Mol.

Cells OS, 625-630, December 31, 2008

Molecules
Communication and
Cells
©2008 KSMCB

A Rapid and Sensitive Screening System for

Human Type I Collagen with the Aim of Discovering
Potent Anti-Aging or Anti-Fibrotic Compounds
Md. Abul Hashem, Kyu-Yeon Jun, Eunyoung Lee, Soyun Lim, Hea-Young Park Choo, and
Youngjoo Kwon*

This study was undertaken with the aim of developing an alpha 1) and `liN^O (collagen, type I, alpha 2), respectively
easy and quick means of analyzing the effect of various (Di Lullo et al., 2002).
compounds on the synthesis and secretion of human type Fibrosis, characterized by an abundant accumulation of ma-
I collagen at the protein level. A modification of the ELISA trix proteins in the extracellular space, is attributed to the persis-
method was used on HFF-1 cells. For the proof of concept, tent overproduction of collagen and other connective tissue
we used thirteen compounds most of which are known to macromolecules. It is closely associated with heart failure and a
be antioxidants. Each compound was tested at concentra- number of other diseases including scleroderma, (Denton et alK,
tions of 0, 10 and 100 µM on HFF-1 cells for 24 h. Thirteen 2003), lung carcinoma (Kobayashi et al., 1999), myocardial
sets of experiments for each compound were performed in fibrosis (Querejeta et alKI 2004), biliary fibrosis (Boigk et alKI
ANOVA with three replicates. Duncan multiple range test 1997), cirrhosis (Friedman, 1993), myocardial necrosis (Sakata
(DMRT) was used to compare the mean values obtained et alK, 2008), and cardiac hypertrophy (Yang et alK, 1997). In
from the treatment groups. From the results it was con- particular, type I collagen levels in the heart is a major determi-
cluded that Vitamin C, undecylenic acid, conjugated li- nant of the proliferation of cardiac fibroblasts and is directly as-
noleic acid, glycolic acid, and citric acid at 100 µM concen- sociated with heart failure. This is because within the cardio-
tration could be used for anti-wrinkling or protection from vascular system, type I collagen usually exists in the form of
premature aging, which requires enhancement of collagen thick fibers with high tensile strength inducing myocardial stiff-
synthesis. Lactic acid, EGCG, resveratrol, and retinol that ness (Querejeta et alKI 2004; Zannad et al., 2001).
can inhibit collagen synthesis effectively in a dose- Skin aging generally results in a reduction in the amount of
dependent manner may be used for anti-fibrosis treatment connective tissue and concomitant disorganization of its structure.
purposes. The change in the structure of human skin connective tissue by
chronological aging and solar ultraviolet (UV) irradiation-induced
premature aging (photoaging) may be attributed to reduced syn-
INTRODUCTION thesis and elevated degradation of type I collagen. Many re-
searchers have tried to develop effective anti-wrinkle agents for
There are two aspects pertaining to collagen synthesis in the cosmetic purposes. Such agents include alpha-ketoglutarate
human body. One is to enhance collagen synthesis for anti- (Son et alK, 2007), retinol (Kafi et al., 2007), 1,4-dihydroxy-2-
wrinkling effects and to prevent premature aging or photoaging. methoxy-7-methylanthraquinone derived from jçêáåÇ~= ÅáíêáÑçäá~
The second aspect is to inhibit or prevent the synthesis of col- (Noni) fruit extracts (Kim et alK, 2005),= m~å~ñ= ÖáåëÉåÖ= C.A.
lagen for anti-fibrosis effects. Collagen is the most abundant Meyer root extracts (Lee et alKI 2007), and asiaticoside (Lee et
protein in the body, making up about 25% of the whole-body alKI 2006), which act through the increased synthesis of type I
protein content. Bundles of collagen molecules called collagen collagen. However, there is no consensus in the structure of
fibers have a distinctive structure of triple helices that provide chemicals reported to reduce or promote the production of type
great strength and elasticity. Collagen fibers are a major com- I collagen. This may be due to the complexity and high cost of
ponent of the extracellular matrix, supporting most of the con- measuring the effect of these compounds on collagen biosyn-
nective tissue in animals. Strong mature type I collagen fibers thesis.
are formed from triple-stranded, rope-like type I procollagen We thus tried to develop a screening technology to identify
molecules by enzymatic process outside the cell. The type I whether certain chemicals could increase or decrease collagen
procollagen molecules consist of two pro-alpha1 chains and synthesis. The present study was thus undertaken to measure
one pro-alpha2 chain, encoded by `liN^N (collagen, type I, any changes in collagen biosynthesis effected by different

College of Pharmacy and Division of Life and Pharmaceutical Sciences, Ewha Womans University, Seoul 120-750, Korea
*Correspondence: ykwon@ewha.ac.kr

Received July 30, 2008; accepted September 4, 2008; published online September 23, 2008

Keywords: anti-fibrosis, anti-wrinkling, collagen biosynthesis, HFF-1 cell, human type I collagen screening system
626 Human Type I Collagen Screening System

pounds in human fibroblast cells using ELISA as a high- in the presence or absence of the test compounds for 24 h. The
throughput screening method. We further justified this proce- culture supernatants were then harvested and measured with a
dure through the estimation of type I procollagen and mRNA sandwich immunoassay kit, which was utilized in accordance
expression of some collagen-related genes. with the instructions of the manufacturer. The measurement
was performed with a microplate reader at 450 nm (Lee=et alKI
MATERIALS AND METHODS 2006).

Cell lines and culture RT-PCR amplification and quantification of `liN^N,

The cell line HFF-1 (SCRC-1041™) purchased from ATCC `liN^O and jjmN mRNA
(http://www.atcc.org) originated from normal human skin fibro- Three genes namely `liN^N [Homo sapiens], `liN^O
blast cells. The cells were cultured in 100 mm dishes until 70 to [Homo sapiens] and ã~íêáñ=ãÉí~ääçéêçíÉáå~ëÉ=N (jjmN) (inter-
80% confluence was reached, after which the cells were stitial collagenase) [Homo sapiens] were studied after treatment
seeded in 96-well tissue culture plates at densities of 2 × 104 of the cells with Vitamin C, EGCG, UDA and resveratrol at con-
per well for collagen synthesis assay. The cells were main- centrations of 0, 10 and 100 μM for 24 h. Total RNA was iso-
tained in Dulbecco’s modified Eagle’s medium (DMEM) sup- lated using the RNeasy Mini kit (Qiagen Inc.) according to the
plemented with 10% heat-inactivated fetal bovine serum (FBS) manufacturer’s instructions. The quality of total RNA extracted
(Hyclone). The atmosphere was humidified and maintained at was monitored by determining the ratio of absorbance at 260
37°C in 5% carbon dioxide and 95% air. The cultures were nm to 280 nm using a Nanodrop spectrophotometer (ND-1000
treated for 24 h as per experimental design with different com- v3.3.0) and 1.5% agarose gel electrophoresis. The concentra-
pounds. tion of total RNA was also determined for normalization prior to
RT-PCR (reverse transcriptase-polymerase chain reaction).
Experimental design RT-PCR analysis was used to measure the level of mRNA
Thirteen compounds used in the present study were purchased gene expression. For RT-PCR analysis, 25 μl of AccessQuick™
from Sigma-Aldrich: ascorbic acid, ferulic acid, caffeic acid, Master Mix (Promega) was added to 1 μl of RNA templates in
undecylenic acid (UDA), conjugated linoleic acid (CLA), epigal- reaction vials, followed by 1 μl of AMV RT, 2 μl of primers and
locatechin gallate (EGCG), resveratrol, curcumin, silibinin, gly- 21 μl of nuclease-free water to make a final volume of 50 μl.
colic acid, citric acid, lactic acid and retinol. The effects of these Total RNA from cultured HFF-1 cells was reverse transcribed at
test compounds at concentrations of 0, 10 and 100 μM on HFF- 45°C for 45 min. The cDNA samples were incubated in a PCR
1 cells were studied for 24 h. Each treatment was conducted machine for an initial predenaturation at 94°C for 5 min, fol-
three times in triplicate. lowed by 28 PCR cycles. Each cycle was run at 95°C for 30 s,
55°C for 30 s, and 68°C for 1 min. Final extension was con-
ELISA for type I collagen ducted at 68°C for 5 min. The following oligonucleotide primers
Enzyme-linked immunosorbent assay (ELISA) was developed were used: for `liN^N, AGC CAG CAG ATC GAG AAC AT
using a rabbit-anti-human type I collagen polyclonal antibody (sense) and TCT TGT CCT TGG GGT TCT TG (antisense); for
(Chemicon) and an ECL anti-rabbit IgG horseradish peroxi- `liN^O, TCA AGG TTT CCA AGG ACC TG (sense) and
dase-linked species-specific whole antibody (from donkey) (GE GTG TCC CCT AAT GCC TTT GA (antisense); for jjmN,
Healthcare). The outline of the procedure is briefly described GAT GTG GAG TGC CTG ATG TG (sense) and TGC TTG
below. After a 24 h-treatment of the cells with the different test ACC CTC AGA GAC CT (antisense); for d^mae (glyceralde-
compounds in 96-well plates, 80 μl of supernatant was trans- hyde 3-phosphate dehydrogenase), AAC GAA TTT GGT CGA
ferred to new plates. The plates were incubated for 1 h at room ACA GC (sense) and TGA GGA GGG ATT CAG TG (an-
temperature (RT) to attach the synthesized collagen to the tisense). The amplified PCR product was electrophoretically
surface of the plate. Thereafter, the supernatants were re- resolved on a 1.5% agarose gel in 1× TAE (Tris-acetate-EDTA)
moved and a blocking solution (0.2% skim milk in PBS) was buffer solution (40 mM Tris-acetate, 1 mM EDTA, pH 8.2-8.4)
added to each well and shaken at 50 to 100 rpm at RT for 1 h. and was visualized with ethidium bromide staining and UV light.
The primary antibody (1/1000 in blocking solution) was then The intensity of each band was scanned and quantified using
added to each well and shaken for 1 h as before. Afterwards, the AlphaImager analysis program v5.5. Relative `liN^N,
the supernatants were removed and each well was washed five `liN^O and jjmN mRNA expression levels were repre-
times with PBS to remove non-specific binding. Secondary sented as a ratio of each gene to d^mae. Each treatment was
antibody (1/1000 in blocking solution) was added to each well repeated thrice.
and shaken as before. Supernatants were then removed and
the wells were washed five times with 0.05% Tween 20 in PBS Statistical analyses
and two times with 0.2% skim milk in PBS. Subsequently, 100 In all experiments, data are expressed as means ± SD, with at
μL of 1% 1,2-phenylenediamine (OPD flakes, Aldrich) substrate least 3 repeats in each experimental group. All data were sub-
solution was added into each well and incubated for 30 min in jected to one-way analysis of variance (ANOVA) and Duncan’s
the dark. The absorbance was then recorded at 450 nm using multiple range (DMRT) test using general linear models in a
the VersaMaxPLUS ROM v11.21 (Molecular Devices) to de- statistical analysis system (SAS) program in order to determine
termine the relative collagen output. differences among the experimental groups. P values less than
0.05 were considered to be statistically significant.
Quantitative detection of type I procollagen carboxy-
terminal peptide RESULTS
The quantity of type I collagen in the cells was determined us-
ing a commercially available kit (Takara Bio Inc., Japan). This Development of a typical ELISA standard curve for type I
kit detected the type I procollagen carboxy-terminal peptide collagen
using monoclonal antibodies, instead of directly measuring the A sensitive and reproducible ELISA for type I collagen was
collagen levels. Human dermal fibroblast cells were incubated developed using two antibodies. The results of this ELISA
 Md. Abul Hashem et al. 627

Fig. 2. Alterations of collagen biosynthesis by the test compounds.

Fig. 1. A typical ELISA standard curve for type I collagen. All the compounds were treated at concentrations of 10 μM and
100 μM, respectively. Bars represent the mean ± S.D. from three
experiments performed in triplicate. The asterisk means signifi-
showed a linear standard curve where the coefficient of deter- cance of p < 0.05 compared to the control (no compound treated).
mination (R2) value was 0.93 and regression coefficient Ä=value
was 1.93 (Fig. 1). This value was sensitive enough to deter-
mine the amount of type I collagen produced by cultured HFF-1
cells.

Compounds increasing type I collagen synthesis

Type I collagen in cultured HFF-1 cells was measured using a
modified ELISA method established in our laboratory. The ef-
fects of the various test compounds on the synthesis of type I
collagen are shown in Fig. 2. All compounds used in the pre-
sent study except for ferulic acid, caffeic acid, curcumin, and
silibinin, showed the ability to increase collagen biosynthesis in
HFF- 1 cells. Vitamin C, UDA, CLA, citric acid and glycolic acid
increased collagen biosynthesis at both concentrations of 10
and 100 μM. Vitamin C at a concentration of 100 μM was the
most effective for collagen biosynthesis.
Fig. 3. Effect of compounds on the synthesis of type I procollagen.
Compounds decreasing type I collagen synthesis The asterisk means significance of p < 0.05 compared to the control.
The treatment of cells with lactic acid, EGCG, resveratrol, and Bars represent the mean ± S.D. from three experiments performed
retinol showed an increase in synthesis of collagen at a con- in triplicate.
centration of 10 μM and a drastic reduction at 100 μM. No dif-
ference was observed between 0 and 10 μM concentrations of
curcumin, but this compound at a concentration of 100 μM, `liN^N differed significantly when the cells were treated with
effectively and significantly reduced type I collagen synthesis (p both 10 and 100 μM of EGCG (p < 0.05, Fig. 4B), with a higher
< 0.05). expression at 10 μM and lower expression at 100 μM, in com-
parison to the control. However, the expression of `liN^O and
Effect of compounds on type I procollagen synthesis jjmN did not differ significantly after treatment of the cells with
The effect of compounds on the synthesis of type I procollagen EGCG (p > 0.05). In case of UDA-treatment, the expression
is shown in Fig. 3. Statistically all compounds showed equiva- levels of all genes were equivalent (p > 0.05, Fig. 4C). `liN^N
lent effects at 10 and 100 μM concentrations on type I procolla- and `liN^O expression were equivalent at all concentrations
gen synthesis, except for EGCG, resveratrol and retinol. How- of resveratrol tested (p > 0.05, Fig. 4D). However, the expres-
ever, in the case of EGCG, resveratrol and retinol there were sion of jjmN=was significantly reduced when compared to the
significant differences at 10 and 100 μM (p < 0.05). The highest control (p < 0.05).
level of collagen synthesis was observed when the cells were
treated with 10 μM resveratrol, which was then drastically re- DISCUSSION
duced at 100 μM (Fig. 3).
In order to develop an easy and quick method of quantitation of
mRNA expression after treating with Vitamin C, EGCG, the effects of various compounds on the synthesis and secre-
UDA and resveratrol tion of type I collagen at the protein level, we established a
The effects of Vitamin C, EGCG, UDA and resveratrol on the modified ELISA method using HFF-1 cells. For the proof of
mRNA expression of `liN^N, `liN^O and jjmN genes are concept, we used thirteen compounds, some of which are
shown in Fig. 4. In case of Vitamin C-treatment (10 and 100 known to be antioxidants while others are collagen-promoting
μM), `liN^O was expressed at higher levels than the control. agents. Type I collagen was measured at the protein level after
Although jjmN was less expressed at a concentration of 100 treatment of HFF-1 cells with Vitamin C and other test com-
μM, this difference in expression was not statistically significant. pounds for 24 h. The results presented here demonstrated that
Expression of `liN^N was equivalent at all concentrations of Vitamin C, UDA, CLA, citric acid and glycolic acid enhanced the
Vitamin C tested (p > 0.05, Fig. 4A). The expression of production of type I collagen in a dose-dependent manner. This
628 Human Type I Collagen Screening System

A B Fig. 4. Alterations in expression level of type I

collagen-related genes after treatment with Vita-
min C (A), EGCG (B), UDA (C), and resveratrol
(D) at different concentrations in HFF-1 cells. The
asterisk means significance of p < 0.05 compared
to the control. Bars represent the mean ± S.D.
from three experiments performed in triplicate.

C D

modified ELISA method was used to effectively and signify- min C increased the level of type I collagen synthesis as well as
cantly quantify type I collagen in comparison to a control group. `liN^O mRNA expression. A possible explanation for these
For validation of this method, we measured the type I procolla- findings is that Vitamin C-treated fibroblasts displayed in-
gen C-peptide (PIP) using type I procollagen C-peptide EIA kit creased transcription of collagen. Previous studies demon-
and also examined the mRNA expression of type I collagen strated that Vitamin C was capable of upregulating íóéÉ=f=Åçää~J=
related genes such as `liN^N, `liN^O, and jjmN. ÖÉå gene expression at the mRNA level in cultured dermal
MMPs are a family of zinc-dependent endoproteinases that fibroblasts in a concentration-dependent manner (Boyera et alKI
play pivotal roles in the dynamic remodeling of the extracellular 1998), which is similar to our findings. Intracellularly accumu-
matrix. Based on substrate preference and structural homology, lated procollagen in Vitamin C-deficient conditions may lead to
MMPs are sub-classified into a number of functional groups: a translational repression of procollagen synthesis. Vitamin C
collagenases, gelatinases, stromelysins, matrilysins, membrane may relieve this block by promoting hydroxyproline formation
type-MMPs and other non-classified MMPs (Visse and Nagase, and, consequently, may enhance the secretion of procollagen
2003). from the cell. It has already been proven that Vitamin C is an
MMPs are frequently overexpressed by various extracellular important factor for the biological expression and deposition of
stimuli including growth factors, cytokines, tumor promoters and collagen in the extracellular matrix (Davidson et alK, 1997) via its
UV radiation. This increase in MMP-related activities may be role as a cofactor for the hydroxylation of proline and lysine
involved in the pathogenesis of diseases such as cancer and residues in procollagen (Myllyla et al., 1984). This hydroxylation
inflammation as well as in physiological processes (Sternlicht is essential for the subsequent assembly of procollagen
and Werb, 2001). It has been also reported that the up- monomers into the triple helix, which in turn is a critical re-
regulation of some MMPs is responsible for the enhanced deg- quirement for the stability of procollagen (Sakakibara et alKI
radation of dermal collagen during chronological and UV- 1973) and its subsequent secretion (Graham et alK, 1995).
induced skin aging (Fisher et alKI 1996; Kawaguchi, 2008). With In the case of EGCG-treatment of cells, type I collagen was
increasing age, the synthesis of collagen becomes lower and increased at 10 μM and decreased at 100 μM treatments and
MMP1 levels increase in sun-protected human skin áå= îáîç `liN^N= mRNA expression showed a similar trend. Another
(Varani et alKI 2000). For this reason, the regulation of MMP study reported that EGCG significantly inhibited cardiac hyper-
activity might be a potential strategy for the prevention and/or trophy in mice and rats, which might be related to a reduction in
treatment of UV-induced skin damage. collagen accumulation and cell proliferation of cardiac tissue
The alteration of type I procollagen levels obtained after (Sheng et alKI 2006). Zhang et alK (2006) showed that green tea
treatment of cells with Vitamin C, citric acid, EGCG and res- extract (GTE) and EGCG dramatically inhibited type I collagen
veratrol followed a similar trend to that of type I collagen. Con- production possibly by interfering with the PI-3K/Akt/mTOR
sidering the trend of data in both cases, this modified ELISA signaling pathway. Nakamuta et alK (2005) reported that EGCG
method is acceptable for the estimation of type I collagen in (50 μM) suppressed type I collagen production in rat hepatic
cultured HFF-1 cells. Moreover, we found some significant stellate cells (HSCs). EGCG also inhibited both collagen pro-
differences in mRNA expression of type I collagen-related duction and collagenase activity (MMP1) in a dose-dependent
genes in a dose-dependent manner after treatment of the cells manner, but did not affect the tissue-inhibitor of matrix metallo-
with Vitamin C and EGCG, especially for `liN^O and proteinase-1 (TIMP-1) production in TWNT-4 (derived from
`liN^N, respectively. From this finding, we were able to corre- human HSC) cells. These findings demonstrated that EGCG
late the elevated mRNA expression of these genes to the in- inhibited collagen production regardless of enhanced collagen
crease in production of type I collagen at the protein level, esti- transcription and suppressed collagenase activity, which sug-
mated using the modified ELISA. gested that EGCG might possess therapeutic potential for the
We demonstrated that the synthesis of type I collagen was treatment of liver fibrosis. This finding correlated with the pre-
enhanced and its secretion was markedly stimulated by Vitamin sent results showing that 10 μM EGCG increased the synthesis
C and other compounds, as shown in Fig. 2. In our study Vita- of collagen while 100 μM decreased it (Fig. 2).
 Md. Abul Hashem et al. 629

A number of studies have reported a consistent effect of citric formed by procollagen peptidase. Multiple tropocollagen mole-
acid on collagen synthesis, which is consistent with our findings cules form collagen fibrils, and multiple collagen fibrils form
(Ruggeri et alK, 2006; Yamamoto et alK, 2006). Furthermore, our collagen fibers, which are available to be deposited in the ex-
results on glycolic acid and lactic acid also correlate with previ- tracellular matrix (Di Lullo et al., 2002; Sakakibara et alKI 1973).
ously published results (Ash et alK, 1998; Bernstein et alK, 2001; The first reported method for the quantitative detection of col-
Hong et alK, 2001; Kneedler et alK, 1998) (Fig. 2). lagen synthesis was competitive radioimmunoassay for procol-
Huang et alK (2004) reported that sodium ferulate (SF) could lagen type I carboxy-terminal peptide using polyclonal antibod-
decrease collagen synthesis by intervening with TGF-β1 activity ies (Taubman et al., 1974). The amount of propeptides cleaved
in HSCs. However, no significant decrease was detected either off from the collagen triple helix molecule during its secretion
in collagen or in procollagen protein levels, in our study (p > outside the cell was assumed to stoichiometrically reflect the
0.05). The amount of type I procollagen was likely to increase amount of collagen molecules synthesized inside the RER.
with treatment of ferulic acid but these changes were not statis- Similar methods using PIP as reference have been used to
tically significant (Fig. 3). study collagen levels with certain health disorders (Boigk et alK,
It was reported that CLA showed no significant impact or 1997; Denton and Black, 2003; Kobayashi et alK, 1999; Quere-
changes in collagen content of the skin in mice (Oikawa et alKI jeta et alK, 2004). The commercially available procollagen type I
=
2006). få îáíêç studies with primary cultured rat mammary C-terminal peptide EIA kit adopted a similar method described
epithelial cells have confirmed the inhibitory effect of CLA on above. Although it is known that the amount of cleaved propep-
cell proliferation and demonstrated that CLA additionally in- tides reflect the amount of collagen synthesized, procollagen
duces apoptosis of these cells (Ip et alKI 1999). These results peptidase is additionally required to synthesize collagen fibrils
were contradictory to our findings where CLA was found to after the C-terminal peptides (propeptides) are cleaved. For this
increase type I collagen synthesis significantly higher than that reason, the direct measurement of collagen molecules may be
of the control (p < 0.05). This variation may be due to different straightforward and useful.
experimental protocols, cell types, concentration of CLA etc. Our data indicated that the synthesis and secretion of type I
Godichaud et alK (2000) stated that trans-resveratrol inhibited collagen in HFF-1 cells were regulated by Vitamin C, UDA, CLA,
the mRNA expression of type I collagen. In our study `liN^O= glycolic acid, citric acid, lactic acid, EGCG, resveratrol, and retinol
mRNA expression was decreased when the cells were treated in a dose-dependent manner. The expression of the íóéÉ=f=Åçää~J
with higher concentrations (100 μM) of resveratrol, which corre- ÖÉå gene by different compounds was also accompanied by an
lated with the above study. Kang et al. (2002) concluded in their elevated production of type I collagen at the protein level. In con-
study that curcumin inhibited collagen synthesis and HSC acti- clusion, this modified method of ELISA can be used for screening
vation áå=îáîç and áå=îáíêç, and thus proved to be a valuable anti- small molecules for the synthesis of type I collagen, which can be
fibrogenic agent. This is in agreement with our findings where used for anti-aging or anti-fibrotic purposes.
type I collagen synthesis was reduced with increasing concen-
trations of curcumin (p < 0.05). ACKNOWLEDGMENT
Silibinin strongly inhibited the growth and survival of human E. Lee and S.-Y. Lim were supported by the Brain Korea 21
endothelial cells via cell cycle arrest and downregulation of Project.
survivin, Akt and NF-κB (Singh et alKI 2005). Silibinin had no
significant effect on type I collagen synthesis in comparison to REFERENCES
the control in our study. Kafi et alK (2007) reported that topical
Ash, K., Lord, J., Zukowski, M., and McDaniel, D.H. (1998). Com-
retinol improved fine wrinkles associated with natural aging. parison of topical therapy for striae alba (20% glycolic
Retinol treatment significantly increased glycosaminoglycan acid/0.05% tretinoin versus 20% glycolic acid/10% L-ascorbic
expression and procollagen. Glycosaminoglycan retains a sub- acid). Dermatol. Surg. OQ, 849-856.
stantial amount of water and increases collagen production, Bernstein, E.F., Lee, J., Brown, D.B., Yu, R., and Van Scott, E.
which is most likely to be responsible for the effacement of (2001). Glycolic acid treatment increases type I collagen mRNA
and hyaluronic acid content of human skin. Dermatol. Surg. OT,
wrinkles. Retinol-treated aged skin resists skin injury success- 429-433.
fully and ulcer formation, promoting an improved appearance Boigk, G., Stroedter, L., Herbst, H., Waldschmidt, J., Riecken, E.O.,
due to the synthesis of greater skin matrix. In our study, retinol and Schuppan, D. (1997). Silymarin retards collagen accumula-
supplement increased type I collagen synthesis and type I pro- tion in early and advanced biliary fibrosis secondary to complete
collagen at a concentration of 10 μM while this was decreased bile duct obliteration in rats. Hepatology OS, 643-649.
at a concentration of 100 μM. This indicated the potential effi- Boyera, N., Galey, I., and Bernard, B.A. (1998). Effect of vitamin C
and its derivatives on collagen synthesis and cross-linking by
cacy of retinol as an anti-wrinkle agent, in a concentration- normal human fibroblasts. Int. J. Cosmet. Sci. OM, 151-158.
dependent manner. Davidson, J.M., Lu Valle, P.A., Zoia, O., Quaglino, D. Jr., and Giro,
The typical process of formation of type I collagen can be M. (1997). Ascorbate differentially regulates elastin and collagen
summarized as follows. Inside the cell, three peptide chains, biosynthesis in vascular smooth muscle cells and skin fibro-
known as preprocollagen, are formed consisting of two alpha-1 blasts by pretranslational mechanisms. J. Biol. Chem. OTO, 345-
and one alpha-2 chain in ribosomes along the Rough Endo- 352.
Denton, C.P., and Black, C.M. (2003). Pulmonary hypertension in
plasmic Reticulum (RER). These peptide chains have a regis- systemic sclerosis, Rheum. Dis. Clin. North Am. OV, 335–349.
tration peptide (propeptide) and a signal peptide on each end of Di Lullo, G.A., Sweeney, S.M., Körkkö, J., Ala-Kokko, L., and San
the peptide chain. Signal peptides are cleaved inside the RER Antonio, J.D. (2002). Mapping the ligand-binding sites and dis-
and then the peptide chains form procollagen. Hydroxylation of ease-associated mutations on the most abundant protein in the
lysine and proline in procollagen is processed with a cofactor, human, type I collagen; J. Biol. Chem. OTT, 4223-4231.
Fisher, G.J., Datta, S.C., Talwar, H.S., Wang, Z.Q., Varani, J., Kang,
Vitamin C, resulting in the formation of the triple helical structure
S., and Voorhees, J.J. (1996). Molecular basis of sun induced
of procollagen inside the RER. Procollagen is shipped to the premature skin ageing and retinoid antagonism. Nature PTV,
Golgi apparatus, where it is packaged and secreted by exocy- 335-339.
tosis. Outside the cell, propeptides attached to each end of Friedman, S.L. (1993). The cellular basis of hepatic fibrosis:
procollagen are cleaved and sequentially tropocollagen is mechanisms and treatment strategies. New Engl. J. Med. POU,
630 Human Type I Collagen Screening System

1828-1835. Querejeta, R., López, B., González, A., Sánchez, E., Larman, M.,
Godichaud, S., Krisa, S., Couronné, B., Dubuisson, L., Mérillon, Martínez Ubago, J.L., and Díez, J. (2004). Increased type I col-
J.M., Desmoulière, A., and Rosenbaum, J. (2000). Deactivation lagen synthesis in patients with heart failure of hypertensive ori-
of cultured human liver myofibroblasts by trans-resveratrol, a gin: relation to myocardial fibrosis. Circulation=NNM, 1263-1268.
grapevine-derived polyphenol. Hepatology PQ, 922-931. Ruggeri, A. Jr., Prati, C., Mazzoni, A., Nucci, C., Di Lenarda, R.,
Graham, M.F., Willey, A., Adams, J., Yager, D., and Diegelmann, Mazzotti, G., and Breschi, L. (2007). Effects of citric acid and
R.F. (1995). Role of ascorbic acid in procollagen expression and EDTA conditioning on exposed root dentin: An immunohisto-
secretion by human intestinal smooth muscle cells. J. Cell. chemical analysis of collagen and proteoglycans. Arch. Oral Biol.
Physiol. NSO, 225-233. RO, 1-8.
Hong, J.T., Kim, E.J., Ahn, K.S., Jung, K.M., Yun, Y.P., Park, Y.K., Sakakibara, S., Inouye, K., Shudo, K., Kishida, Y., Kobayashi, Y.,
and Lee, S.H. (2001). Inhibitory effect of glycolic acid on ultravio- and Prockop, D.J. (1973). Synthesis of (Pro-Hyp-Gly) n of de-
let-induced skin tumorigenesis in SKH-1 hairless mice and its fined molecular weights. Evidence for the stabilization of colla-
mechanism of action. Mol. Carcinog. PN, 152-160. gen triple helix by hydroxypyroline. Biochim. Biophys. Acta PMP,
Huang, J., Hu, J.H., Qiu, L., and Cai, Z. (2004). Mechanisms of 198-202.
sodium ferulate inhibition of collagen synthesis in hepatic stellate Sakata, Y., Chancey, A.L., Divakaran, V.G., Sekiguchi, K.,
cells. Yao Xue Xue Bao, PV, 577-580. Sivasubramanian, N., and Mann, D.L. (2008). Transforming
Ip, M.M., Masso-Welch, P.A., Shoemaker, S.F., Shea-Eaton, W.K., growth factor-beta receptor antagonism attenuates myocardial
and Ip, C. (1999). Conjugated linoleic acid inhibits proliferation fibrosis in mice with cardiac-restricted overexpression of tumor
and induces apoptosis of normal rat mammary epithelial cells in necrosis factor. Basic Res. Cardiol. NMP, 60-68.
primary culture. Exp. Cell. Res. ORM, 22-34. Sheng, R., Gu, Z., Xie, M., Zhou, W., and Guo, C. (2006). EGCG
Kafi, R., Kwak, H.S., Schumacher, W.E., Cho, S., Hanft, V.N., Ham- inhibition against collagen accumulation and cell proliferation in
ilton, T.A., King, A.L., Neal, J.D., Varani, J., Fisher, G.J., Voor- cardiac hypertrophy. Zhongguo Yao Li Xue Bao OO, 1095-1099.
hees, J.J., and Kang, S. (2007). Improvement of naturally aged Singh, R.P., Dhanalakshmi, S., Agarwal, C., and Agarwal, R. (2005).
skin with vitamin A (retinol). Arch. Dermatol. NQP, 606-612. Silibinin strongly inhibits growth and survival of human endothe-
Kang, H.C., Nan, J.X., Park, P.H., Kim, J.Y., Lee, S.H., Woo, S.W., lial cells via cell cycle arrest and downregulation of survivin, Akt
Zhao, Y.Z., Park, E.J., and Sohn, D.H. (2002). Curcumin inhibits and NF-κB: implications for angioprevention and antiangiogenic
collagen synthesis and hepatic stellate cell activation áåJîáîç and therapy. Oncogene OQ, 1188-1202.
áåJîáíêç. J. Pharm. Pharmacol. RQ, 119-126. Son, E.D., Choi, G.H., Kim, H.K., Lee, B.S., Chang, I.S., and
Kawaguchi, H. (2008). Endochondral ossification signals in cartilage Hwang, J.S. (2007). Alpha-ketoglutarate stimulates procollagen
degradation during osteoarthritis progression in experimental production in cultured human dermal fibroblasts, and decreases
mouse models. Mol. Cells OR, 1-6. UVB-induced wrinkle formation following topical application on
Kim, S.W., Jo, B.K., Jeong, J.H., Choi, S.U., and Hwang, Y.I. (2005). the dorsal skin of hairless mice. Biol. Pharm. Bull. PM, 1395-
Induction of extracellular matrix synthesis in normal human fi- 1399.
broblasts by anthraquinone isolated from Morinda citrifolia Sternlicht, M.D., and Werb, Z. (2001). How matrix metallopro-
(Noni) fruit. J. Med. Food. U, 552-555. teinases regulate cell behavior. Annu. Rev. Cell Dev. Biol. NT,
Kneedler, J.A., Sky, S.S., and Sexton L.R. (1998). Understanding 463-516.
alpha-hydroxy acids. Dermatol. Nurs. NM, 247-254. Taubman, M.B., Goldberg, B., and Sherr, C. (1974). Radioimmuno-
Kobayashi, T., Gabazza, E.C., Taguchi, O., Risteli, J., Risteli, L., assay for human procollagen. Science NUS, 1115-1117.
Kobayashi, H., Yasui, H., Yuda, H., Sakai, T., Kaneda, M., et al. Varani, J., Warner, R.L., Gharaee-Kermani, M., Phan, S.H., Kang,
(1999). Type I collagen metabolites as tumor markers in patients S., Chung, J.H., Wang, Z.Q., Datta, S.C., Fisher. G.J., and
with lung carcinoma. Cancer UR, 1951-1957. Voorhees, J.J. (2000). Vitamin A antagonizes decreased cell
Lee, J., Jung, E., Kim, Y., Park, J., Park, J., Hong, S., Kim, J., Hyun, growth and elevated collagen-degrading matrix metallopro-
C., Kim, Y.S., and Park, D. (2006). Asiaticoside induces human teinases and stimulates collagen accumulation in naturally aged
collagen I synthesis through TGF receptor I kinase (TRI kinase)- human skin. J. Invest. Dermatol. NNQI=480-486.
independent Smad signaling. Planta. Medica. TO, 324-328. Visse, R., and Nagase, H. (2003). Matrix metalloproteinases and
Lee, J., Jung, E., Lee, J., Huh, S., Kim, J., Park, M., So, J., Ham, Y., tissue inhibitors of metalloproteinases: structure, function, and
Jung, K., Hyun, C.G., et al. (2007). Panax ginseng induces hu- biochemistry. Circ. Res. VO, 827-839.
man type I collagen synthesis through activation of Smad signal- Yamamoto, Y., Uede, K., Yonei, N., Kishioka, A., Ohtani, T., and
ing. J. Ethnopharmacol. NMV, 29-34. Furukawa, F. (2006). Effects of alpha-hydroxy acids on the hu-
Myllyla, R., Majamaa, K., Gunzler, V., Hanauske-Abel, H.M., and man skin of Japanese subjects: the rationale for chemical peel-
Kivirikko, K.I. (1984). Ascorbate is consumed stoichiometrically ing. J. Dermatol. PP, 16-22.
in the uncoupled reactions catalyzed by prolyl 4-hydroxylase Yang, C.M., Kandaswamy, V., Young, D., and Sen, S. (1997).
and lysyl hydroxylase. J. Biol. Chem. ORV, 5403-5405. Changes in collagen phenotypes during progression and re-
Nakamuta, M., Higashi, N., Kohjima, M., Fukushima, M., Ohta, S., gression of cardiac hypertrophy. Cardiovasc. Res. PS, 236-245
Kotoh, K., Kobayashi, N., and Enjoji, M. (2005). Epigallocate- Zannad, F., Dousset, B., and Alla, F. (2001). Treatment of conges-
chin-3-gallate, a polyphenol component of green tea, sup- tive heart failure: interfering the aldosterone-cardiac extracellular
presses both collagen production and collagenase activity in he- matrix relationship. Hypertension PU, 1227-1232.
patic stellate cells. Int. J. Mol. Med. NS, 677-681. Zhang, Q., Kelly, A.P., Wang, L., French, S.W., Tang, X., Duong,
Oikawa, D., Nakanishi, T., Nakamura, Y.N., Yamamoto, T., Yama- H.S., Messadi, D.V., and Le, A.D. (2006). Green tea extract and
guchi, A., Shiba, N., Iwamoto, H., Tachibana, T., and Furuse, M. (-) -epigallocatechin-3-gallate inhibit mast cell-stimulated type I
(2005). Modification of skin composition by conjugated linoleic collagen expression in Keloid fibroblasts via blocking PI-3K/Akt
acid alone or with combination of other fatty acids in mice. Br. J. signaling pathways. J. Invest. Dermatol. NOS, 2607-2613.
Nutr. VQ, 275-281.