Enhancement of wound healing by curcumin in animals

GURMEL S. SIDHU, PHD-, ANOOP K. SINGH, PHD,°b; DEEPA THALOOR, PHD °,b; KRISHNA K. BANAUDHA, PHDC;
GYANENDRA K. PATNAIK, PHD ; RIKHAB C. SRIMAL, PHD , RADHA K. MAHESHWARI, PHDa

Tissue repair and wound healing are complex processes that involve inflammation, granulation, and remodeling of the
tissue . In this study, we evaluated the in vivo effects of curcumin (dfeurloylmethane), a natural product obtained from
the rhizomes of Curcuma longo on wound healing in rats and guinea pigs . We observed faster wound closure of punch
wounds in curcumin-treated animals in comparison with untreated controls . Biopsies of the wound showed reepithelial-
ization of the epidermis and increased migration of various cells including myofibroblasts, fibroblasts, and macrophages
in the wound bed . Multiple areas within the dermis showed extensive neovascularization, and Masson's Trichrome stain-
ing showed greater collagen deposition in curcumin-treated wounds . Immunohistochemical localization of transform-
ing growth factor-R1 showed an increase in curcumin-treated wounds as compared with untreated wounds. In situ
hybridization and polymerase chain reaction analysis also showed an increase in the mRNA transcripts of transforming
growth factor-(31 and fibronectin in curcumin-treated wounds . Because transforming growth factor-R1 is known to en-
hance wound healing, it may be possible that transforming growth factor-(31 plays an important role in the enhance-
ment of wound healing by curcumin. (WOUND REP REG 1998;6:167-177)

Tissue repair and wound healing are complex pro-

DEPC Diethylpyrocarbonate
cesses that involve clotting, fibrin-fibronectin (FN) ECM Extracellular matrix
deposition, inflammation, fibroplasia, neovasculariza- FN Fibronectin
tion, wound contraction, and reepithelialization . 1,2 In- GAPDH Glyceraldehyde 3-phosphate dehydrogenase
teractions of different cell types, extracellular matrix LMN Laminin
(ECM) proteins, and their receptors are involved dur- PBS Phosphate buffered saline solution
PCR Polymerase chain reaction
ing these biological significant processes, which are
RT-PCR Reverse transcription-PCR
mediated by cytokines and growth factors.' ECM pro- SM Smooth muscle
TGF-(31 Transforming growth factor-(31
From the Center for Combat Casualty and Life Sustain-
ment Research,' Department of Pathology, Uni-
formed Services University of the Health Sciences, teins such as fibronectin, laminin (LMN), tenascin,
Bethesda, MD,, and the Birla Institute of Technol- and collagen are not only important for structural
ogy and Science b Pilani and Central Drug Re- remodeling of the tissue but also because they are
search Institute, Lucknow, India. involved in the early phase of the wound healing.
The opinions or assertions contained herein are the pri- ECM protein interactions support the adhesion of fi-
vate views of the authors and should not be con-
strued as official or necessarily reflecting the broblasts and endothelial cells.',' Several growth fac-
views of the Uniformed Services University of the tors/cytokines alone or in combination play an
Health Sciences or the Department of Defense. important role during tissue repair and enhance the
Reprint requests: Rodho K. Moheshworl, PhD, Department normal healing process in animals and humans."
of Pathology, Uniformed Services University of the The release of various growth factors into the wound,
Health Sciences 4301 Jones Bridge Road, Be- such as platelet-derived growth factor, the epidermal
thesdo, MD 20814.
Copyright © 1998 by The Wound Healing Society. growth factor family, basic fibroblast growth factors,
1067-1927 $5.00 +0 transforming growth factor R-1 (TGF-R1), and trans-
167

168 SIDHU ET AL .
forming growth factor-a, affect recruitment, activa-
tion, migration, and differentiation of various cell
types in the wound bed.' °,11 The induction of growth
factor(s) in the wound would provide evidence for the
consideration of a compound as a vulnerary agent.
The present study was undertaken to evaluate
the in vivo effect of curcumin (difeurloylmethane) on
wound healing in rats and guinea pigs . Curcumin, a
major component of the food flavor turmeric, has been
isolated from the rhizomes of Curcuma longa and has
been widely used for centuries in indigenous medicine
for the treatment of a variety of inflammatory condi-
12
tions and arthritis. The other attractive feature of
curcumin to explore as a vulnerary agent is that de-
spite being eaten daily for centuries in Asian coun-
tries, curcumin has not been reported to be toxic. 12
Various parameters were used to assess healing
in this study, including wound closure by contraction,
morphology, and presence of ECM proteins and TGF
(31 in the wound region . Our results suggest that an
increased biosynthesis of ECM and TGF-(3 in the
wound may play an important role in the enhance-
ment of wound healing by curcumin in animals.
MATERIALS AND METHODS
Male Swiss albino strain rats (250-300 gm) were pur-
chased from Charles River laboratory. The animals
were used in compliance with the U.S. Public Health
Service policy on humane care and use of animals.
Adult guinea pigs were obtained from National
Laboratory Animal Center, Central Drug Research
Institute, Lucknow, India. Animals were housed in a
room with 12-hour light and 12-hour dark cycle and
were fed Purina Fox Chow and water ad libitum.
Guinea pig experiments were performed at the Cen-
tral Drug Research Institute, and tissues were
brought to the Uniformed Services University of
Health Sciences for histology, histochemistry, and in
situ hybridization.
Animal protocol
Animals were anesthetized using pentobarbital, as
previously described. 13 Hair on the dorsal side of each
animal was shaved, and the skin was cleansed with
70% ethanol. An 8-mm skin biopsy punch (Acuderm
Inc., Ft . Lauderdale, Fla.) was used to create full-
thickness dorsal cutaneous wounds under aseptic con-
ditions. Four wounds, with two on each side, were
created, and thereafter, the animals were individually
caged.
WOUND REPAIR AND REGENERATION
MARCH-APRIL 1998
Initially, various doses of curcumin (10, 20, 40,
60, 80, and 100 mg/kg; Central Drug Research Insti-
tute) were administered either orally or applied top
ically daily or once at the time of wounding or
alternate days in guinea pigs to evaluate its wound-
healing capacity. The dose-response studies showed
that curcumin was most effective at a concentration
from 30-50 mg/kg when administered orally daily for
wound healing. Therefore, we selected a dose of 40 mg/
kg for these studies. Curcumin was dissolved in water
and given orally daily for 7-11 days to the guinea
pigs, and water was given orally to the vehicle control
animals.
To investigate whether systemic administration
of curcumin would enhance wound healing, we se-
lected the intraperitoneal route. The pharmacokinet
ics of curcumin in animals have been previously
reported. 14-16 Because intraperitoneal administration
of curcumin of 40 mg/kg yields higher blood levels
than oral administration at 40 mg/kg, we treated the
animals with the same dose to find out whether higher
blood levels of curcumin would heal the wound faster
than the same dose given orally. Curcumin (40 mg/
kg intraperitoneal) was injected 18 hours prior to the
creation of wounds in rats, on the day of wounding,
and daily for 6 days . Phosphate-buffered saline solu-
tion (PBS) was injected in the vehicle control animals.
The experiment in rats was terminated early because
of faster natural healing as compared with guinea
pigs .
62uanfitation of wound area
On the first day of photography, the camera (Nikon
8008S) was locked and fixed on a stand after adjusting
the focus on the wound. Every time thereafter, the
animals were placed under the same fixed camera
and photographed to eliminate an error in the size
of the wounds caused by photography. A Macintosh
11 cx computer, equipped with MacImage 2 .3 (copy-
right 1986-90), was used to scan photographic images
of the wounds . Images were analyzed, and the area
was measured using NIH image 1 .30u program .
Morphological studies
Six rats from each group were euthanized on day 3
and 6, guinea pigs from each group on day 7 and 11 of
wounding using pentobarbital anesthesia . Wound
tissues were excised and cut into halves ; one half was
placed in formalin (10% formaldehyde in PBS) for
histology, hematoxylin and eosin staining, imm-
unohistochemistry, and in situ hybridization, and the
second half was quickly washed in cold sterile
WOUND REPAIR AND REGENERATION
VOL 6, NO. 2
diethylpyrocarbonate (DEPC)-treated PBS for RNA
preparation.
Masson's Trichrome method for staining collagen
was performed on paraffin sections, as described, 13
and photomicrography was performed. Collagen is
stained blue with Masson's Trichrome stain.
Immunohistochemical staining for TGF-P 1 and a
actin
Immunostaining was performed using a monoclonal
anti-human TGF-P antibody (Genzyme, Cambridge,
Mass .) by an indirect avidin-biotin-immunoperoxi-
dase technique (Vectastain ABC Elite, Vector labora-
tories, Burlingame, Calif), as specified by the
manufacturer. Briefly, sections were placed on poly-
i.-lysine-coated slides, deparaffinized, and hydrated.
Endogenous peroxidase activity was blocked with hy-
drogen peroxide in methanol for 10 minutes. Nonspe-
cific staining was blocked with normal serum, and
the sections were incubated with monoclonal anti-
human TGF-R antibody for 10-16 hours at 4° C. To
assure that the reaction of antibodies was specific,
sections from each test were incubated with normal
serum IgG separately. Slides were washed with PBS.
Biotinylated secondary antibody IgG (H + L) was
added for 1 hour, followed by avidin-biotin-peroxidase
complex for 30 minutes. Slides were stained with di-
aminobenzidine and counter stained with Harris he-
matoxylin/eosin. Slides were examined under a
Olympus light microscope equipped with a camera,
and photomicrography was performed.
Immunostaining for a actin was also performed
in curcumin-treated wounds to determine whether
myofibroblasts were present in the granulation tissue
of the wound. An anti-(x smooth muscle (SM) actin
antibody (IgG2a murine monoclonal ; Boehringer Man-
nheim Biochemicals, Indianapolis, Ind.) was used in
an indirect avidin-biotin-immunoperoxidase tech-
nique as described earlier here .
Polymerase chain reaction
On day 3 and 4 postwounding, animals were eutha-
nized, wounds were excised and washed in PBS con-
taining DEPC, and total RNA was immediately
extracted from the tissues with the using a RNAzol
B kit (Tel-Test Inc., Friendswood, Tex.), as specified
by the manufacturer . The RNA was further purified
by lithium chloride precipitation and analyzed by elec-
trophoresis in a denaturating 1.5% formaldehyde gel
to ascertain that the RNA was undegraded.
RNA was reverse transcribed, as described by Ra-
ghunath et al. 17 The cDNA was amplified by poly-
SIDHU ET AL . 169
merase chain reaction (PCR) using sense and
antisense specific primers. Briefly, to 3 pl of reverse
transcriptase (RT) mix, the following components
were added: 6p1 of dNTP mix (100 mM); 1Opl of 1OX
PCR buffer and 0 .5pl ofTaq polymerase (5 U/pb ; 0.4p1
of sense primer (1 gg/~t1) and 0 .4p1 of antisense primer
(1 pg/pl) ; and 79.7pl of distilled water. The samples
were incubated in a DNA thermocycler (Perkin Elmer
Cetus, Gene Amp PCR system 9600, Foster City, CA)
for 25 cycles. To confirm that equal amounts of RNA
were taken iri each PCR reaction in the experiment,
primers for the "housekeeping gene," glyceraldehyde
3-phosphate dehydrogenase (GAPDH), were used for
comparison . The location of primers for nucleotide
sequence amplification of amplified DNA product on
Southern blot were bases 1271-1292 (sense), 1510-
1491 (antisense), and 1401-1421 (probe) for TGF-(3
and 1657-1677 (sense), 2070-2050 (antisense), and
probe (1830-1850) for FN. GAPDH primers were to
the rat sequence, bases 388-405 (sense), 581-562 (an-
tisense), and 531-549 (probe). The sequences for the
antisense probes were selected from cDNA for TGF-
(3, FN, or GAPDH, were examined for sequence ho-
mology in the GENE BANK, and were found to be
gene specific.
The final PCR products from each sample were
run on 1% GTG agarose (FMC Corp., Rockford, Me.),
blotted to nytran membrane (Schleicher and Schuell,
Keene, N.H .), and hybridized to 32p (NEN, Boston,
Mass.), end labeled oligo-probes at 42° C for 8 hours,
and then exposed to x-ray film .
In situ hybridization
In situ hybridizations were performed as previously
described 18 with slight modifications . Briefly, the
slides were fixed in 4% paraformaldehyde acetylated
for 10 minutes and were dehydrated and hybridized
for 18 hours at 42° C. Plasmid pBR322 containing
murine cDNA was a generous gift from Dr. Helen
Coon, USUHS, Bethesda, Md. Murine TGF-(31 cDNA
insert was excised and purified prior to labeling with
[a-35 S] dCTP (ICN Biomedical Inc ., Irvine, Calif;
1000 Ci[37TBq]/mmol) with random primers. Hybrid-
ization was performed using a solution containing
O.l pg/ml "S-labeled cDNA probe, 1Omm dithiothrei-
tol, 500pg/ml acetylated bovine serum albumin, 0.3M
sodium chloride, 50% deionized formamide, 10% (w/
v) dextran sulfate, 500pg/ml tRNA, 0 .02% polyvi-
nylpyrrolidone, 0 .02% Ficoll (Sigma Chemical Co., St.
Louis, MO), 1pmM EDTA, and 50 mmol/L Tris-HCl,
pH7 .6. Posthybridization was performed by sequen-
tial incubation of the slides in a progressively more
 WOUND REPAIR AND REGENERATION
170 SIDHU ET AL . MARCH-APRIL 1998

Figure 1 Effect of curcumin treatment on wound closure in guinea pigs . A, control . B, curcumin treated .

Figure 2 Quantitation of wound contraction in response to cur-

cumin in guinea pigs . At the indicated times postwounding, an-
imals were photographed, and the wounds were quantitated
using an image analysis program . The data represent the mean
SD of 12 wounds from six animals in each group . All values in
the treated wound are significant (p < 0 .001, t test) compared
with the untreated wounds (Placebo) .

Figure 3 (Opposite page, top) Hematoxylin/eosin staining of the

wounded guinea pig tissues . A, Seven-day untreated wound
shows epithelial loss, loose crust over the area of epithelial loss,
and a focally extensive layer of neutrophils/eosinophils adja-
cent to the area of epithelial loss . B, Seven-day curcumin-
treated wound that shows migration of the epithelium over the
wound . The dermal area is diffusely infiltrated with a moderate
number of fibroblasts and macrophages . Multiple areas within
the dermis exhibit mild to moderate neovascularization . C,
Eleven-day postinjury, wound showing epithelium regeneration .
Hyperplastic and migrating epithelium adjacent to the wound
was seen, and dermal collagen fibers are loosely arranged . D,
Eleven-day curcumin-treated wound showing complete reepi-
thelialization with a 2-3 layer thick epidermis that is formed . Col-
lagen fibers are compact and well aligned (magnification
x10) .
Figure 4 Higher power magnification of the dermal region of the wound showing granulation tissue . A, Seven-day postwounding, the
untreated wound shows infiltration of cells with a loose reticular tissue . B, Seven-day curcumin-treated wound exhibits matured gran-
ulation tissue with ample fibrogenesis and neovascularization . C, Eleven-day untreated wounds show immature granulation tissue
with numerous cells, mild neovascularization, and loosely arranged collagen . D, Eleven-day curcumin-treated wound shows fully ma-
tured granulation tissue with compact and well-aligned collagen fibers (magnification x40) .
 WOUND REPAIR AND REGENERATION
172 SIDHU ET AL . MARCH-APRIL 1998

Figure 5 Hematoxylin and eosin stained sections of 3- and 7-day punch wounds in rats : A, Third-day postwounding untreated control
wound tissue with epithelial loss and adjacent focally extensive layer of neutrophilsleosinophils . B, Third day curcumin-treated wound
with formation of granulation tissue . Granulation tissue is infiltrated with a large number of cells and a moderate number of vessels . C,
Seven-day postinjury, untreated control wound with a thin layer of migrating epithelium over the dermis . Granulation tissue shows ex-
tensive neovascularization and infiltration of a large number of cells . D, Seven-day curcumin-treated wound with complete reepithe-
lialization of the wound with multiple layers of thick epidermis . Connective tissue is formed, and collagen fibers are arranged,
indicating complete healing of the wound (magnification x10) .

dilute saline-citrate solution containing 5 mM dithio-
threitol and 0.1% triton-X, dehydrated, and dipped
in NTB-3 emulsion for autoradiography. After expo-
sure for 7-10 days at 4° C, the slides were developed
with D-19 developer, stained with eosin, and mounted
with permount .The cells were examined and photo-
graphed with a Zeiss microscope equipped with cam-
era attachments.
Masson's Trichrome and immunohistochemical
staining and in situ hybridization were performed
on 8-10 wounds with 10-12 serial sections of
each wound to assess the reproducibility of the
wound observations .
compared with control. Wounds in the treated or
untreated guinea pigs were quantitated from
photographs at 5, 7, 8, 9, and 11 days postwounding .
The area of the curcumin-treated wounds was
significantly decreased as compared with control
(Figure 2) untreated wound . It is apparent from the
figure that in untreated animals, the wound area at
9-days postwounding is the same as that of 7-day
curcumin-treated animals, thereby indicating a 2-day
enhancement in wound closure. The results were
consistently reproducible and statistically significant
when compared with control wounds . Curcumin
treatment was also effective in reducing closure time
in rats by approximately 35%.

RESULTS Morphological changes in curcumin treated

Visual observation of 7-day-old punch wounds in wounds
guinea pigs (Figure 1) showed that curcumin Multiple cross-sections of hematoxylin and eosin
treatment resulted in a greater closure of wounds as stained sections of curcumin-treated and untreated
WOUND REPAIR AND REGENERATION
VOL 6, NO . 2
wound tissues of rats and guinea pigs were examined
for epithelial regeneration, formation of granulation
tissue, and fibrogenesis . In the untreated wounds,
epithelium was not intact, and the epithelial cells
adjacent to the area ofloss did not exhibit hyperplasia,
in the guinea pig model day 7 postwounding or in the
rat model on the day 4 postinjury. The wound itself
showed a number of inflammatory cells, such as neu-
trophils and lymphocytes, and scattered fibroblasts
at this time (Figures 3, A, and 5, A) . In contrast, the
curcumin-treated wounds showed organization and
initiation of fibrogenesis and epithelial regeneration
at these times (Figures 3, B, and 5, B) . Neovascular-
ization was also more extensive in treated wounds
(Figures 4, B, and 5, B) as compared with untreated
wounds (Figures 4, A, and 5, A) . In the curcumin-
treated wounds, there was complete reepithelializa-
tion with 2-3 layers of the epidermis (Figures 3, D,
and 5, D), whereas the untreated wounds were at an
earlier stage of epithelial differentiation (Figures 3,
C and 5, C) at these times. At day 11, whereas the
untreated wounds showed loose reticular collagen,
congested vessels, and persisting inflammatory infil-
trate (Figure 4, C), the curcumin-treated wounds dem-
onstrated compact and well-aligned collagen, showing
that the wound was essentially healed (Figure 4, D).
The extent of collagen deposition in the wounds
was examined by Masson's Trichrome staining. Cur-
cumin treatment resulted in greater collagen content
in the wounds as compared with untreated wounds
(Figure 6, A and B) . Untreated wounds had a loose
reticular arrangement of collagen, whereas collagen
was compact and well aligned in the treated wounds
(Figure 6 C, and D) .
Immunohistochemical localization of TGF-(31
TGF-(31 was localized in curcumin-treated and
untreated wounds on day 3 in which only a faint
staining of TGF-R was seen in untreated wounds of
rats and guinea pigs (Figure 7, A) . TGF-(31, however,
was significantly increased and found predominantly
in the epidermis and granulation tissue of curcumin-
treated wounds (Figure 7, B) . Migrating epithelium
over the lost tissue area and hyperplastic epithelium
adjacent to the migrating epithelium showed
maximum staining for TGF-(31 in both animals (Figure
7 C, and D) . Curcumin-treated wounds consistently
showed a greater influx of fibroblasts, which were
positive for TGF-(31 staining compared with untreated
wounds (Figure 7) . To determine if the infiltrating cells
were myofibroblasts, sections were stained with SM a
actin. As is apparent from Figure 8, many of the
infiltrating cells in the dermis were positive for SM a-
SIDHU ET AL . 173
actin staining, indicating that myofibroblasts were
abundant in the granulation tissue of the wound of
curcumin-treated animals.
Determination ofFN and TGF-(31 mRNA
Immunohistochemical localization showed increased
levels of TGF-(31 protein in the wounded tissues of
curcumin-treated animals. To determine whether cur-
cumin treatment was enhancing the biosynthesis of
TGF-(31, we ascertained the relative abundance of
mRNA transcripts using in situ hybridization and
RT-PCR . The amplification of TGF-(31 gene by RT-
PCR and its hybridization with TGF-(31 oligonucle-
otide probe showed an increase in the transcript abun-
dance of TGF-(31 in curcumin-treated rat wounds (3-
days postwounding) compared with untreated wounds
(Figure 9, A) . An increase in FN mRNA transcripts
was also seen in curcumin-treated rat wounds com-
pared with untreated wounds at the same time (Fig-
ure 9, B) . The RTPCR products of GAPDH indicated
that an equal amount of RNA was taken for the am-
plification of TGF-(31 and FN.
Curcumin treatment did not appear to have
any effect on the housekeeping gene (GAPDH) as
compared with control . To monitor, the reproduc
ibility of mRNA transcripts by PCR, RTPCR
amplifications were performed on RNA separately
prepared from 16 wounds from four rats . The
PCR amplification and hybridization of the blot
with the probe were repeated by more than one
individual in the laboratory. The results obtained
from the wounds of the same rat as well as
different rats in normal and curcumin-treated
groups were reproducible .
We then examined the expression of TGF-(31 in
untreated and curcumin-treated wounds by in situ
hybridization . The procedure was performed on 8-10
wounds with 10-12 serial sections of each . The level
of grains in the vector control wounded tissue was
low, indicating the specificity of the probe (Figure 10,
A) . A few grains, however, were observed in wounded
skin 3-days postinjury when the tissue was hybridized
with the TGF-(31 cl)NA, indicating the presence of
TGF-P1 (Figure 10, B) in normal wound healing. Pres-
ence of a large number of grains in curcumin-treated
wounds indicated an increase in the presence of TGF-
(31 mRNA compared with untreated wounds (Figure
10, C) .
DISCUSSION
Tissue repair and wound healing are complex pro-
cesses that involve inflammation, fibroplasia, neovas-
Figure 6 Masson's Trichrome staining for collagen : Three-day postwounding untreated (A) and curcumin-treated (B) tissues sections .
Six-day untreated (C) and curcumin-treated (D) rat wounds (magnification 40x) .
WOUND REPAIR AND REGENERATION
VOL 6, NO . 2 SIDHU ET AL . %75

Figure 8 SM a-actin staining in rat wound . Three-day untreated wound (A) and curcumin-treated wounds (B) . Numerous infiltrating
cells in the wound bed of curcumin-treated wounds show positive staining for SM a actin, indicating that myofibroblasts are abun-
dant in the granulation tissue . Photomicrographs (magnification 100x, oil immersion) were taken of the granulation tissue near the
epidermal/dermal junction .

Figure 10 In situ hybridization of TGF-ßl mRNA in rat wounds . A, A day 3 wound section hybridized with plasmid vector pBR322 . B, Day
3 untreated wound and day 3 curcumin-treated wounds (C) . Black grains indicate mRNA hybridization for TGF-ß . Lack of grains on
hybridization with a plasmid from which the TGF-ß cDNA was excised indicates the degree of specificity of the probe . Numerous cells
are moderately labeled with silver grains in the untreated wound, whereas the administration of curcumin stimulated both the num-
ber and intensity of positive cells . Photomicrographs (magnification 20x) were taken of the granulation tissue near the dermal and
epidermal junction .

<-- Figure 7 Immunohistochemical localization of TGF-ßl . Three-day postwounding untreated (A) and curcumin-treated (B) tissues
sections . Six-day untreated (C) and curcumin-treated (D) wounds in rats . Brown indicates the staining for TGF-ßl (magnification 40x) .

176 SIDHU ET AL .
cularization, wound contraction, and resurfacing of
the wound defect with epithelium . l9 To understand
the mechanism of wound repair, various models such
as polyvinyl sponges, dead space chambers, punch
models, linear incision models, and full-thickness
punch biopsy models have been used in the past . Each
model has several advantages and disadvantages over
the other. We used full-thickness punch wounds to
study the effect of curcumin on wound healing. This
model enabled us to study wound closure caused by
contraction. A faster wound closure was observed in
curcumin-treated wounds as compared with un-
treated wounds, suggesting that curcumin enhanced
wound healing. The contraction of the wound corre-
lated with morphologic analysis of the tissues. Cur-
cumin-treated wound tissues showed a large number
of infiltrating cells, such as macrophages, neutrophils,
and fibroblasts, as compared with untreated wounds .
During granulation tissue formation, as contraction
proceeds and resistance increases, fibroblasts differ-
entiate into myofibroblasts . 2o,21 The presence of my-
ofibroblasts is considered to be characteristic of tissue
undergoing contraction. 22,23 The apparent greater
number of myofibroblasts in curcumin-treated
wounds may be partially responsible for the fast
wound contraction.
The process of postinflammatory wound healing
and tissue regeneration is characterized by a complex,
multicomponent cascade of degradative and biosyn
thetic reactions that direct underlying cell-cell and
cell-ECM interactions .2 ECM proteins are well known
to be important in many phases of wound healing.
WOUND REPAIR AND REGENERATION
MARCH-APRIL 7998
Figure 9 PCR gels of FN and TGF-P1 ampli-
fled mRNA in rat wounds . Left panel, TGF-
ß : Lane l, day 3 untreated; Lane 2, day 3
curcumin-treated ; Right panel, FN : Lane
l, day 3 untreated ; Lane 2, day 3 treated
wounds . GAPDH was used as an internal
control .
FN helps in the phagocytosis of wound debris and
promotes the migration of wound fibroblasts, endot-
helial cells, and epithelial cells over the wound. Col-
lagen type III is increased in the early granulation
tissue and is later replaced by collagen type 1.24 These
two collagens impart tensile strength to the healing
wound. 25 Collagen and FN are also chemotactic for
fibroblasts .26 Migrating and proliferating epidermis
induces the production of LMN, which has been shown
to be important in the remodeling of tissues.26 En-
hanced deposition of collagen, LMN, and FN proteins
by curcumin in wounds suggests that curcumin indi-
rectly helps in the remodeling of tissues, as well as
providing strength to the wounded tissues.
Migration of cells represents a potential source
of growth factors required for the regulation of bio-
logical processes during wound healing. TGF-ßl is
important in wound healing, as it stimulates the ex-
pression of FN and collagen by fibroblasts and in-
creases the rate of formation of granulation tissue in
vivo . 7,27,28 Increased levels of TGF-P1 protein were
observed, along with the increased steady state level
of TGF-(31 mRNA in curcumin-treated wounds com-
pared with untreated wounds . The increased levels
of TGF-P1 may be due to both its release by the mi-
grating cells as well as an increase in the transcription
of TGF-P1 mRNA at the wound site . Induction of
TGF-P1 occurs upon exposure to proteases, the acidic
pH of the wound, and upon activation of macrophages,
which then leads to the release of various growth
factors that are essential for wound repair. Therefore,
the endogenous production of TGF-P1 by curcumin
WOUND REPAIR AND REGENERATION
VOL 6, NO. 2
during the initial stages of wound healing could lead
to increased granulation tissue deposition . TGF-P1
stimulates keratinocyte production of FN and col-
lagen and its deposition in ECM.29 TGF-(31 also stim-
ulates epidermal migration and mediates epidermal
regrowth . Thus, TGF-(31 may be responsible for the
faster reepithelialization observed during wound
healing in curcumin-treated wounds .
It has previously been shown that the sup-
plementation of TGF-R accelerates the wound healing
in Adriamycin-treated animals, which shows reduced
endogenous production of TGF-R expression levels at
wound sites. 30 Therefore, curcumin may be useful in
conditions of impaired wound healing, as it appears to
induce endogenous production of TGF-(31 in the
wound. Whether curcumin could induce TGF-(31 in
conditions of impaired wound healing remains to be
elucidated .
ACKNOWLEDGMENTS
This work was supported by Grant G174FQ and
G174DE from the Naval Medical Research and De-
velopment Command. The authors thank Major Ste-
iffle of the Armed Forces Radiobiology Research
Institute, Bethesda, for histopathology examination
of wound tissue, and Dr RN. Raghunath and Mr J.W
Sklarsh for their help with in situ hybridization and
animal work.
REFERENCES
1. Martin P, Hopkinson-Wooley J, McClusky J. Growth factors and
cutaneouswoundrepair.Prog Growth Factor Res 1992 ;4 :24-44 .
2. Hunt TK . A retrospective perspective on the nature ofwoundin
growth factors and other aspects of wound healing. In : Barbul
A, Pines E, Caldwell M, Hunt TK, editors. Biological and clin-
ical implications . Philadelphia: Alan R Liss, 1988 :XIII-XX .
3. McDonald JA. Receptors for extracellular matrix components .
Am J Physiol 1989 ;257 :L331-7.
4. Mackie EJ, Halfter W, Liverani D. Induction of tenascin in heal-
ing wounds . J Cell Biol 1988 ;107 :2757-67 .
5. Fujikawa LS, Foster CS, HavristTJ, LaniganJM, Cavin RB. Fi-
bronectin in healing rabbit corneal wounds . Lab Invest
1981 ;45:120-6 .
6. Brown RL, Breeden MP, Greenhalgh DG. PDGF and TGF-a act
synergistically to improve wound healing in genetically dia-
betic mouse. J Surg Res 1994 ;56:562-7 .
7. Mustoe TA, Pierce GF, Thomason A, Gramates P, Sporn MB,
Deuel TF. Accelerated healing of incisional wounds in rats in-
ducedby transforming growth factor-(3. Science 1987 ;237 :1333-
5.
8. Kiristy CP, Antoniades HN, Carlson MR, Beaulieu MT, D'An-
drea M, Lynch SE . Combination of platelet derived growth fac-
tor-BB and insulin-like growth factor-I is more effective than
platelet derived growth factor-BB in stimulating complete
healingoffull thicknesswounds in "older"diabetic mice. Wound
Rep Reg 1995 ;3 :340-50.
SIDHU ET AL . 177
9. Falanga V Growth factors and wound healing. Dermatol Clin
1993 ;11:667-75.
10 . Rothe M, Falanga V Growth factors: their biology and promise
in dermatologic diseases and tissue repair. Arch Dermatol
1989 ;125 :1390-8.
11 . Mustoe TA, Pierce GF, Morishima C, Deuel TF. Growth factor
induced acceleration of tissue repair through direct and in-
ductive activities in a rabbit dermal ulcer model. J Clin Invest
1991 ;87:694-703 .
12 . Ammon HP, Wahl MA . Pharmacology ofCurcuma longa. Planta
Med 1991 ;57:1-7 .
13 . Bhartiya D, Sklarsh JW, Maheshwari RK. Enhanced wound
healing in animal models by interferon and interferon inducer.
J Cell Physiol 1992 ;150 :312-9 .
14 . Ravindranath V, Chandrasekhara N. Absorption and tissue
distribution of curcumin in rats . Toxicology 1980 ;16:259-65 .
15 . Singh S, Khanna M, Sarin JPS. High pressure liquid chro-
matographic determination of curcumin in biological fluids . In-
dian Drugs 1981 ;18:207-9 .
16 . Ravindranath V, Chandrasekhra N. Metabolism of curcumin-
studies with [3H] curcumin. Toxicology 1982 ;22:337-44.
17 . Raghunath PN, Sidhu GS, Coon HC, Liu K, Srikantan V, Ma-
heshwari RK. Interferons upregulate the expression of laminin
and its receptor LBP-32 in cultured cells. JBiol Regul Homeost
Agents 1993;7 :22-30 .
18 . Diamond SL, Sharefkin JB, Dieffenbach C, Frasier-Scott K,
McIntire LV, Eskin SG . Tissue plasminogen activator mes-
senger RNAlevels increase in cultured human endothelial cells
exposed to laminar shear stress . J Cell Physiol 1990;143 :364-
71 .
19 . Skalli O, Gabbiani G. The biology of the myofibroblast rela-
tionship to wound contraction and fibrocontractive disease. In :
Clark RAF, Henson PM, editors. The molecular and cellular bi-
ology of wound repair. New York : Plenum Press, 1988 :373-402 .
20 . Rudolph R, Berg JV, Ehrlich HP. Wound contraction and scar
contracture . In : Cohen IK, DiegelmanRF, Lindbad WJ, editors.
Wound healing: biochemical andclinical aspects. Philadelphia :
Saunders, 1992 :96-114.
21 . Petroll WM, Cavangah HD, Barry P, Andrews P, Jester JV
Quantitative analysis of stress fibre orientation during corneal
wound contraction . J Cell Sci 1993 ;104 :353-63.
22 . Darby 1, Gabbiani G. a-Smooth muscle actin is transiently ex-
pressedby myofibroblasts during experimental wound healing.
Lab Invest 1990 ;63:21-9.
23 . Desmouliere A, Geinoz A, Gabbiani F, Gabbiani G. Trans-
forming growth factor-a induces a-smooth muscle actin ex-
pression in granulation tissue myofibroblasts and in quiescent
and growing cultured fibroblasts. J Cell Biol 1993 ;122 :103-11.
24 . Kirsner RS, Eaglestein WH . The wound healing process. Der-
matol Clin 1993 ;11:629-40.
25 . Madden JW, PeacockEE . Studies on the biology of collagen dur-
ing wound healing: rate of collagen synthesis and deposition in
cutaneous wounds of the rat. Surgery 1968 ;64:288-94.
26 . Raghow R. The role of extracellular matrix in post inflam-
matory wound healing and fibrosis . FASEB J 1994 ;8 :823-31.
27 . Varga J, Rosenbloom J, Jimenez SA. Transforming growth fac-
tor R (TGF-(3) causes a persistent increase in steady-state
amounts of type 1 and type III collagen and fibronectin mRNAs
in normal dermal fibroblasts. Biochem J 1987 ;247 :597-604 .
28 . Quaglino D, Nanney LB, Kennedy R, Davidson JM . Trans-
forming growth factor-R stimulates wound healing and mod-
ulates extracellular genes expression in pig skin . Lab Invest
1990;63:307-12.
29 . McKenzie RC, Sauder DN. The role of cytokines in inflam-
mation and immunity. J Invest Dermatol 1990;95:1058-7S .
30 . Lawrence WT Norton JA, Sporn MB, Gorschboth C, Groten-
dorst GR . The reversal of adriamycin induced healing im-