Original Article

Effects of Centella asiatica

ISSN No 2230-7885
CODEN JPBSCT

Ethanolic Extract Encapsulated NLM Title

DOI
J Pharm Biomed Sci
http://dx.doi.org/10.20936/jpbms/160246

in Chitosan Nanoparticles on Linda Yulianti1*,

Proliferation Activity of Skin Kusmarinah Bramono2,
Etik Mardliyati3,
Fibroblasts and Keratinocytes, Hans-Joachim Freisleben4,5
1 
Biomedical Sciences Doctoral Program,

Type I and III Collagen Faculty of Medicine, University of Indonesia,

Salemba Jakarta, Indonesia

Synthesis and Aquaporin 3

2 
Department of Dermatovenerology, Faculty
of Medicine, University of Indonesia,
Salemba Jakarta, Indonesia
Expression In Vitro 3 
Centre for Pharmaceutical and Medical
Technology, Agency for the Assessment
and Application of Technology,
ABSTRACT
4 
Medical Research Unit, Faculty of
Background  The activity of skin cell proliferation, collagen synthesis and skin hydration Medicine, University of Indonesia,
decrease with the process of aging; therefore, the skin looks dull, dry and sagging. Salemba Jakarta, Indonesia
Aquaporin 3 (AQP3) is a key protein that plays a major role in skin cell proliferation and 5 
German–Indonesian Medical Association
skin hydration. Retinoic acid (RA) is still considered for anti-aging treatment, but it fre- (DIGM e. V.)
quently shows side effects, such as skin irritation. Centella asiatica (CA) formulation in
chitosan nanoparticles has a promising potential as anti-aging cosmetic.  Address reprint requests to:
*Dr. Linda Yulianti, MD, Biomedical
Aim  The aim of this study was to evaluate the effects of CA ethanolic extract encapsu-
Sciences Doctoral Program, Faculty of
lated in chitosan nanoparticles (CAEE + CNP) on skin cell proliferation, collagen synthesis Medicine, University of Indonesia,
and AQP3 expression in vitro, when compared to RA. Salemba Jakarta, Indonesia
Methods  Microculture tetrazolium assay was conducted to analyse the proliferation of E-mail: lindajuliantiwijayadi@gmail.com
normal human dermal fibroblasts (NHDF) and normal human epidermal keratinocytes  Article citation: Yulianti L, Bramono K,
(NHEK) at 24, 48 and 72 h. Type I and III collagen synthesis was evaluated at the same Mardliyati E, Freisleben HJ. Effects of
time points using ELISA. Aquaporin 3 expression at 24 h was evaluated using immunocy- Centella asiatica ethanolic extract
tochemistry and measured quantitatively using ImageJ software. All treatments involved encapsulated in chitosan nanoparticles
several concentrations of CAEE + CNP and RA through serial dilution. on proliferation activity of skin fibroblasts
Results  Collagen type I and III synthesis of NHDF and NHEK was neither significantly and keratinocytes, type I and III
different from untreated controls nor from RA-treated cells. Nevertheless, CAEE + CNP collagen synthesis and aquaporin 3
expression in vitro. J Pharm Biomed Sci
stimulated the proliferation of both NHDF and NHEK. Additionally, AQP3 expression in both
2016;06(05):315–327.
cell types was upregulated by CAEE + CNP.
Available at www.jpbms.info
Conclusion  CAEE + CNP is a promising formulation for anti-aging activity by inducing
skin cell proliferation and AQP3 expression. The clinical trials are still needed to evaluate Statement of originality of work: The
skin hydration in vivo. manuscript has been read and approved by all
the authors, the requirements for authorship have
Keywords  aquaporin 3, Centella asiatica, chitosan nanoparticles, collagen, fibroblasts, been met, and that each author believes that the
hydration, keratinocytes, proliferation manuscript represents honest and original work.
Source of funding: None.
Acknowledgment: The authors wish to thank
Dani, IntanRazari, Ita and Labibah for laboratory
INTRODUCTION assistance and Indra Kusuma for the cell
Skin aging is a complex process controlled by genetic determination (chrono- cultures of NHDF and NHEK.
logic aging) and under the influence of external factors (extrinsic aging). Competing interest / Conflict of interest: The
Skin aging leads to various modifications in cells and tissue1. author(s) have no competing interests for financial
support, publication of this research, patents
A number of studies have shown that aging process results in reduced
and royalties through this collaborative research.
proliferation activity in both keratinocytes and fibroblasts. Furthermore, cell All authors were equally involved in discussed
renewal rate and the synthesis of type I and type III collagen in skin fibro- research work. There is no financial conflict with
blasts decreased naturally in skin as the years go by. This loss of regenerating the subject matter discussed in the manuscript.
power is both, cause and a measure of skin aging1. Disclaimer: Any views expressed in this paper
Aging skin is manifested by the diminished water content of stratum cor- are those of the authors and do not reflect the
neum (SC). Age-related trans-epidermal water loss (TEWL) can be ­evaluated official policy or position of the Department of
by removing the superficial epidermal layer by tape stripping. In aged skin, Defense.

Copyright © 2016

Received Date: 22 April 2016 – Accepted Date: 14 May 2016 – Published Online: 21 May 2016
 316 Linda Yulianti
the barrier repair capacity is significantly impaired2.
Hydration of the SC determines the appearance and the
physical properties of the skin. The aging process reduces
skin surface hydration, which results in dry, dull, coarse
and saggy appearance with loss of elasticity. Other visi-
ble signs include pigmentation and appearance of benign
skin tumours. Moreover, the epidermis becomes thinner
with age and as a result wrinkles develop2–4.
Cosmeceutical anti-aging formulations contain active
ingredients to achieve local therapeutic effects, and should
not have harmful systemic side effects2. Well known
anti-aging ingredients like vitamin A or retinoic acid (RA)
and its derivatives (retinoids), have harmful side effects
by irritating the skin and causing dry, flaky skin, which is
sensitive to light5. Moreover, usage of these retinoids con-
taining formulations is limited, as they may cause contact
RA dermatitis and have teratogenic effects5,6. Therefore,
it is necessary to develop innovative cosmeceutical anti-­
aging formulations with active ingredients which are
effective and stable, yet safe and nontoxic. Additionally,
they should be compatible with other active ingredients
and be delivered efficiently to the target cells7,8.
Indonesian biodiversity promises great potential
for selection and development of herbal cosmetics and
cosmeceuticals, which are effective and safe. Natural
products contain a wealth of interesting and possibly
beneficial cosmepharmaceutically active compounds.
For example, Pereda et al.9 reported that hydroglycolic
Piptadenia colubrina extract can improve skin hydration.
Centella asiatica (CA) (gotu kola, Hydrocotyle asiatica) is
often found in Indonesia and has been used for vari-
ous medicinal purposes since 1,000 years ago, CA has
been applied for wound healing and the treatment of
asthma, ulcers, leprosy, lupus erythematosus, psoriasis,
vein disease and cancer10–13. CA stimulates the synthe-
sis of collagen for skin tissue regeneration14–16. The bio-
logically active ingredients in CA are triterpenes namely
asiatic acid, madecassic acid, asiaticoside and madecassoside12,13.
Although CA extracts (CAE) possess high potential of
biological activities, their clinical usage is limited due
to poor physical stability, with CA extracts having high
hygroscopic index. Powder extracts are promptly liq-
uefied within a few minutes when exposed to normal
environment. Therefore, nanoparticles which entrap
the extract and protect it from external moisture
should be developed for stabilisation17.
Chitosan is a biopolymer produced from chitin by
partial deacetylation. The polysaccharide chitin is the
main compound of the exoskeletons in crustaceans and
fungal cell walls. Chitin nanofibrils are applied in cos-
metics as carriers to increase penetration18,19, as well as
active ingredients for anti-aging cosmetics as it helps
to maintain the integrity of skin barrier and increases
the capacity of corneocytes to store water20. Chitosan
consists of acetylated and deacetylated units. Later, they
are composed of β-(1,4)-d-glucosamine, whereas,
the former units consist of N-acetyl-d-glucosamine.
Chitosan is hypoallergenic and has natural antibacterial
properties, it is biocompatible and biodegradable and
its low cost warrants biomedical application in wound
healing21,22. Furthermore, chitosan is used in anti-­aging
skin cosmetics, both as penetration enhancer and as
active anti-aging compound19.
Nanocarrier systems as innovative drug and cosmetic
delivery technology have been widely used23,24. This
technology is commonly based on lipid carriers, such as
liposomes and solid lipid nanoparticles with 100–300
nm in diameter23–25. Nanoparticles have unique physi-
cal properties which make them ideal for various skin
care products26. For example, a lipid nanocarrier (LNC)
system loaded with tocopheryl acetate (TA) exhibited
the ability to enhance skin hydration26. Apart from lip-
id-based carriers, polymeric nanoparticles are being devel-
oped as novel delivery systems, including those based on
chitin/chitosan27,28.
Aquaporins (AQPs) as water channels are the key
factors in skin hydration mechanisms29. The most abun-
dant AQP in the skin is AQP3, a water/glycerol trans-
porting protein. However, its expression in normal
human dermal fibroblasts (NHDF) and normal human
epidermal keratinocytes (NHEK) decreases with the
aging process. Non-expression of AQP3, particularly in
the basal layer leads to reduced SC hydration and skin
elasticity30. Therefore, AQP3 is a key protein target for
future anti-aging treatment to improve durability, tex-
ture and quality of the skin surface31. Naturally active
compounds that can stimulate AQP3 expression would
be effective hydrating agents as emollients in anti-aging
cosmetics31.
RA influences keratinocyte differentiation, reduces
epidermal cell adhesion and facilitates corneocyte exfo-
liation from the surface of epidermis. In the dermis,
retinoids regulate fibroblast proliferation, induce angio-
genesis and play an essential role in the synthesis of col-
lagen and elastin fibrils32. Moreover, it was reported that
RA stimulates AQP3 expression in NHEK33.
In this study, we evaluated the effect of CA ethanolic
extract (CAEE) encapsulated into chitosan nanoparticles
(CAEE + CNP) on the proliferative activity of NHDF and
NHEK, on the synthesis of type I and type III collagen
in NHDF and determined the ability of CAEE + CNP to
increase AQP3 expression in NHDF and NHEK.
MATERIALS AND METHODS
Materials
RA was purchased from BASF, Germany and chitosan
with a deacetylation degree >75% from Sigma-Aldrich,
USA. All other chemicals were obtained from Merck,
Germany; the assay kits used were purchased from
Sigma-Aldrich, via their subsidiaries in Indonesia or
Singapore. CA, from Tawangmangu, was used to prepare
the ethanolic extract in the Centre for Pharmaceutical
and Medical Technology, Agency for the Assessment and
Application of Technology (BPPT PUSPITEK), Serpong.
J Pharm Biomed Sci | Vol. 06 No. 05 | 315–327
 Effects of CA ethanolic extract in chitosan nanoparticles 317

CAEE was encapsulated into chitosan nanoparticles by Fibroblast collagen synthesis activity was evaluated at 24
ionic gelation method34 using Sigma-Aldrich chitosan and 48 h after treatment and compared against untreated
with a deacetylation degree of >70%. NHDF as negative control. Results for collagen synthesis
Serial concentrations of all compounds used were were reported in mg/mL.
obtained by 1 + 1 dilution steps with the respective sol-
vent: CAEE + CNP (100, 50, 25, 12.5, 6.25 and 3.125
Immunocytochemistry test
mg/mL); RA (10, 5, 2.5, 1.25, 0.625 and 0.3125 mg/mL)
and (1, 0.5, and 0.25 mg/mL). Immunocytochemistry is a specific protein detection
Serial concentrations of all compounds used were technique using a specific antibody (Ab) to recognise
obtained by 1 + 1 dilution steps with the respective the antigen presented on cells. Immunochemical stain-
solvent: CAEE + CNP from 100, 50, 25, 12.5, 6.25 ing was done by using anti-aquaporin 3 Ab, Ab125219.
mg/mL down to 3.125 mg/mL; RA from 10, 5, 2.5, Then, the expression of AQP3 in both NHDF and NHEK
1.25, 0.625 mg/mL down to 0.3125 mg/mL to deter- was quantitatively analysed by using ImageJ.
mine proliferation activity of NHDF and NHEK. RA
dilution steps from 1, 0.5 mg/mL to 0.25 mg/mL were
STATISTICAL ANALYSIS
used to determine type I and III collagen synthesis and
in immunocytochemistry. The data obtained from this study are presented as
images, tables or graphs. To assess significance across the
different treatment groups, unpaired t-test and one-way
Cell cultures of human fibroblasts and ANOVA test were used, followed by post-hoc Tukey test.
keratinocytes Significance level was set at P < 0.05 using SPSS software.
In this study, primary NHDF and NHEK cells were
derived from male children’s foreskin and cultured
RESULTS
in Roswell Park Memorial Institute (RPMI) medium
supplemented with 10% fetalbovine serum (FBS) and Figure 1 shows the transmission electron micrographs
1% penicillin–streptomycin from the Cell Culture (TEM) of CNP and CAEE + CNP.
Laboratory Faculty of Medicine, University of Gajah Particle size distribution of CAEE + CNP analysed
Mada, Yogyakarta. NHDF cell culture from 7th passage with Beckman Particle Size Analyzer CV. Main peak was
was used to test the synthesis of type I and III col- obtained at 96.3 ± 26.6 nm with a polydispersity index
lagen as well as for the immunocytochemical test in (PI) of 0.308. The cumulative percentage of this peak
the Integrated Laboratory of Yarsi University, Jakarta. amounts to 100%, which means that there is only one
Meanwhile, NHEK cell culture was used in the same size population with uniform distribution.
laboratory to test keratinocyte proliferation and immu- Figure 2 compares cell proliferative activity of cul-
nocytochemistry. Early, the cell passages between 2 and tured NHDF after exposure with serial dilutions of
7 were used in all experiments. CAEE + CNP and RA at 24, 48 and 72 h. At 24 h stim-
ulation, fibroblast proliferation is high for 6.25 mg/mL
CAEE + CNP with ~160% of control. This is significantly
MTT assay of fibroblast and keratinocyte higher as compared to untreated control and RA-treated
proliferation cells. At 48 h, both compounds demonstrate similar results
MTT assay (MTT = 3-(4,5-dimethylthiazol-2-yl)-2,5-­ with maximum effects seen on 3.125 mg/mL concentra-
diphenyl-tetrazolium bromide) was adapted to measure tion. After 72 h, RA shows dominant effects on fibroblast
fibroblast and keratinocyte cell proliferation. Cells were proliferation. The statistical analysis with unpaired t-test
seeded at a ratio of 1:3 and 1 × 106 cells/mL density, revealed no significant difference between the two groups.
before subjected to treatment with 0.3125, 0.625, 1.25, Figure 3 compares type I collagen synthesis in cul-
2.5, 5 and 10 mg/mL of RA and 6 concentration (3.125, tured NHDF after 24, 48 and 72 h of exposure with
6.25, 12.5, 25, 50, 100 mg/mL) of CAEE + CNP. The CAEE + CNP and RA. The different concentrations of
MTT assay was repeated three times. The number of CAEE + CNP were obtained by serial dilution from 12.5
viable cells was evaluated at 24, 48 and 72 h after treat- mg/mL down to 3.125 mg/mL and with RA by serial
ment, as an indicator of fibroblast proliferation, fur- dilution from 0.1 mg/mL down to 0.025 mg/mL.
thermore at 24 and 48 h for keratinocytes proliferation. In general, CAEE + CNP and RA exerted similarly
Subsequently, obtained values were normalised against positive effects as compared to negative control. After 24
untreated control group (100%). and 48 h of exposure, CAEE + CNP and RA did not differ
from control. At 72 h, however, both CAEE + CNP and
RA showed greater type I collagen synthesis, as com-
Evaluation on type I and III collagen synthesis pared to control group. Comparing for all time periods
using ELISA (24, 48 and 72 h) through one-way ANOVA test, there
Type I and III collagen synthesis was examined by using were no statistically significant differences between
designated ELISA kit Alpha 1 from CUSABIO Life science, US. CAEE + CNP, RA and control at all concentrations.

J Pharm Biomed Sci | Vol. 06 No. 05 | 315–327

 318 Linda Yulianti

a b

Fig. 1  Chitosan nanoparticles (CNP) (a) and Centella asiatica ethanolic extract (CAEE) encapsulated into chitosan nanoparticles (CAEE +
CNP) (b); magnification ×100,000; black bars = 50.0 nm.

Fig. 2  Comparison of cell proliferative activity of NHDF without treatment (control), after exposure to Centella asiatica ethanolic extract
encapsulated into chitosan nanoparticles (CAEE + CNP) and retinoic acid (RA) at various concentrations at 24, 48 and 72 h. The con-
centrations of CAEE + CNP were obtained by serial dilution from 100 mg/mL down to 3.125 mg/mL and RA from 10 mg/mL down to
0.3125 mg/mL.

Figure 4 compares type III collagen synthesis in cul-
tured human fibroblasts after 24, 48 and 72 h of exposure
to CAEE + CNP and RA. The different concentrations of
CAEE + CNP were obtained by serial dilution from 12.5
mg/mL down to 3.125 mg/mL and with RA by serial
dilution from 1 mg/mL down to 0.25 mg/mL. Type III
collagen synthesis generally increases over time (with its
highest value at 48 h) and fluctuates slightly more than
type I collagen (compare with controls in Figs. 3, 4).
After 24 h, results from CAEE + CNP, RA and con-
trols were not significantly different at 3.125 mg/mL
(P = 0.151) using one-way ANOVA test.
After 48 h of exposure to CAEE + CNP and RA, the
lowest concentration, synthesis of type III collagen pro-
vided a higher yield than control. At 72 h, control had
the highest value. Overall, it appears that the amount of
synthesised collagen does not strictly depend on the time
exposure and concentrations of the test compounds.

J Pharm Biomed Sci | Vol. 06 No. 05 | 315–327

 Effects of CA ethanolic extract in chitosan nanoparticles 319

Fig. 3  Comparison of type I collagen synthesis in NHDF without treatment (control) and after 24, 48 and 72 h of exposure to Centella
asiatica ethanolic extract encapsulated into chitosan nanoparticles (CAEE + CNP) and retinoic acid (RA) at various concentrations for 24,
48 and 72 h. The concentrations of CAEE + CNP were obtained by serial dilution from 12.5 mg/mL down to 3.125 mg/mL and RA from
1 mg/mL down to 0.25 mg/mL.

Fig. 4  Comparison of type III collagen synthesis in NHDF without treatment (control) and after 24, 48 and 72 h of exposure to Centella
asiatica ethanolic extract (CAEE) encapsulated into chitosan nanoparticles (CAEE + CNP) and retinoic acid (RA) at various concentrations
for 24, 48 and 72 h. The concentrations of CAEE + CNP were obtained by serial dilution from 12.5 mg/mL down to 3.125 mg/mL and
with RA by serial dilution from 1 mg/mL down to 0.25 mg/mL.

After 48 h, there is no statistically significant difference compared to 24 h exposure, though there is an increase
at all the concentrations tested between CAEE + CNP, RA in collagen synthesis with CAEE + CNP and RA.
and control (P-value >0.05) using one-way ANOVA After 72 h, CAEE + CNP and RA showed no statisti-
test. These results are not significantly different when cally significant differences using one-way ANOVA test,

J Pharm Biomed Sci | Vol. 06 No. 05 | 315–327

 320 Linda Yulianti
although there was a decrease in type III collagen syn-
thesis activity with CAEE + CNP and RA compared with
24 and 48 h of exposure.
Figure 5 compares cell proliferative activity of cul-
tured NHEK after exposure with CAEE + CNP and RA at
serial dilutions for 24 and 48 h.
After 24 h of treatment, RA showed maximum result
with 155% of control at a concentration of 5 mg/mL,
while CAEE + CNP gave 145% at 3.125 mg/mL. Overall,
CAEE + CNP stimulated lower proliferative effect than
RA. Unpaired t-test statistical analysis showed no signif-
icant difference.
Keratinocyte proliferation at 48 h looked similar for
the two compounds tested. CAEE + CNP at all concentra-
tions (between 3.125 and 25 mg/mL) and RA (between
0.3125 and 5 mg/mL) stimulated proliferation to 110–
120% of control (100%). Interestingly, the two high-
est concentrations of CAEE + CNP and the highest RA
concentration showed lower proliferative activity than
control (<100%).
Figure 6 shows AQP3 expression in NHDF with-
out and with exposure to CAEE + CNP for 24 h, as
detected by immunocytochemistry. The results showed
an increase in protein AQP3 expression in NHDF after
exposure for 24 h to CAEE + CNP at 3.125, 6.25 and
12.5 mg/mL. Semi-quantitatively, we were able to
obtain optimal results at the highest concentration of
12.5 mg/mL.
Figure 7 shows protein expression of AQP3 in cul-
tured NHDF without and with exposure to RA for 24 h,
as detected by immunocytochemistry. Similar AQP3
expression was seen at all concentration, although with
a reverse trend as compared to CAEE + CNP.
Fig. 5  Comparison of cell proliferative activity of NHEK without treatment (control) and after exposure to Centella asiatica ethanolic extract
encapsulated into chitosan nanoparticles (CAEE + CNP) and retinoic acid (RA) at various concentrations for 24 and 48 h. The concentrations
of CAEE + CNP were obtained by serial dilution from 100 mg/mL down to 3.125 mg/mL and RA from 10 mg/mL down to 0.3125 mg/mL.
The results showed an increase in AQP3 protein
expression in NHDF after exposed to RA at concentration
of 0.25, 0.5 and 1 mg/mL. Semi-quantitatively, optimal
results were gained at 0.25 mg/mL concentration of RA.
Figure 8 shows AQP3 expression in cultured NHDF
before and after treatment with CAEE + CNP and RA
for 24 h, as quantitatively analysed with computer
software ImageJ.
Quantified particle density from ImageJ analysis
demonstrated that CAEE + CNP stimulated the expres-
sion of AQP3 protein in a concentration-dependent
manner, with optimal concentration at 12.5 mg/mL.
In contrast, the expression of AQP3 in dermal fibroblast
after treated with RA decreased with increasing concen-
tration (0.25–0.5 and 1 mg/mL).
By using one-way ANOVA test, we obtained a sta-
tistically significant difference on AQP3 expression in
cultured NHDF (P < 0.001) after being exposed with
CAEE + CNP and RA at concentration of 3.125 and 0.25
mg/mL, respectively, compared with control. Further­
more, with post-hoc Tukey test, there is also a signif-
icant difference between CAEE + CNP groups and RA
groups (P < 0.001).
Another significant difference on AQP3 expression
in NHDF was also obtained by exposing 6.25 mg/mL
of CAEE + CNP and 0.5 mg/mL of RA compared with
control (P < 0.05). The following post-hoc Tukey test
showed a significant difference between CAEE + CNP
and RA (P = 0.006).
This study revealed a significant difference of AQP3
expression in NHDF after being exposed to CAEE + CNP
and RA at concentration of 12.5 and 1 mg/mL, respec-
tively, compared with control (P < 0.05). Another
J Pharm Biomed Sci | Vol. 06 No. 05 | 315–327
 Effects of CA ethanolic extract in chitosan nanoparticles 321

a b

c d

Fig. 6  Aquaporin 3 (AQP3) expression (brown colour, pointed by arrows) in NHDF: (a) without treatment (control); and after 24 h exposed to
CAEE + CNP at concentration of (b) 3.125 mg/mL; (c) 6.25 mg/mL; and (d) 12.5 mg/mL; immunochemical staining with anti-aquaporin 3
antibody Ab125219.

statistical analysis was conducted by using post-hoc Tukey
test, and there is no significant difference between CAEE
+ CNP and RA.
Figure 9 shows protein expression of AQP3 in cul-
tured NHEK without and with exposure to CAEE + CNP
for 24 h, as detected by immunocytochemistry. After 24
h of exposure to CAEE + CNP concentration at 3.125
and 6.25 mg/mL, there was an increased expression of
AQP3 protein in the cytoplasm of keratinocytes com-
pared to control, while the exposure of 12.5 mg/mL
actually decreased AQP3 expression. These results indi-
cate that AQP3 expression in cultured NHEK showed
inverted concentration dependence upon exposure with
CAEE + CNP. Similarly, quantitative analysis was done
using ImageJ software.
Figure 10 shows protein expression of AQP3 in cul-
tured NHEK without and with exposure to RA for 24 h,
as detected by immunocytochemistry. AQP3 expression
on the keratinocytes increased after exposure to RA at
concentration of 0.025, 0.05 and 0.1 mg/mL. The inten-
sity of protein AQP3 expression in keratinocytes increases
proportionally with the increasing of RA concentration.
Figure 11 shows AQP3 expression in cultured NHEK
as quantitatively analysed using computer software ImageJ.
Qualitative results of immunocytochemistry were anal-
ysed by using ImageJ software to obtain quantitative data
for further statistical evaluation.
From the immunocytochemistry results on kerati-
nocytes, the quantitative analysis based on colour den-
sity showed that CAEE + CNP at concentration of 3.125
and 6.25 mg/mL stimulated AQP3 protein expression
slightly higher than control. For RA stimulation, how-
ever, there is a linear correlation between AQP3 protein
expression and the treatment concentration.
Statistical analysis by using one-way ANOVA test
showed no significant difference on AQP3 expression in
cultured NHEK after being exposed with CAEE + CNP
and RA at concentration of 3.125 and 0.25 mg/mL,
respectively, compared with control.
After exposure with 6.25 mg/mL of CAEE + CNP
and 0.5 mg/mL of RA showed a significant difference
on AQP3 expression in NHEK (P < 0.05) compared with
control, as statistical analysis was conducted by one-way
ANOVA test. Further statistical analysis using post-hoc
Tukey test showed no significant difference between
CAEE + CNP and RA.
Using one-way ANOVA test, we obtained another
significant difference on AQP3 expression in NHEK after

J Pharm Biomed Sci | Vol. 06 No. 05 | 315–327

 322 Linda Yulianti

a b

c d

Fig. 7  Aquaporin 3 (AQP3) expression (brown colour, pointed by arrows) in NHDF: (a) without treatment (control); and after 24 h exposed
to retinoic acid (RA) at concentrations of: (b) 0.25 mg/mL, (c) 0.5 mg/mL, and (d) 1 mg/mL; immunochemical staining with anti-aquaporin
3 antibody Ab125219.

Fig. 8  Quantitative analysis of AQP3 expression in cultured NHDF by computer software ImageJ. *Statistically significant difference
found between 3.125 mg/mL of CAEE + CNP and RA 0.25 mg/mL using one-way ANOVA and post-hoc Tukey tests (P < 0.05).
**Statistically significant difference found between 6.25 mg/mL of CAEE + CNP and RA 0.5 mg/ mL using post-hoc Tukey test (P < 0.05).
***Statistically significant difference found between 0.25 mg/mL of RA and control using post-hoc Tukey test (P < 0.05). ****Statistically
significant difference found between 0.5 mg/mL of RA and control using post-hoc Tukey test (P < 0.05).

J Pharm Biomed Sci | Vol. 06 No. 05 | 315–327

 Effects of CA ethanolic extract in chitosan nanoparticles 323

a b

c d

Fig. 9  Aquaporin 3 (AQP3) expression (brown colour, pointed by arrows) in NHEK: (a) without treatment (control); and after 24 h exposed
to CAAE + CNP at concentration of: (b) 3.125 mg/mL; (c) 6.25 mg/mL and (d) 12.5 mg/mL; immunochemical staining with anti-aquaporin
3 antibody Ab125219.

a b

c d

Fig. 10  Aquaporin 3 (AQP3) expression (brown colour, pointed by arrows) in NHEK: (a) without treatment (control) and after 24 h exposed
to retinoic acid (RA) at concentration of: (b) 0.25 mg/mL; (c) 0.5 mg/mL and (d) 1 mg/mL; immunochemical staining with anti-aquaporin
3 antibody Ab125219.

J Pharm Biomed Sci | Vol. 06 No. 05 | 315–327

 324 Linda Yulianti
Fig. 11  Quantitative analysis of AQP3 expression in cultured keratinocytes by computer software ImageJ. *Statistically significant difference
found between 12.5 mg/mL of CAEE + CNP and control using one-way ANOVA and post-hoc Tukey tests (P < 0.05). **Statistically sig-
nificant difference found between 0.5 mg/mL of RA and control using post-hoc Tukey test (P < 0.05). ***Statistically significant difference
found between 1 mg/mL of RA and control using post-hoc Tukey test (P < 0.05). ****Statistically significant difference found between 12.5
mg/mL of CAEE + CNP and control using post-hoc Tukey test (P < 0.05).
being exposed with 12.5 mg/mL of CAEE + CNP and
1 mg/mL of RA compared with control (P < 0.05).
The following post-hoc Tukey test showed no significant
difference between CAEE + CNP and RA (P < 0.05).
DISCUSSION
RA is a potent stimulator of keratinocyte proliferation
and regulator of cell differentiation. It increases the
thickness of stratum granulosum and is widely used in
cosmetic anti-aging treatment of the skin35. Since RA
has harmful side effects, it is interesting to search for
alternatives, i.e., other active substances with better
pharmacological profile (bioavailability, toxicity) as
compared to RA.
In this study, we used CAEE, which had shown better
proliferative activity in preliminary experiments than the
aqueous extract (data not shown). A study in Malaysia had
also proven that ethanolic extracts have better biological
effects than aqueous extract of CA36. Chitosan–alginate
nanoparticles developed as carriers for CA extracts were
considered as promising systems to stabilise the extracts17.
Hence, we used our own CNP to encapsulate our CAEE
(CAEE + CNP).
Leonida (2012) investigated the activity of chitosan
as antimicrobial, wound-healing and anti-aging com-
pounds. The authors reported that ‘nano-sizing’ enhanced
its effectiveness23. Moreover, when well-known sub-
stances are prepared into nanoparticles, they may dis-
play completely different properties and can behave in
an unpredictable manner. This means that nanoparticles
could re-invent the properties of substances currently
used in cosmetic dermatology to create novel chemical
entities from the old ones8.
In this study, we used chitosan with a degree of
deacetylation more than 85% (not consistent with
‘Materials and Methods’ Section). A study by Howling
et al. (2001) demonstrated that chitosan with a relatively
high degree of deacetylation strongly stimulated fibro-
blast proliferation while chitosan with lower levels of
deacetylation showed less activity22.
Proliferation of cultured NHDF
From our results, it can be concluded that the prolifera-
tive effect of CAEE + CNP is not strictly concentration-­
dependent but rather time-dependent (Fig. 3). After
24 h of exposure, CAEE + CNP showed a generally better
proliferative effect on NHDF (with at a peak at 6.25 mg/
mL = 160% of controls) than RA. After 48 h, there are
no significant differences between CAEE + CNP and RA.
After 72 h, the effect of RA dominated the picture with
a peak of about 180% of controls (Fig. 3).
A study by Wu et al.37 investigated the effect of four
triterpenes from CA on the proliferation of NHDF. At low
concentrations (1, 3 and 10 µM), all compounds tested
did not enhance proliferation compared with the control
group. Meanwhile, Song et al.38 showed that madecasso-
side, a bioactive compound of CA-inhibited proliferation
of keloid fibroblasts in a time and concentration-­
dependent manner. A study by Lu et al. (2004)39 found
that asiaticoside, another compound of CA, strongly
induced cell-cycle progression, proliferation and col-
lagen synthesis in dermal fibroblasts. Asiaticoside also
increased migration of keratinocytes by ~20% in wound
healing process)40.
In the dermis, RA was found to regulate fibroblast
proliferation and play a role in aging prevention because
J Pharm Biomed Sci | Vol. 06 No. 05 | 315–327
 Effects of CA ethanolic extract in chitosan nanoparticles
of its inhibitory effects on the expression of metallo-­
proteinases41. Varani et al.35 demonstrated that the abil-
ity of RA to stimulate NHDF proliferation depends on
the concentration of Ca2+ in the extracellular environ-
ment. Although RA exerts the above effects and has also
such clinical activity in dermal tissue, all-trans-RA (ATRA)
reduced proliferation in NHDF cultures42. Moreover,
retinoids generally decreased NHDF proliferation after
48 h of incubation.
Collagen expression in cultured NHDF
Type I and III collagens are the major structural and
functional components of skin connective tissue. Hence,
we were interested to investigate the efficacy of CAEE
+ CNP on the biosynthesis of these collagens in NHDF
compared to RA. We found that the effect of CAEE + CNP
on type I collagen synthesis at 24 and 48 h of exposure
was not significantly different from RA and controls. At
72 h, control values were slightly decreased, but CAEE
+ CNP and RA maintained type I collagen synthesis on
the same level.
Type III collagen synthesis fluctuated and generally
increased over time with its highest control value at 48 h;
exposure to CAEE + CNP and RA at their lowest concen-
trations slightly stimulated biosynthesis at 48 h. At 72 h,
control value was higher than all other values; in other
words, all the treatment had inhibitory effects on type
III collagen synthesis.
In a study from Malaysia, CA was shown to have stim-
ulatory effect on collagen synthesis in a concentration-­
dependent manner. At 50 mg/mL, their CA extract
enhanced 3-fold collagen production, compared to the
control. Their data indicated higher proliferative stimu-
lation by CA than by vitamin C which was used as a pos-
itive control and enhanced collagen synthesis of 2-fold
over negative controls43.
Proliferative activity of cultured NHEK
Our findings revealed that CAEE + CNP and RA increased
proliferation activity at certain concentrations; however,
the effect is not strictly concentration-dependent. The
picture differed between 24 and 48 h of incubation.
The highest results were obtained with each of the two
compounds at 24 h: RA (5 mg/mL, 155%) >CAEE +
CNP (3.125 mg/mL, 145%). All other results including
all results at 48 h were not significantly different from
controls.
In literature, exposure of NHEK to RA increases pro-
liferative activity; RA has regulatory effects on epidermal
keratinisation, keratinocyte differentiation and inflam-
mation41. In keratinocytes, RA activates peroxisome pro-
liferator-activated receptor (PPAR) β/γ in parallel to the
activation of RA receptor (RAR); consequently, PPAR β/γ
is central to the maintenance of skin permeability, bar-
rier integrity and keratinocyte survival during inflam-
mation and wound healing. RA regulates keratinocytes
J Pharm Biomed Sci | Vol. 06 No. 05 | 315–327
325
proliferative activity and differentiation through retinoid
receptors of RAR and RXR44.
The expression of AQP3 in NHDF
Aquaporin 3 (AQP3) is the most common aquaporin
in skin. AQP3 is a membrane transporter of water and
glycerol expressed in plasma membrane in the basal
layer of human epidermis30. The function of AQP3 in
NHDF is not totally clear yet, but it may be involved in
cell migration which occurs in wound healing process45.
AQP3 expression in human skin increased as a response
to stress in diseases such as atopic eczema and to various
agents such as RA46.
The expression level of AQP3 in NHDF decreases sig-
nificantly beyond the age of 60 years. This finding might
explain why wound healing process is slower in old age
than in youngsters, suggesting that AQP3 may play an
important role in the aging of human skin47. Li et al.45
found that the expression of AQP3 protein in sun-pro-
tected human skin decreased with skin aging.
The research by Cao et al.48 showed that the expres-
sion of AQP3 in NHDF increased after treatment with
EGF in concentration- and time-dependent patterns. Ryu
et al.49 investigated the effect of AQP3 overexpression
in HPMC after incubated with TGF caused fibrosis, but
the same effect have not yet found in HDF fibroblast49
Our study demonstrates that CAEE + CNP formulation
increases the expression of AQP3 in NHDF after 24 h
of incubation in a concentration-dependent pattern. RA
also influenced the expression of AQP3 in NHDF, but the
result decreased with increasing concentrations (Fig. 8).
The expression of AQP3 in NHEK
Hydration of the SC is an important determinant of skin
appearance. Reduced SC hydration was found in aged
skin, making it looked dry, dull and saggy2. The expres-
sion of AQP3 in NHEK decreases with aging45.
In skin epidermis, as a water channel, AQP3 facili-
tates cell migration in wound healing and as a glycerol
transporter enhances keratinocyte proliferation and dif-
ferentiation. Therefore, AQP3 is also a key target protein
for anti-aging treatment30. The correlation was found
among AQP3 expression, epidermal ATP concentration
and cell proliferation as AQP3-facilitated glycerol trans-
port and lipid synthesis in the epidermis30.
ATRA increases AQP3 expression in NHEK through
RAR-γ46. ATRA of 0.05% increased AQP3 expression in
human skin explants after application for 24 h33. This
is consistent with our finding that RA increased AQP3
expression in keratinocytes at concentrations of 0.25
and 0.5 mg/mL, with optimal concentration at 1 mg/mL.
A study by Song et al. showed the AQP3 overexpression
induced by ATRA, in which time and dose dependent
will be followed by increased water transport, keratino-
cyte, skin permeability and TEWL, and caused dry skin.
Nicotinamide attenuates AQP3 overexpression induced
 326 Linda Yulianti
by RA through inhibitor of EGFR in cultured human
skin keratinocyte50.
AQP3 is the abundant skin aqua-glycero-porin
functioning as a channel for both water and glycerol.
Thus, AQP3 has an important role in skin hydration51.
The recent discovery of AQP3 functions has led to sev-
eral studies of other novel compounds51, motivated by
the relation between AQP3 expression, water content
and moisturising effect on skin. Several cosmetic com-
panies have marketed cosmetics-containing ingredients
that are claimed to increase AQP3 expression in skin.
The increase of AQP3 expression is claimed to have anti-­
aging, anti-wrinkle and other beneficial effects on skin
appearance31.
Currently, research is under way to investigate com-
pounds with the ability to stimulate AQP3 expression,
thereby increasing skin hydration endogenously. Recent
studies of herbal AQP3 stimulants included P. colubrina9,
green Coffea arabica L.52 and extract of Ajugaturga turkesnita
(a plant from Central Asia)51. The increase of AQP3 expres-
sion in NHEK was claimed to have moisturising effect
on skin31. Our research indicates that after 24 h of expo-
sure, CAEE + CNP can induce AQP3 expression in kera-
tinocytes at concentrations of 3.125 and 6.25 mg/mL.
Our study demonstrates that CAEE + CNP increased
the expression of AQP3 in NHEK ~1.6-fold, com-
pared to negative control at an optimal concentration
of 3.125 mg/mL after incubation for 24 h. This result
suggests that CAEE + CNP increases skin hydration and
that CAEE + CNP can be used as an active compound for
cosmetic skin care/moisturiser.
SUMMARY AND CONCLUSION
CAEE + CNP can be used as an effective formulation to
stimulate the proliferation of fibroblasts and keratino-
cytes during short 24 h exposure time. The best effect
of CAEE + CNP for the proliferation of fibroblasts and
keratinocytes was also obtained at low concentrations
(6.25 and 3.125 mg/mL, respectively).
Efficacy of CAEE + CNP on stimulating type I and
III collagen synthesis in fibroblasts was not signifi-
cantly different from RA, after exposure to all concen-
trations and time periods tested. CAEE + CNP increase
AQP3 expression in fibroblasts and keratinocytes after
24 h of exposure at optimal concentrations of 12.5 and
3.125 mg/mL, respectively.
Thus, CAEE + CNP can be used as an active formu-
lation for novel anti-aging cosmetic that regulates skin
hydration by increasing skin cell proliferation and AQP3
expression. In the future, pre-clinical and clinical studies
are needed for further validation.
REFERENCES
  1. Guillou MJ, Bordes S, Soulie C, Closs B. Stimulation of dermal
and epidermal metabolism: an approach to antiaging. Cosmet
Toiletries. 2005;120(6):97–102.
  2. Baumann L. Skin ageing and its treatment. J Pathol. 2007;211:
241–251.
 3. Yaar M. Clinical and histological features of intrinsic versus
extrinsic skin aging. In: Gilchrest B, Krutmann J, (eds): Skin aging.
Berlin: Springer. 2006. pp. 9–21.
  4. Zouboulis CC, Makrantonaki E. Clinical aspects and molecular
diagnostics of skin aging. Clin Dermatol. 2011;29(1):3–14.
  5. Rittie L, Fisher GJ, Voorhess JJ. Retinoid therapy for photoaging.
In: Gilchrest B, Krutmann J, (eds): Skin aging. Berlin: Springer.
2006. pp. 143–156.
  6. Rawlin AV. Retinoids and retinoic acid treatment of skin aging.
In: Rhein LD, Fluhr JW, (eds): Aging skin: current and future ther-
apeutic strategies. USA: Allured Books. 2010. pp. 219–234.
  7. Menta RC, Fitzpatrick RE. Growth factors and aging skin treat-
ment. In: Rhein LD, Fluhr JW, (eds): Aging skin: current and future
therapeutic strategies. USA: Allured Books. 2010. pp. 495–512.
  8. Draelos ZD. New delivery systems for novel compounds. In:
Sadrick NS, Lup M, Berso DS, Draelus ZD, (eds): Cosmet Sci.
2008. pp. 32–39.
  9. Pereda MC, Dieamant GC, Eberlinc S, Werkac RM, Colombic D,
Queiroz ML, et al. Expression of differential genes involved in
the maintenance of water balance in human skin by Piptadenia
colubrina extract. J Cosmet Dermatol. 2010;9(1):35–43.
10. Sirisa-ard P, Charumanee S, Rahuruk N, Inchai N, Pholsongkram
K. Development of Centella asiatica (Linn.) urban silicone
transdermal patch for wound healing. J Metals Mater Miner.
2010;20(3):169–171.
11. Bhavna D, Jyoti K. Centella asiatica: the elixir of life. Int J Res
Ayurveda Pharm. 2011;2(2):431–438.
12. Gohil KJ, Patel JA, Gajjar AK. Pharmacological review on
Centella asiatica: a potential herbal cure-all. Indian J Pharm Sci.
2010;72(5):546–556.
13. Zainol NA, Voo SC, Sarmidi MR, Aziz RA. Profiling of Centella asiat-
ica (L.) urban extract. Malaysian J Anal Sci. 2008;12(2):322–327.
14. Chanchal D, Swarnlata S. Novel approaches in herbal cosmetics.
J Cosmet Dermatol. 2008;7(2):89–95.
15. Pittella F, Dutra RC, Junior DD, Lopes MTP, Barbosa NR. Antioxidant
and cytotoxic activities of Centella asiatica (L) Urb. Int J Mol Sci.
2009;10(9):3713–3721.
16. Paocharoen V. The efficacy and side effect of oral Centella asi-
atica extract for wound healing promotion in diabetic wound
patients. J Med Assoc Thail. 2010;93(7):166–170.
17. Okonogi S, Sirithunyalug J, Chen Y. Nano-encapsulation of Centella
asiatica, bioactive extract. Paper presented at the XVIth International
conference on bioencapsulation, Dublin, Ireland. 2008.
18. Morganti P, Fabrizi G, Palombo P, Ruocco E, Morganti G, Cardilo
A. Chitin-nanofibrils: a new active cosmetic carrier. J Appl Cos­
metol. 2008;26(3):113–128.
19. Morganti P, Carezzi F, Ciotto PD. Chitin nanoparticles as innova-
tive delivery system. Person Care. 2012;5(2):55–58.
20. Morganti P, Morganti G, Muzzarelli RAA, Muzzarelli C. Chitin nano-
fibrils: a natural compound for innovative cosmeceuticals. In:
Anthony J, O’Lenick JR, (eds): Natural and organics in cosmetics:
trends and technology. USA: Allured Books. 2010. pp. 383–393.
21. Cao Z, Mo X, Zhang K, Fan Y, Lin A, He C, et al. Fabrication of
chitosan or silk fibro in composite nanofibers for wound-chessing
applications. Int J Mol Sci. 2010;11(9):3529–3539.
22. Howling GI, Dettmar PW, Goddard PW, Hampson FC, Dornish M,
Wood EJ. The effect of chitin and chitosan on the proliferation
of human skin fibroblast and keratinocytes in vitro. Biomaterials.
2001;22(22):2959–2966.
23. Leonida M. Nano-sizing chitosan for wound healing, anti-aging.
Cosmet Toiletries. 2012;127(7):532.
J Pharm Biomed Sci | Vol. 06 No. 05 | 315–327
 Effects of CA ethanolic extract in chitosan nanoparticles
24. Nasir A. Nanotechnology and dermatology: part 1: potential of
nanotechnology. Clin Dermatol. 2010;28(4):458–466.
25. Moddaresi M, Tamburic S, Williams S, Jones SA, Zhao Y, Brown
MB. Effects of lipid nanocarriers on the performance of topical
vehicles in vivo. J Cosmet Dermatol. 2009;8(2):136–143.
26. Wiesenthal A, Hunter L, Wang S, Wickliffe J, Wilkerson M.
Nanoparticles: small and mighty. Int J Dermatol. 2011;50(3):
247–254.
27. Tiyaboonchai W. Chitosan nanoparticles: a promising system for
drug delivery. Naresuan Univ J. 2003;11(3):51–66.
28. Wang X, Chi N, Tang X. Preparation of estradiol chitosan nanopar-
ticles for improving nasal absorption and brain targeting. Eur J
Pharm Biopharm. 2008;70(3):735–740.
29. Draelos Z. Aquaporins: an introduction to a key factor in the mech-
anism of skin hydration. J Clin Aesthet Dermatol. 2012;5(7):53–56.
30. Hara-Chikuma M, Verkman AS. Roles of aquaporin-3 in the epi-
dermis. J Invest Dermatol. 2008;128(9):2145–2151.
31. Verkman AS. A cautionary note on cosmetics containing ingre-
dients that increase aquaporin 3 expression. Exp Dermatol.
2008;17(10):871–872.
32. Jurzak M, Latocha M, Goiniczek K, Kapral M, Garncarczyk A,
Pierzchala E. Influence of retinoids on skin fibroblasts metabo-
lism in vitro. Acta Pol Pharm. 2008;65(1):85–91.
33. Bellemère G, Von Stetten O, Oddos T. Retinoic acid increases
aquaporin 3 expression in normal human skin. J Invest Dermatol.
2008;128(3):542–548.
34. Fan W, Yan W, Xu Z, Ni H. Formation mechanism of monodisperse,
low molecular weight chitosan nanoparticles by ionic gelation
technique. Colloids Surf B Biointerfaces. 2012;90:21–27.
35. Varani J, Shayevitz J, Perry D, Mitra RS, Nickoloff BJ, Voorhees JJ.
Retinoic acid stimulation of human dermal fibroblast prolifera-
tion is dependent on suboptimal extracellular Ca2+ concentra-
tion. Am J Pathol. 1990;136(6):1275–1281.
36. Hamid AA, Shah ZM, Muse R, Mohamed S. Characterization of
antioxidative activity of various extracts of Centella asiatica (L)
urban. Food Chem. 2002;77(4):465–469.
37. Wu F, Bian D, Xia Y, Gong Z, Tan Q, Chen J, et al. Identification of
major active ingredients responsible for burn wound healing of
Centella asiatica herbs. Evid Based Complementary Altern Med.
2012;2012:13.
38. Song J, Dai Y, Bian D, Zhang H, Xu X, Xia Y, et al. Madecassoside
induces apoptosis of keloid fibroblasts via a mitochondrial-­
dependent pathway. Drug Dev Res. 2011;72(4):315–22.
39. Lu L, Ying K, Wei S, Fang Y, Liu Y, Lin H, et al. Asiaticoside induc-
tion for cell-cycle progression, proliferation and collagen
J Pharm Biomed Sci | Vol. 06 No. 05 | 315–327
327
synthesis in human dermal fibroblasts. Int J Dermatol. 2004;
43(11):801–807.
40. Lee JH, Kim HL, Lee MH, You KE, Kwon BJ, Seo HJ, et al.
Asiaticoside enhances normal human skin cell migration,
attachment, and growth in vitro wound healing model.
Phytomedicine. 2012;19(13):1223–1227.
41. Daly TJ, Weston WL. Retinoic effect on fibroblast proliferation
and collagen synthesis in vitro and on fibrotic disease in vivo.
J Am Acad Dermatol. 1986;4(2):900–902.
42. Shigematsu T, Tajima S. Modulation of collagen synthesis
and cell proliferation by retinoids in human skin fibroblasts.
J Dermatol Sci. 1995;9(2):142–145.
43. Hashim P, Sidek H, Helan MHM, Sabery A, Palanisamy UD, Ilham
M. Triterpene composition and bioactivities of Centella asiatica.
Molecules. 2011;16(2):1310–1222.
44. Jean J, Soucy J, Pouliot R. Effects of retinoic acid on keratinocyte
proliferation and differentiation in a psoriatic skin model. Tissue
Eng Part A. 2011;17(13–14):1859–1868.
45. Li J, Tang H, Hu X, Chen M, Xie H. Aquaporin 3 gene and protein
expression in sun-protected human skin decreases with skin
ageing. Aus J Dermatol. 2010;51(2):106–112.
46. Hara-Chikuma M, Verkman AS. Aquaporin 3 facilitates epider-
mal cell migration and proliferation during wound healing.
J Mol Med. 2007;86(2):221–231.
47. Verkman AS. Aquaporins at a glance. J Cell Sci. 2011;124(13):
2107–2112.
48. Cao C, Sun Y, Healey S, Bi Z, Hu G, Wan S, et al. EGFR-mediated
expression of aquaporin 3 is involved in human skin fibroblast
migration. Biochem J. 2006;400:225–234.
49. Ryu HM, Oh EJ, Park SH, Kim CD, Choi JY, Cho JH, et al. Aquaporin
3 expression is up-regulated by TGF-beta1 in rat peritoneal
mesothelial cells and plays a role in wound healing. Am J Pathol.
2012;181(6):2047–57.
50. Song X, Xu A, Pan W, Wallin B, Kivlin R, Lu S, et al. Nicotinami­
deattenuates aquaporin 3 overexpression induced by retinoic
acid through inhibition of EGFR/ERK in cultured human skin
keratinocytes. Int J Mol Med. 2008;22(2):229–36.
51. Dumas M, Sadick NS, Nobless E, Juan M, Lachmann-Weber N,
Boury-Jamot M, et al. Hydrating skin by stimulating biosynthesis
of aquaporins. J Drugs Dermatol. 2007;6(6):20–24.
52. Velazquez PMC, Dieamant GC, Eberlin S, Noqueira C, Colombi
D, Stasi LCD, et al. Effect of green Coffea arabica L. seed oil on
extracellular matrix components and water-channel expression
in in vitro and ex vivo human skin models. J Cosmet Dermatol.
2009;8(1):56–62.