- a significant correlation (Pearson r ¼ 0.63, P ¼ 0.002,
tracted from the ear thickness at the different time-points Figure 5A). No correlation was found between the ear-
after zymosan injection to calculate ear swelling. In the swelling and IL-6 production in the strychnine supple-
vehicle-supplemented animals, the sham injection (PBS) mented whole blood cell assay (Pearson r ¼ 0.20,
induced an ear swelling of 79 mm and 112mm after 3 and P ¼ 0.377, Figure 5B).
6 h respectively. This PBS-induced swelling was subtracted
from the zymosan-induced ear swelling to acquire the Discussion
‘zymosan specific’ skin-swelling. The ear swelling of the Collagen is a natural component of the diet, found in animal
CH supplemented animals was compared to the control products such as meat and fish. Small peptides and amino
animals (vehicle supplemented). CH inhibited the zymosan acids derived from CH are detectable in blood.10,19 The pre-
induced ear-swelling in a concentration dependent manner
sent study demonstrated that orally administered CH
at 3 h after injection (Figure 3A). At 6 h post zymosan chal-
increases plasma glycine levels in mice in a concentration-
lenge the effect was no longer significant (Figure 3B).
dependent manner. Anti-inflammatory effectiveness was
indicated by counteraction of the zymosan-induced ear
Cytokine levels in an LPS-induced whole blood assay swelling at 3 h after zymosan injection (5 h after CH
Blood was incubated, ex vivo, in the presence of LPS with intake). Glycine serum levels, tested for the most effective
and without strychnine. After 20 h, the supernatant was anti-inflammatory dose (50 mg/day), were more than

(a) (b)
500 ** ** **
mM Gly 8 h after intake

* 800
400
mM Glycine

600
300 **
400
200

100 200

0 0
0 12.5 25 50 0 1 3 8
CH mg/day hr after CH intake

Figure 2 Plasma concentrations of Gly, 8 h after intake of vehicle (0) or different concentrations CH (a). Plasma concentration of Gly before (0) and 1, 3 and 8 h after,
intake of 50 mg CH (b). Data are means (mM)  SEM, n ¼ 6. Significant differences from no CH intake (0) are indicated *P < 0.05, **P < 0.01

(a) 175
(b)
150 200
ear swelling (mm)

ear swelling (mm)

125
** **
150
100
75 100

50
50
25
0 0
0 12.5 25 50 0 12.5 25 50
CH mg/day CH mg/day

Figure 3 Zymosan-induced ear inflammation (swelling) 3 (a) and 6 (b) hours after zymosan injection. Data are means (mm)  SEM, n ¼ 6. Significant differences from
no CH intake (0) are indicated **P < 0.01

Downloaded from ebm.sagepub.com at UCSF LIBRARY & CKM on March 15, 2015
 Hartog et al. Collagen hydrolysate inhibits zymosan-induced inflammation 801
..........................................................................................................................
(a) 0.25 (b) 0.3

0.20

IL-6 (ng/ml)

IL-6 (ng/ml)
0.2
0.15
*
0.10
0.1
0.05

0.00 0.0
0 12.5 25 50 0 12.5 25 50
CH mg/day CH mg/day
(c1) 0.40 (c2) 0.35
0.35 0.30 p< 0.05
0.30 0.25

IL-6 ng/ml
IL-6 ng/ml

0.25
0.20
0.20
0.15
0.15
0.10 0.10
0.05 0.05
0.00 0.00
–strychnine + strychnine –strychnine + strychnine
Control CH (12.5 mg)
(c3) 0.35 (c4) 0.35
0.30 0.30
0.25 p < 0.05 IL-6 ng/ml 0.25 p< 0.01
IL-6 (ng/ml)

0.20 0.20
0.15 0.15
0.10 0.10
0.05 0.05
0.00 0.00
–strychnine + strychnine –strychnine + strychnine
CH (25 mg) CH (50 mg)

Figure 4 LPS-induced (1 mg/mL) IL-6 production by whole blood cells, without (A) and in the presence of strychnine (10 M, B). Data are means (ng/mL)  SEM, n ¼ 6.
Significant differences from no CH intake (0) are indicated *P < 0.05. Figure C shows LPS-induced IL-6 production by whole blood cells with and without strychnine in
control condition (1) and after intake of 12.5 mg (2), 25 mg (3) and 50 mg (4) CH. Significant differences between the control and strychnine condition as detected by T-
test are indicated

(a) 300 (b) 300

ear swelling (mm)

ear swelling (mm)

200 200

100 100

0 0
0.0 0.1 0.2 0.3 0.4 0.5 0.0 0.1 0.2 0.3 0.4 0.5
IL-6 (ng/ml) IL-6 (ng/ml)

Figure 5 The correlation between ear-swelling 6 h after zymosan injection and LPS induced IL-6 production in whole blood (n ¼ 21) in the absence (a) and presence
(b) of strychnine

doubled at the time point of induction of the inflammation combination with synovial inflammation) CH is one of the
(2 h after CH intake). candidate disease modifying osteoarthritis drugs
Effectiveness of CH has been tested in different disease (DMOADs).22 Inflammatory cells play a role in osteoarth-
models. In the ethanol-induced ulcer model in rats, a single ritis and the aforementioned animal models in which the
dose of CH resulted in a strong reduction of ulcer forma- effects of CH are described. The results of this study suggest
tion19,20 and, in a spontaneous hyperlipidemic mouse that modulation of the inflammatory response by CH could
model, C57BL/6.KOR-Apoe(shl), CH feeding showed contribute to the protective effectiveness as detected in the
besides a lipid-lowering effect a decrease in plasma IL-6 different studies.
and TNF-a levels.21 Moreover, in osteoarthritis (which is Eight hours after intake of different amounts of CH,
associated with significant functional impairment in blood was collected and stimulated with LPS in the

Downloaded from ebm.sagepub.com at UCSF LIBRARY & CKM on March 15, 2015
802 Experimental Biology and Medicine Volume 238 July 2013
..........................................................................................................................
presence or absence of strychnine. The neurotoxin strych- 8. Hartog A, Leenders I, van der Kraan PM, Garssen J. Anti-inflammatory
nine acts as an antagonist on the GlyR, neutralizing the anti- effects of orally ingested lactoferrin and glycine in different zymosan-
induced inflammation models: evidence for synergistic activity. Int
inflammatory effects of glycine on immune cells.5 It was
Immunopharmacol 2007;7:1784–92
found that the LPS-stimulated IL-6 production was lower 9. Li X, Bradford BU, Wheeler MD, Stimpson SA, Pink HM, Brodie TA,
after intake of 25 mg CH, demonstrating a systemic effect. Schwab JH, Thurman RG. Dietary glycine prevents peptidoglycan
The addition of strychnine significantly increased the pro- polysaccharide-induced reactive arthritis in the rat: role for glycine-
duction of IL-6 in the CH groups but not in the control gated chloride channel. Infect Immun 2001;69:5883–91
group. Recent literature indicates that extracellular pH 10. Stoffels B, Turler A, Schmidt J, Nazir A, Tsukamoto T, Moore BA,
Schnurr C, Kalff JC, Bauer AJ. Anti-inflammatory role of glycine in
changes influence glycine receptor functionality.23 Both,
reducing rodent postoperative inflammatory ileus. Neurogastroenterol
the plasma and the cell cultures showed a physiological Motil 2011;23:76–87,e8
pH, excluding pH-induced receptor modifications. LPS 11. Alarcon-Aguilar FJ, Almanza-Perez J, Blancas G, Angeles S, Garcia-
induced IL-6 production and the detected ear swelling at Macedo R, Roman R, Cruz M. Glycine regulates the production of pro-
6 h after zymosan injection were highly correlated. inflammatory cytokines in lean and monosodium glutamate-obese
Although the plasma glycine levels were low at the time mice. Eur J Pharmacol 2008;599:152–8
of LPS stimulation, this correlation in combination with the 12. Walrand S, Chiotelli E, Noirt F, Mwewa S, Lassel T. Consumption of a
functional fermented milk containing collagen hydrolysate improves
strychnine results suggest that the GlyR are involved, point-
the concentration of collagen-specific amino acids in plasma. J Agric
ing to a contribution of glycine in the CH-induced anti- Food Chem 2008;56:7790–5
inflammatory effect. 13. Benito-Ruiz P, Camacho-Zambrano MM, Carrillo-Arcentales JN,
Collectively, these data show that CH is able to modulate Mestanza-Peralta MA, Vallejo-Flores CA, Vargas-Lopez SV, Villacis-
inflammatory responses. This effect might be constituted by Tamayo RA, Zurita-Gavilanes LA. A randomized controlled trial on the
inhibiting pro-inflammatory cytokine production via the efficacy and safety of a food ingredient, collagen hydrolysate, for
improving joint comfort. Int J Food Sci Nutr 2009;60:99–113
GlyR.
14. Bello AE, Oesser S. Collagen hydrolysate for the treatment of osteo-
arthritis and other joint disorders: a review of the literature. Curr Med
Author contributions: All authors participated in the inter- Res Opin 2006;22:2221–32
pretation of the study data and review of the manuscript; 15. Carpenter RL, Peel JB, Carpenter MR, Lowndes J, Angelopoulos T,
AH and JG designed the experiments; AH, MC and GdV Rippe JM. Effectiveness of a collagen hydrolysate-based supplement on
conducted the experiments; AH wrote the manuscript. joint pain, range of motion and muscle function in individuals with
mild osteoarthritis of the knee: a randomized clinical trial. Ann Rheum
Dis 2005;64:476–476
ACKNOWLEDGEMENTS 16. Bruyere O, Zegels B, Leonori L, Rabenda V, Janssen A, Bourges C,
Reginster JY. Effect of collagen hydrolysate in articular pain: a 6-month
This research was performed within the framework of TI randomized, double-blind, placebo controlled study. Complement Ther
Pharma project T1-103. Med 2012;20:124–30
17. Hougee S, Hartog A, Sanders A, Graus YM, Hoijer MA, Garssen J, van
den Berg WB, van Beuningen HM, Smit HF. Oral administration of the
REFERENCES
NADPH-oxidase inhibitor apocynin partially restores diminished car-
tilage proteoglycan synthesis and reduces inflammation in mice. Eur J
1. Froh M, Thurman RG, Wheeler MD. Molecular evidence for a glycine-
Pharmacol 2006;531:264–9
gated chloride channel in macrophages and leukocytes. Am J Physiol
18. Erdo F, Torok K, Aranyi P, Szekely JI. A new assay for antiphlogistic
Gastrointest Liver Physiol 2002;283:G856–63
activity: zymosan-induced mouse ear inflammation. Agents Actions
2. Wheeler M, Stachlewitz RF, Yamashina S, Ikejima K, Morrow AL,
1993;39:137–42
Thurman RG. Glycine-gated chloride channels in neutrophils attenuate
19. Castro GA, Sgarbieri VC, Carvalho JE, Tinti SV, Possenti A. Protective
calcium influx and superoxide production. FASEB J 2000;14:476–84
effect of collagen derivates on the ulcerative lesions caused by oral
3. Zhong Z, Wheeler MD, Li X, Froh M, Schemmer P, Yin M, Bunzendaul H,
administration of ethanol. J Med Food 2007;10:154–8
Bradford B, Lemasters JJ. L-glycine: a novel antiinflammatory, immu-
20. Castro GA, Carvalho JE, Tinti SV, Possenti A, Sgarbieri VC. Anti-
nomodulatory, and cytoprotective agent. Curr Opin Clin Nutr Metab Care
ulcerogenic effect of a whey protein isolate and collagen hydrolysates
2003;6:229–40
against ethanol ulcerative lesions on oral administration to rats. J Med
4. Ikejima K, Iimuro Y, Forman DT, Thurman RG. A diet containing glycine
Food 2010;13:83–90
improves survival in endotoxin shock in the rat. Am J Physiol
21. Zhang Y, Kouguchi T, Shimizu K, Sato M, Takahata Y, Morimatsu F.
1996;271:G97–103
Chicken collagen hydrolysate reduces proinflammatory cytokine pro-
5. Spittler A, Reissner CM, Oehler R, Gornikiewicz A, Gruenberger T,
duction in C57BL/6.KOR-ApoEshl mice. J Nutr Sci Vitaminol (Tokyo)
Manhart N, Brodowicz T, Mittlboeck M, Boltz-Nitulescu G, Roth E.
2010;56:208–10
Immunomodulatory effects of glycine on LPS-treated monocytes:
22. Van Vijven JP, Luijsterburg PA, Verhagen AP, van Osch GJ,
reduced TNF-alpha production and accelerated IL-10 expression. FASEB
Kloppenburg M, Bierma-Zeinstra SM. Symptomatic and chondropro-
J 1999;13:563–71
tective treatment with collagen derivatives in osteoarthritis: a system-
6. Blancas-Flores G, Alarcon-Aguilar FJ, Garcia-Macedo R, lmanza-
atic review. Osteoarthr Cartilage 2012;20:809–21
Perez JC, Flores-Saenz JL, Roman-Ramos R, Ventura-Gallegos JL,
23. Song YP, Schlesinger F, Ragancokova D, Calixto R, Dengler R,
Kumate J, Zentella-Dehesa A, Cruz M. Glycine suppresses TNF-alpha-
Krampfl K. Changes in extracellular pH affect glycine receptor channels
induced activation of NF-kappaB in differentiated 3T3-L1 adipocytes.
expressed in HEK 293 cells. Eur J Pharmacol 2010;636:59–64
Eur J Pharmacol 2012;689:270–7
7. Hasegawa S, Ichiyama T, Sonaka I, Ohsaki A, Okada S, Wakiguchi H,
Kudo K, Kittaka S, Hara M, Furukawa S. Cysteine, histidine and glycine
exhibit anti-inflammatory effects in human coronary arterial endothelial (Received October 25, 2012, Accepted January 23, 2013)
cells. Clin Exp Immunol 2012;167:269–74

Downloaded from ebm.sagepub.com at UCSF LIBRARY & CKM on March 15, 2015