Plant Foods for Human Nutrition
2006 Springer Science+Business Media, Inc.
Effect of Different Flavonoids on Collagen Synthesis in Human Fibroblasts
& DIRK VANDEN BERGHE
Department of Molecular Medicine, Rudjer Boskovic Institute, Zagreb, Croatia;
Department of Molecular Genetics, Rudjer Boskovic Institute,
Department of Pharmaceutical Sciences, University of Antwerp, Antwerp, Belgium (
author for correspondence; e-mail:
Abstract In this study, we discovered that ﬂavonoids belonging to the sub-
classes: (ﬂavanone, ﬂavone, and ﬂavonol) display differential effects on the
synthesis of collagen in human dermal ﬁbroblasts. At 80 µg/ml ﬂavonoids
, 5,7-pentahydroxyﬂavone, 3-methyl quercetin, and 7-
hydroxyﬂavone signiﬁcantly decreased the total protein concentration
which was a direct consequence of their cytotoxic effect, while narin-
genin exhibited no effect on total collagen and total protein concentration.
,7-tetramethyl ether, 4
-hydroxyﬂavanone, ﬂavanone, and
ﬁsetin signiﬁcantly decreased collagen concentration while morin, rutin,
and chrysin increased collagen concentration without changing the over-
all protein concentration. The initial screening performed in this study
enables the identiﬁcation of compounds that exert signiﬁcant effects on
ﬁbroblast function and show potential as starting material for pharma-
ceutical preparations targeted against various disorders centered around
disturbed collagen metabolism.
Keywords: Collagen, Fibroblasts, Flavonoids, Tissue culture, Woessner
Flavonoids are a large group of naturally occurring, low
molecular weight polyphenolic compounds present in all
plants and are, therefore, an integral part of human diet.
Fruits, vegetables, and beverages such as tea and red wine
are especially rich sources of ﬂavonoids. The dietary in-
take of ﬂavonoids has been estimated to vary from 100 to
1000 mg/day . Perhaps the most active area of ﬂavonoid
research at the present time is the possible medicinal con-
tribution that ﬂavonoids make to human health. Flavonoids
have been shown to possess, both in vitro and in vivo, a wide
range of biochemical and pharmacological effects, includ-
ing antiinﬂammatory and antiallergic effects . While ma-
jor focus of the studies on ﬂavonoids has been their antiox-
idant properties, there is an emerging view that ﬂavonoids
act as modulators of cell signaling . Understanding the
cellular effects of ﬂavonoids is important for predicting
which dietary ﬂavonoids might be most beneﬁcial in vivo.
Fibroblasts are responsible for the synthesis of collagen,
a ﬁbrous protein of extracellular matrix (ECM) and major
constituent of connective tissue (such as skin, tendons, liga-
ments, cartilage, and bones) responsible for scar formation.
Human ﬁbroblasts in culture synthesize both types I and III
collagen, with type I accounting for 70–90% of the total .
Studies on the effects of ﬂavonoids on collagen synthesis
have in the past few years progressed to the point of sug-
gesting the potential therapeutic use of these plant derived
compounds in the treatment of various medical conditions
centered on disrupted collagen metabolism, for example,
delayed wound healing, excessive scarring, skin aging, etc.
[5–8]. It is worth mentioning that ﬂavonoids from plant
extracts are available as over-the counter ‘natural’ ingredi-
ents in a form of oral supplements or skin care cosmetic
products. They are recommended for daily consumption
and can be bought in regular drug stores or easily ordered
via internet. The list of manufacturer’s claimed beneﬁts for
these ﬂavonoids-based products is long and sometimes not
scientiﬁcally founded. Therefore, any additional scientiﬁc
interest in ﬂavonoids should be welcomed because it will
ultimately drive future research that will distinguish truly
efﬁcacious ingredients from misleading claims. We are re-
porting results of preliminary screening of 11 ﬂavonoids
commonly present in diet, in terms of their inﬂuence on
collagen synthesis. Figure 1 shows names of ﬂavonoids
and corresponding general structures.
Materials and Methods
Fibroblast culture was established from a sample of neona-
tal human foreskin under Institutional Review Board Ex-
empted protocol as described . Cells were grown in
Dulbecco’s modiﬁed Eagle’s medium (DMEM) (Invit-
rogen Grand Island, NY) supplemented with 10% fetal
bovine serum FBS (Gemini-Bio-Products, Woodland, CA).
Flavonoids were purchased from Sigma–Aldrich corpora-
tion with the exception of 3-methyl quercetin, a natural
ﬂavonoid, isolated from plant material by Prof. D. Vanden
Berghe. Cells at 50% conﬂuence (2.5 ×10
dish) were treated with 80 µmol/l (0.05% DMSO/DMEM),
40 µmol/l (0.025% DMSO/DMEM), and 20 µmol/l
(0.0125% DMSO/DMEM) ﬂavonoid concentrations. The
concentration range was selected on the basis of a pre-
vious “trial and error” experiment where it has been
shown that higher concentrations (above 100 µmol/l) of
tested ﬂavonoids were in most cases toxic to cells. 0.05%
DMSO/DMEM, 0.025% DMSO/DMEM, and 0.0125%
DMSO/DMEM solutions were used as controls in order
to rule out the possible effect of DMSO on ﬁbroblasts.
All concentrations were tested in triplicates. Ascorbic acid
Figure 1. Names and general structures of ﬂavonoids.
solution was added every day in concentration of 50 µg/ml
to the medium. After 5 days, cells were collected using
rubber policeman in 1 ml of 10 mM PBS. Cell membranes
were disrupted using sonicator and total protein amount was
quantiﬁed using Bio-Rad protein assay kit (Bio-Rad Labo-
ratories, Hercules, CA). The hydroxyproline was quantiﬁed
according to Woessner, method suitable for cases when the
proportion of hydroxyproline to other amino acids in the
sample is very small . 0.5 ml of sonicated samples
pules (Wheaton Science Products, Millville, NJ). The vials
were ﬂame-sealed and placed in 115
C oven for 18 hrs.
The samples were dried under N
gas in 50
C water bath.
The residue was resuspended in 250 µl deionized H
50 µl was mixed with 450 µl of deionized H
O to per-
form assay. The standard was prepared in glass test tubes
using 10 µg/ml hydroxyproline standard (Sigma–Aldrich,
St. Louis, MO). Following, were the incubations of sam-
ples: 20 min incubation with 250 µl of 0.05 M chloramine
T solution [chloramine T solution = 1.41 g sodium p-
toluenesulfonylchloramide (chloramine T, Sigma–Aldrich,
St. Louis, MO)] dissolved in 20 ml dH
O, 30 ml
methyl cellosolve (2-methoxyethanol, Fisher Scientiﬁc,
Pittsburgh, PA) and 50 ml citric acid buffer]; 5 min
incubation with 250 µl of 3.15 M perchloric acid
(Fisher Scientiﬁc, Pittsburgh, PA); 20 min incubation with
250 µl of 20% p-dimethylbenzaldehyde (Fisher Scientiﬁc,
Pittsburgh, PA) in 60
C water bath dissolved in methyl
cellosolve to a ﬁnal volume of 50 ml and heated to 60
dissolve the reagent. After cooling to room temperature, the
optical density of samples was read at 557 nm in a Milton-
Roy spectrophotometer Model 601 (Milton-Roy, Ivyland,
The data was analyzed using one-way analysis of vari-
ance (ANOVA) accepting
p < 0.05 as level of signiﬁcance.
Calculation of the total collagen was based on the analy-
sis of total hydroxyproline which is for the most purposes,
a reasonably speciﬁc marker for collagen, as the amount
present in elastin and in the Clq component of complement
(the only two potential sources) is negligible by comparison
. The conversion factor of hydroxyproline to collagen
was taken as 7.8.
Results and Discussion
The results obtained clearly show that tested ﬂavonoids
exert differential effects on the total protein concentra-
tion and total collagen concentration in ﬁbroblast cultures
(Table 1, Figure 2). Quercetin (3,3
droxyﬂavone) showed cytotoxic effect and signiﬁcantly de-
creased total protein and total collagen concentration in a
dose-dependent manner at all tested concentrations (more
Table 1. Effects of three different concentrations
of ﬂavonoids on protein synthesis
in neonatal human
ﬁbroblasts expressed in µg/dish
Compound 80 µmol/l 40 µmol/l 20 µmol/l
-Hydroxyﬂavanone 915.87 ±58.30 942.47 ±45.89 936.50 ±38.45
963.07 ±81.50 948.93 ±31.59 946.80 ±25.28
Flavanone 914.20 ±9.25 1060.07 ±90.95 950.60 ±32.34
±25.04 894.13 ±57.27 963.53 ±43.05
954.40 ±39.92 987.57 ±31.96 913.13 ±71.93
1055.27 ±71.03 952.80 ±37.87 990.30 ±28.73
938.10 ±35.26 949.53 ±21.54 949.50 ±60.04
915.30 ±57.06 943.00 ±48.93 907.17 ±16.06
919.13 ±10.61 938.10 ±62.48 943.00 ±33.43
Control 957.10 ±39.38 950.57 ±25.03 947.33 ±23.72
Each ﬂavonoid was tested in three concentrations: 80, 40, and 20 µmol/l.
Values were calculated using means of three measurements.
Signiﬁcant difference from control values was determined by ANOVA (single factor analysis) where
p < 0.05;
p < 0.01.
than 50% inhibition) which can be attributed to its ability of
inducing DNA damage . Overall, 80 µmol/l was deter-
mined to be concentration at which ﬂavonoids exert maxi-
mum effect on the concentration of collagen in ﬁbroblasts.
At 80 µmol/l, ﬁsetin, quercetin-tetramethylether, ﬂavanone,
-hydroxyﬂavanone decreased collagen concentra-
tion with no effect on the overall protein concentration
(Table 1, Figure 2). Such cases, where the overall protein
amount remains unaffected while the collagen has either
been decreased or increased, are of particular interest be-
cause at these concentrations ﬂavonoids speciﬁcally act on
collagen production and at the same time are not harmful
to the cells.
One of the possible explanations for collagen inhibit-
ing activity of various ﬂavonoids could lie in their ability
to interfere with the biosynthesis of the precursor colla-
gen molecule known as procollagen. In the study involving
genistein and quercetin, it has been shown that ﬂavonoids
have the ability to directly inﬂuence the synthesis of colla-
gen by inhibiting type I procollagen mRNA in concentra-
tions ranging from 6.25 to 70 µmol/l, indicating effect at a
pretranslational level . Similarly, quercetin was shown
to have growth-inhibitory effect on keloid derived ﬁbrob-
last and signiﬁcantly inhibited collagens I and III expres-
sion [7,14]. The inhibition of collagen by some ﬂavonoids
is a property relating positively to cases of deranged heal-
ing resulting in excessive scarring. Compounds capable of
reducing collagen could modulate the deposition of extra-
cellular matrix thus favoring the repair process with very
On the other hand, at 80 µmol/l concentration morin,
rutin, and chrysin increased collagen concentration 42, 61,
and 42% respectively as compared to control (p = 0.01
and p < 0.01) without affecting the total protein concentra-
tion (Table 1, Figure 2). Collagenolytic enzymes, precisely
matrix metalloproteinases (MMPs) are neutral proteinases
that are inhibited by decrease in free zinc concentrations
caused by chelating agents. With regards to this, ﬂavonoid
catechin, isolated from green tea, is currently in clinical
trials for the treatment of MMP-medicated diseases .
Because ﬂavonoids have already been proven to be very
effective in chelating metal ions [16–18] it is possible that
they potentially act as collagenase inhibitors which would
result in the overall increase in the measurable collagen con-
centration. Novel results suggest that certain ﬂavonols have
strong inhibitory activity against collagenase . Simi-
larly, catechins, in particular ( −)epigallocatechin gallate
(EGCG) has been shown to inhibit membrane type I matrix
Figure 2. Effect of 80 µmol/l (A), 40 µmol/l (B), and 20 µmol/l (C) ﬂavonoid concentrations on collagen
production, expressed as µg of total collagen per dish. t-bars represent standard deviation. Signiﬁcant
difference from control values determined by ANOVA (single factor analysis) where
p < 0.05;
p < 0.01.
metalloproteinase which hydrolyzes type I collagen .
Additionally, certain ﬂavonoids have the ability to bind di-
rectly to collagen ﬁbers, thus making them resistant to the
action of mammalian collagenase . Flavonoids of sub-
class proanthocyanidins display high afﬁnity for collagen
and elastin ﬁber in addition to preventing their enzymatic
hydrolysis by matrix metalloproteinases [22, 23]. This im-
plies that the treatment of ﬁbroblasts with these ﬂavonoids
would result in increase of measurable extracellular
Cutaneous aging resulting in wrinkled, rough, and pig-
mented skin is considered to be as a consequence of the
interplay of extrinsic damage to skin by UV radiation, in-
trinsic increases in collagen-degrading matrix metallopro-
teinases, and decreased collagen synthesis . As men-
tioned previously, some ﬂavonoids are capable of increasing
collagen and inhibiting the degradation of matrix (reduction
in dermal thickness). Flavonoids can also have photopro-
tective properties against UVB radiation .
Collagen stimulating activity observed for some ﬂavo-
noids might be beneﬁcial in the cases of delayed healing of
cutaneous wounds. By inhibiting MMP, ﬂavonoids could
increase the rate and the amount of collagen synthesized
by ﬁbroblasts necessary for the formation of new wound
matrix, thereby speeding up the healing process.
Overall, it is reasonable to consider the use of ﬂavonoids,
having the ability to accelerate wound healing, con-
trol scar formation and prevent or slow-down the pro-
cess of skin aging for dermatological and cosmetic
Knowing the important function of collagen in main-
taining structure, storing energy, altering its diameter and
modality of packing and its role in physiological processes
such as aging and wound healing, the ability of ﬂavonoids
to inﬂuence the collagen synthesis should be further ex-
plored. The initial screening performed in this study, by
testing the activity of ﬂavonoids on collagen synthesis, en-
ables the identiﬁcation of compounds that exert signiﬁcant
effects on ﬁbroblast function and show potential as start-
ing material for various antiwrinkling cosmetic prepara-
tions, wound healing or even antiﬁbrotic drug development
This study was performed within the Masters degree pro-
gram at the Faculty of Medicine and Pharmaceutical Sci-
ences, University of Antwerp, Belgium. The authors would
like to thank Prof. Karen Reiser, Department of Internal
Medicine, University of California at Davis, for helping in
analysing collagen data.
1. Aherne SA, O’Brien NM (2002) Dietary ﬂavonols: Chemistry, food
content, and metabolism. Nutrition 18: 75–81.
2. Middleton E, Kandaswami C (1992) Effectsof ﬂavonoids on immune
and inﬂammatory cell functions. Biochem Pharmacol 43: 1167–
3. Williams RJ, Spencer JP, Rice-Evans C (2004) Flavonoids: Antiox-
idants or signalling molecules? Free Radic Biol Med 36(7): 838–
4. Nimmi ME (1988) Collagen. In: Biotechnology. Boca Raton: CRC
Press, pp 3–250.
5. Kapoor M, Howard R, Hall I, et al. (2004) Effects of epicatechin
gallate on wound healing and scar formation in a full thickness
incisional wound healing model in rats. Am J Pathol 165(1): 299–
6. Adcocks C, Collin P, Buttle DJ (2002) Catechins from green tea
(Camellia sinensis) inhibit bovine and human cartilage proteogly-
can and type II collagen degradation in vitro. J Nutr 132(3): 341–
7. Phan TT, Sun L, Bay BH, et al. (2003a) Dietary compounds inhibit
proliferation and contraction of keloid and hypertrophic scar-derived
ﬁbroblasts in vitro: Therapeutic implication for excessive scarring. J
Trauma 54(6): 1212–1224.
8. Moon HI, Kwak JH, Zee OP, et al. (2005) Regulation effect of 2
trihydroxyisoﬂavone on the expression of matrix metalloproteinase
-1,2 in ultraviolet-B irradiated primary cultured old aged human skin
ﬁbroblasts. Biol Pharm Bull 28(11): 2173–2175.
9. Stipcevic T, Piljac T, Isseroff RR (2005) Di-rhamnolipid from Pseu-
domonas aeruginosa displays differential effects on human ker-
atinocyte and ﬁbroblast cultures. J Dermatol Sci 40(2): 141–143.
10. Woessner JF (1961) The determination of hydroxyproline in tissue
and protein samples containing small proportions of this imino acid.
Arch Biochem Biophys 93: 440.
11. Reisser KM (1996) Extracellular matrix. In: Marshall VW, Cole TR
(eds.) Encyclopedia of Gerontology, vol. 1. San Diego: Academic
Press, pp 519–529.
12. Formica J, Regelson W (1995) Review of the biology of quercetin
and related bioﬂavonoids. Food Chem Toxicol 33(12): 1061–
13. Qi LH, Kang LP, Zang JP, et al. (2001) Antiﬁbrotic effects of
genistein and quercetin in vitro. Yao Xue Xue Bao 36(9): 648–
14. Phan TT, See P, Tran E, et al. (2003b) Suppression of insulin-like
growth factor signalling pathway and collagen expression in keloid-
derived ﬁbroblasts by quercetin: Its therapeutic potential use in the
treatment and for prevention of keloids. Br J Dermatol 148(3): 544–
15. Mandal M, Mandal A, Das S, Chakraborti T, Sajal C (2003) Clinical
implications of matrix metalloproteinases. Mol Cell Biochem 252(1–
16. McDonald M, Mila I, Scalbert A (1996) Precipitation of metal ions
by plant polyphenols: Optimal conditions and origin of precipitation.
J Agric Food Chem 44(2): 599–606.
17. de Souza RF, De Giovani WF (2005) Synthesis, spectral and
electrochemical properties of Al(III) and Zn(II) complexes with
ﬂavonoids. Spectrochim Acta A Mol Biomol Spectrosc 61(9): 1985–
18. Scatterﬁeld M, Brodbelt JS (2000) Enhanced detection of ﬂavonoids
by metal complexation and electrospray ionization mass spectrome-
try. Anal Chem 72(24): 5898–5906.
19. Sin BY, Kim HO (2005) Inhibition of collagenase by nat-
urally occuring ﬂavonoids. Arch Pharm Res 28(10): 1152–
20. Yamakawa S, Asai T, Uchida T, et al. (2004) ( −)Epigallocatechin
gallate inhibits membrane-type 1 matrix metalloproteinase, MT-1-
MMP, and tumor angiogenesis. Cancer Lett 210(1): 47–55.
21. Kuttan R, Donnelly PD, Di Ferrante N (1981) Collagen treated with
( +)catechin becomes resistant to the action of mammalian collage-
nase. Experientia 37: 221–223.
22. Hyun CG, Hwang YS, Choi YR, et al. (2003) Development of
collagenase-resistant collagen and its interaction with adult human
dermal ﬁbroblasts. Biomaterials 24: 5099–5113.
23. An BJ, Kwak JH, Park JM, et al. (2005) Inhibition of enzyme ac-
tivities and the antiwrinkle effect of polyphenol isolated from the
persimmon leaf (Diospyros kaki folium) on human skin. Dermatol
Surg 37: 848–854.
24. Fisher GJ, Wang ZQ, Datta SC, et al. (1997) Pathophysiology of
premature skin aging induced by ultraviolet light. N Engl J Med
25. Hsu S (2005) Green tea and the skin. J Am Acad Dermatol 5(6):