RESEARCH ARTICLE

Anti-inflammatory effects of methyl ursolate obtainedfrom a chemically derived crude extract of apple peels: potentialuse in rheumatoid arthritis

Tatiana A. Padua • Bianca S. S. C. de Abreu • Thadeu E. M. M. Costa •

Marcos J. Nakamura • Lıgia M. M. Valente • Maria das Gracas Henriques •

Antonio C. Siani • Elaine C. Rosas

Received: 12 July 2013 / Accepted: 26 January 2014

� The Pharmaceutical Society of Korea 2014

Abstract Ursolic acid (UA), a pentacyclic triterpene acid

found in apple peels (Malus domestica, Borkh, Rosaceae),

has a large spectrum of pharmacological effects. However,

the vegetal matrix usually produces highly viscous and

poorly soluble extracts that hamper the isolation of this

compound. To overcome this problem, the crude EtOH–

AcOEt extract of commercial apple peels was exhaustively

treated with diazomethane, after which methyl ursolate

(MU) was purified by column chromatography and char-

acterized spectrometrically. The anti-inflammatory effects

of UA and MU (50 mg/kg) were analyzed by zymosan-

induced paw edema, pleurisy and in an experimental

arthritis model. After 4 h of treatment with UA and MU,

paw edema was reduced by 46 and 44 %, respectively.

Both UA and MU inhibited protein extravasation into the

thoracic cavity; tibio-femoral edema by 40 and 48 %,

respectively; and leukocyte influx into the synovial cavity

after 6 h by 52 and 73 %, respectively. Additionally, both

UA and MU decreased the levels of mediators related to

synovial inflammation, such as KC/CXCL-1 levels by 95

and 90 %, TNF-a levels by 76 and 71 %, and IL-1b levels

by 57 and 53 %, respectively. Both the compounds were

equally effective when assayed in different inflammatory

models, including experimental arthritis. Hence, MU may

be considered to be a useful anti-inflammatory derivative to

overcome the inherent poor solubility of UA for formu-

lating pharmaceutical products.

Keywords Ursolic acid � Apple peels � Zimosan-induced

arthritis

Introduction

The triterpene ursolic (3b-hydroxy-olea-12-en-28-oic) acid

(UA) has an important role as a chemical marker in some

medicinal plants and phytopharmaceutical products

(Takeoka et al. 2000; Baricevic et al. 2001). This com-

pound presents antioxidative and hepatoprotective activi-

ties, which have recently been comprehensively reviewed

along with other closely related triterpene acids (Liu 2005).

UA is also considered to be nutritionally important in the

prevention of some chronic diseases, such as diabetes and

cancer (He and Liu 2007; Lee et al. 2010). Furthermore,

the immunomodulatory and anti-inflammatory actions of

UA have been reported along with its ability to induce

apoptosis on diverse cell lines (Ying et al. 1991; Mitaine-

Offer et al. 2002; Ikeda et al. 2008; Sultana and Saify

2012). In experimental animal models, UA inhibits cell

migration, vascular permeability, paw swelling, and the

expression of pro-inflammatory cytokines, such as INF-cand TNF-a (Chattopadhyay et al. 2002; Ahmad et al.

2006). In addition, UA inhibits paw swelling, plasma PGE2

levels, and radiological changes in the joints caused by

complete Freund’s adjuvant-induced arthritis in rats (Kang

et al. 2008).

UA is widely distributed in the plant kingdom and is

found in the apple (Malus domestica Borkh., Rosaceae)

T. A. Padua � B. S. S. C. de Abreu � T. E.

M. M. Costa � M. J. Nakamura � M. G. Henriques �A. C. Siani (&) � E. C. Rosas

Department of Natural Products, Medicines and Drugs

Technology Institute (Farmanguinhos), Oswaldo Cruz

Foundation, Rua Sizenando Nabuco 100, Manguinhos,

Rio de Janeiro, RJ 21041-250, Brazil

e-mail: siani@far.fiocruz.br

L. M. M. Valente

Chemistry Institute, Federal University of Rio de Janeiro, Ilha do

Fundao, Rio de Janeiro, RJ, Brazil

123

Arch. Pharm. Res.

DOI 10.1007/s12272-014-0345-1

cuticle (Szakiel et al. 2012), wherein it constitutes

30–35 % of the total lipophilic content. In this specific fruit

matrix, UA is normally accompanied by approximately

10 % of its isomer, oleanolic acid (Bringe et al. 2006).

Nevertheless, the isolation of UA from apple peels is

usually hampered by the extremely poor solubility of such

a molecule in aqueous or hydro-alcoholic mixtures (Jin

et al. 1997) and by the fact that organic solvents usually

lend an intractable consistency to the resulting extracts due

to the co-extraction of the polymeric cutin (Kolattukudy

1980; Huelin 1959). Hence, derivatives furnished by the

chemical derivatization of either the alcohol at C3 or the

carboxyl group at C17 in the pentacyclic ursane structure

would likely favor their separation from crude apple peel

extracts. In this context, a series of semi-synthesized

compounds have been reported as pharmacologically

active (Gnoatto et al. 2008; Ma et al. 2005).

This study addresses the isolation of ursolic methyl ester

obtained from Fuji apple peels by the previous esterifica-

tion of the crude organic extract with diazomethane. The

anti-inflammatory abilities of methyl ursolate (MU) and

UA (Fig. 1) were compared by assaying both of the com-

pounds simultaneously in the zymosan-induced models of

paw edema, pleurisy, and arthritis.

Materials and methods

Plant material and chemicals

Commercial apple fruits of the Fuji variety (15.0 kg) were

acquired from the central supplier of food and grocery in

the municipality of Campinas, Sao Paulo state, Brazil.

Standard UA (article U6753; chemical purity checked by

gas chromatography as[97 %) and Diazald� for preparing

diazomethane were purchased from Sigma-Aldrich (St

Louis, MO, USA). Zymosan, dexamethasone, potassium

diclofenac, UA, phosphate buffered saline (PBS), buffer

perborate, o-phenylenediamine dihydrochloride (OPD),

Bradford reagent, bovine serum albumin (BSA), ethylene

diaminetetraacetic sodium salt (EDTA) and concanavalin

A were purchased from Sigma Chemical Co. DMSO for

biological tests, ethyl ether, ethyl acetate, n-hexane,

dichloromethane, methanol and acetone for chromatogra-

phy were purchased from VETEC Ltd. Purified anti-murine

TNF-a, CXCL1/KC and IL-1b mAbs, biotinylated anti-

TNF-a, CXCL-1/KC and IL-1b mAbs, and recombinant

TNF-a, CXCL-1/KC and IL-1b were all obtained from

R&D Systems.

Extraction, methylation and separation

Apples were manually peeled using a kitchen peeler to

generate thin slices of fresh material that were intermit-

tently oven-dried three times at 100 �C for 4 h and then for

7 additional days at 45 �C under a constant air stream. The

resulting pieces of peel (442.3 g, 2.95 %) were ground by

successive pulsing in a kitchen blender and then sieved

(Bertel, BR) to form particle sizes between 850 and

1,000 lm. The grounded dry peels (40.68 g) in a 1 L of a

mixture of EtOH–AcOEt (85:15) were subjected to ultra-

sound (Unique USC-1850A) for 10 min. The suspension

was filtered and the procedure was repeated. The soluble

filtrates were pooled, and the solvent was removed in a

rotary evaporator to yield 7.63 g (18.7 %) of a viscous

yellowish plastic-like solid, which was dried overnight

under high vacuum. Part of the first extract (3.17 g) was

repeatedly treated with freshly prepared diazomethane in

ethyl ether. After drying, this methylated crude extract

(2.4 g) was subjected to a silica gel 60 (0.063–0.200 mm

Merck) column (U 3.0 cm 9 h 52 cm) and was eluted with

an AcOEt:n-hexane gradient (0–100 %) followed by

AcOEt:CH2Cl2 1:1, CH2Cl2 and CH2Cl2:acetone 1:1 to

generate 300 fractions (15–20 mL). These were pooled

according to the similarity in TLC (ceric sulfate as coloring

agent) to produce 9 groups. The group eluted between

CH2Cl2 and CH2Cl2:acetone 1:1 afforded 236 mg (10 %)

of pure MU (methyl 3b-hydroxyurs-12-en-28-oate).

Methyl ursolate

Colorless solid; IV (KBr, cm-1): 3,439 (b, O–H),

2,920–2,851 (s, C–H), 1,692 (s, C=O acid), 1,462 (m),

1,364 (w), 1,268 (w), 1,186 (w), 1,032 (w), 995–949 (split);

EI-MS m/z 470 [M?], 455 [M? – 15], 437, 410 [M? –

CH3COOH], 395, 377, 262 (RDA, 100 %), 249, 233, 203

(RDA), 189, 175, 161, 147, 133, 119 for C31H50O3. 13C

RMN (CHCl3, 100 MHz) d 39.7 (C1), 27.9 (C2), 79.7

(C3), 39.4 (C4), 56.0 (C5), 19.07 (C6), 33.7 (C7), 39.8

(C8), 48.3 (C9), 37.7 (C10), 24.0 (C11), 126.3 (C12), 138.9

(C13), 42.7 (C14), 28.8 (C15), 24.9 (C16), 48.8 (C17), 53.6

HO

CO2RH

H

Fig. 1 R = H (ursolic acid); R = CH3 (methyl ursolate)

T. A. Padua et al.

123

(C18), 40.2 (C19), 38.8 (C20), 30.7 (C21), 37.3 (C22), 28.2

(C23), 15.5 (C24), 16.3 (C25), 17.7 (C26), 24.3 (C27),

178.8 (C28), 16.6 (C29), 21.86 (C30) (Seebacher et al.

2003).

Animals

Male Swiss-44 and C57Bl/6 mice (20–30 g) from colony

(CECAL-FIOCRUZ) were maintained with a 12 h light/

dark cycle with controlled temperature and free access to

food and fresh water. All the experiments were conducted

in accordance with the ethical guidelines of the Interna-

tional Association for the Study of Pain (Zimmermann

1983) and the institutional guidelines for animal use

(CEUA L-0052/08).

Treatments

Mice that were fasted overnight received UA or MU

(50 mg/kg) orally (p.o.) in a final volume of 200 lL in 5 %

DMSO 1 h prior to stimulation. Dexamethasone (10 mg/kg,

100 lL) was administered intraperitoneally (i.p.) and

potassium diclofenac (100 mg/kg) was administered orally

(p.o.) 1 h prior to stimulation and were used as reference

drugs. The equivalent volume of vehicle was administered

to the control groups.

Cytotoxicity

Cellular viability in the presence and absence of the UA or

MU was determined using an Alamar Blue assay (Invit-

rogen). Mouse cell line J774-A was plated into black flat-

bottomed 96-well plates at a density of 2.5 9 106 cells/

well. After 1 h of incubation in a controlled atmosphere

(5 % CO2, 37 �C), the cells received fresh medium with or

without Tween 20 (3 %), DMSO (0.5 %) and UA or MU

(0.001–100 lM) in a quadruplicate assay. After 21 h of

further incubation, 20 lL of an Alamar Blue solution was

added to each well, and after 3 h the fluorescence was read

using SpectraMax M5/M5e microplate reader (Molecular

Devices; kexc = 555 nm, kem = 585 nm).

Macrophage activation and nitric oxide production

Mouse macrophages J77A-4 (1 9 105 cells/well in RPMI

with 10 % fetal bovine serum and gentamicin) were plated

into 96-well culture plates and allowed to adhere for 24 h.

The macrophages were incubated with different concen-

trations of UA and MU for 1 h. The control cells were

incubated with dexamethasone (50 nM). Following incu-

bation, the macrophages were either non-stimulated or

stimulated with IFN-c rich media in the presence or

absence of 37.5 ng/mL LPS at a final volume of 200 lL of

RPMI complete media. After 24 h at 37 �C/5 % CO2, the

amount of nitrite, a metabolite of nitric oxide (NO), was

measured in the supernatant from macrophages by the

Griess method (Moncada et al. 1991). The absorbance was

measured at 540 nm in a plate reader (Softmax Pro

190-Molecular Devices).

Paw edema

Preliminarily, the paw edema was induced in male C57Bl/6

pre-treated orally with different doses of MU (5–100 mg/kg)

and submitted to zymosan intraplantar injection (i.pl.,

100 lg/paw) diluted in sterile saline to a final volume of

50 lL (Henriques et al. 1987). Control paw received an i.pl.

injection of equal volume of sterile saline. In addition, to

compare the selected dose of MU (50 mg/kg) with UA

(50 mg/kg), Swiss-44 mice were separately pre-treated

orally with both compounds and also submitted to paw

edema model. In both experiments, 4 h after the stimulus, the

paw edema was evaluated by plethysmography (Plethys-

mometer 7140, Ugo Basile) and the values were expressed as

the difference between stimulated paw (lL) and non-stim-

ulated paw volume (lL) of each animal.

Pleurisy

One hour after the respective treatments, pleurisy was

induced in Swiss mice by an intrathoracic (i.t.) injection of

zymosan (100 lg/cavity) diluted in sterile saline to a final

volume of 100 lL, according to the technique of Spector as

modified for mice by Henriques et al. (1990). The control

group received an i.t. injection of an equal volume of

vehicle. The mice were euthanized 4 h after stimulus by

carbon dioxide inhalation, and their thoracic cavities were

washed with 1 mL of PBS containing EDTA (10 mM), pH

7.4. Total leukocyte counts were performed in an automatic

particle counter Z2: Counter (Coulter, Beckman Coulter).

Differential cell counts were performed using cytospin

smears (Shandon) stained by the May–Grumwald–Giemsa

method under light microscopy (1,0009). The counts were

reported as the number of cells (9106) per cavity. Pleural

washes were centrifuged at 400g during 10 min. Total

protein content of supernatants was quantified by the

Bradford method, according to the manufacturer’s

instructions.

Experimental arthritis model

Joint inflammation was induced by an intra-articular (i.a.)

injection of zymosan (500 lg/cavity in 25 lL of sterile

saline) by inserting a 27.5 G� needle through the supra-

patellar ligament into the left knee joint cavity, according

to the technique of Keystone, as previously modified by

Potential use in rheumatoid arthritis

123

Penido (Keystone et al. 1977; Penido et al. 2006). The

control group animals (C57Bl/6 mice) received an i.a.

injection with an equal volume of sterile saline. Knee joint

swelling was evaluated by the measurement of the trans-

verse diameters of each knee joint by a digital caliper

(Digmatic Caliper, Mitutoyo Corporation). The values of

knee joint thickness were expressed as the difference of the

knee joint diameter before and after the induction of

articular inflammation. The results were expressed as D in

the thickness of the knee joint in millimeters (mm). After

6 h of joint inflammation induction, mice were euthanized

by carbon dioxide inhalation. Knee synovial cavities were

washed with 300 lL of PBS containing EDTA (10 mM) by

the insertion of a 21 G needle into the mice knee joints, and

the synovial washes were recovered by aspiration. Total

leukocyte counts were performed in an automatic particle

counter-Z2 Counter (Coulter, Beckman Coulter). Differ-

ential cell counts were made using cytospin smears

(Shandon) stained by the May–Grumwald–Giemsa method

under light microscopy (1,0009). The counts were reported

as the number of cells per cavity (9105). The synovial

washes were centrifuged at 400g for 10 min. TNF-a,

CXCL-1/KC and IL-1b levels in supernatants were quan-

tified as described below.

ELISA assay

Levels of TNF-a, IL-1b, and keratinocyte-derived che-

mokine CXCL1/KC in the synovial washes were evaluated

by the sandwich ELISA assay, using matched antibody

pairs from R&D Systems (Minneapolis, MN, USA;

Quantikine), according to the manufacturer’s instructions.

Statistical analysis

Results were expressed as the mean ± SEM and were

statistically evaluated by a one-way ANOVA followed by

the Student–Newman–Keuls test. The significance level

was set at p B 0.05.

Results

To investigate cellular viability and to compare the effects

of UA and MU, an Alamar blue assay was performed. Only

the higher concentration (100 lM) of both UA and MU

reduced 25 % and 30 % of the cellular viability, respec-

tively. In the NO assay, UA (10 lM) inhibited only 9.4 %

of NO production. All other concentrations of UA and MU

were not able to inhibit NO production in LPS-stimulated

cells.

Other studies, such us Kang et al. 2008, have demon-

strated the ability of UA to inhibit the edema in inflammatory

models. To confirm the anti-inflammatory properties of UA

and to investigate the effect of MU, the experimental model

of paw edema was utilized. Preliminarily we carried out the

pre-treatment with different doses of MU (5–100 mg/kg) to

choose its effective dose. Doses of 50 and 100 mg/kg sig-

nificantly inhibited the zymosan-induced paw edema after

4 h (Fig. 2a). Therefore, subsequent experiments were car-

ried out at dose fixed as 50 mg/kg. For comparison between

the effect of UA and MU. the paw edema model was then

repeated. The i.pl. injection of zymosan induced an edema

after 4 h that was reduced after pre-treatment with UA

and MU (both 50 mg/kg) by 46 and 44 %, respectively

(Fig. 2).

The rodent zymosan-induced pleurisy is an inflamma-

tory model characterized by an increase of local vascular

permeability and a robust cell migration (Utsunomiya et al.

1998). The injection of zymosan in the pleural cavity of

mice induced a leukocyte influx and exudation 4 h after

stimulus. Pre-treatment with UA and MU did not inhibit

Fig. 2 Dose-response effect of methyl ursolate (MU) (5–100 mg/kg,

p.o., 1 h) on paw edema induced by zymosan, with potassium

diclofenac (Diclo., 100 mg/kg) as reference drug (a). Comparison

between effects of oral pre-treatment with ursolic acid (UA) and MU

(50 mg/kg, p.o., 1 h) on paw edema induced by zymosan (b).

*p \ 0.05 compared to zymosan; one-way ANOVA

T. A. Padua et al.

123

the leukocyte influx, whereas it inhibited the protein

extravasation (19 and 21 %, respectively) into the thoracic

cavity (Fig. 3d). Pre-treatment with the reference drug,

potassium diclofenac, reduced the cellular influx and pro-

tein extravasation as shown in Fig. 3a–d.

UA exerts some effects on articular inflammation in

animal models. To confirm this activity and to compare it

with the activity of MU, the zymosan-induced knee joint

inflammation was performed. The i.a. injection of zymosan

induced knee swelling and leukocyte influx into the syno-

vial cavity 6 h after stimulus. Pre-treatment with UA and

MU inhibited 40 and 48 % of tibio-femoral junction

edema, 52 and 73 % of the leukocyte influx and 48 and

74 % of the neutrophil influx, respectively (Fig. 4). Pre-

treatment with the reference drug, dexamethasone, signif-

icantly reduced the zymosan-induced arthritis to a similar

extent as MU (Fig. 4a–d).

Studies elsewhere have reported the important role of

some inflammatory mediators in the induction and

maintenance of inflammation on synovial tissue (Penido

et al. 2006; Verri et al. 2006). To investigate the

mechanisms by which UA and MU interfere in the

synovial inflammation, the concentrations of pro-inflam-

matory cytokines, TNF-a and IL-1b and the chemokine,

CXCL-1/KC, in synovial washes generated 6 h after

zymosan stimulus were assessed (Fig. 5). Stimulation

with zymosan increased the levels of these inflammatory

mediators, and pre-treatment with UA and MU signifi-

cantly inhibited the production of TNF-a (76 and 71 %,

respectively), IL-1b (57 and 53 %, respectively) and

CXCL-1/KC (95 and 90 %, respectively). These effects

were leveled against dexamethasone, which was used as

the positive control (Fig. 5a–d).

Discussion

The cytotoxic effects of UA (He and Liu 2007) on several

tumor cell lines revealed that high concentrations of UA

can also be harmful to healthy cells (Yamaguchi et al.

2008). With a few exceptions, MU has been shown to exert

a very similar behavior as UA in diverse tumoral cell

Fig. 3 Effect of the oral pre-treatment with ursolic acid (UA) and

methyl ursolate (MU) (50 mg/kg, p.o., 1 h) before pleurisy induced

by zymosan on total leucocyte (a), neutrophils (b) or mononuclear

cells migration (c) and on protein extravasation (d). Analysis was

performed 4 h after the stimulation. The reference drug used was

Diclofenac Potassium (Diclo).*p \ 0.05 compared to saline and?p \ 0.05 compared to zymosan; one-way ANOVA

Potential use in rheumatoid arthritis

123

lineages (Ma et al. 2005). Thus, our study confirmed that

both the compounds are equally cytotoxic.

The nitric oxide (NO)-producing system is a useful tool

for screening new substances with biological activity and is

recognized as an important factor in immunological,

nociceptive and inflammatory responses, although some

controversy has been raised recently in relation to the latter

process (Chaves et al. 2011). Our results showed that

neither UA nor MU resulted in the inhibition of NO pro-

duction in J774-A macrophages stimulated with LPS, other

than a small reduction caused by the highest dose (10 lM)

of UA. This result demonstrated that these substances do

not exhibit large effect on LPS stimulation.

UA presented a significant anti-inflammatory activity in

the carrageenan-induced rat paw edema and ear edema

induced either by tetradecanoylphorbol acetate or croton

oil (Ismaili et al. 2002; Park et al. 2004; Miceli et al. 2005).

Our observation in the zymosan-induced paw edema and

pleurisy models confirmed the anti-inflammatory activity

of UA as reported in the literature and, moreover, sug-

gested that MU presents a similar effect on paw edema.

These anti-inflammatory effects have been attributed to

free radical scavenging (Miceli et al. 2005) and the inhi-

bition of different chemical mediators involved in the

inflammatory reaction, such as histamine release and the

production of PGE2 (Ikeda et al. 2008). Interestingly, UA

and MU (both at 50 mg/kg) showed similar effects on

zymosan-induced pleurisy; both substances inhibited only

protein extravasation. Based on the effect of UA and MU

on the paw edema and pleurisy induced by zymosan, we

suggest that these substances inhibit the increased vascular

permeability in these models of inflammation.

Some studies have demonstrated the effects of UA on

different experimental models of arthritis on rodents (Ah-

mad et al. 2006; Kang et al. 2008). In this regard a poly-

saccharide from yeast cell walls, zymosan is capable of

stimulating inflammatory cytokine production and it has

served as a model for study of innate immune responses

(Underhill et al. 1999). The Toll-like receptor 2 (TLR2)

recognizes zymosan, acting in collaboration with CD14

and TLR6 (Song 2012). TLRs ligands like zymosan indu-

ces the activation of NF-jB and the production of

inflammatory cytokines such as IL-1b, CXCL1/KC and

TNF-a, that are responsible for triggering the inflammatory

Fig. 4 Effect of the oral pre-treatment with ursolic acid (UA) and

methyl ursolate (MU) (50 mg/kg, p.o., 1 h) on experimental arthritis

induced by zymosan on knee-joint edema formation (a), Total

leucocyte (b), neutrophils (c) and mononuclear cells influx (d).

Analyses were performed 6 h after stimulation. The reference drug

used was dexamethasone. *p \ 0.05 compared to saline and?p \ 0.05 compared to zymosan; one-way ANOVA

T. A. Padua et al.

123

process (Frasnelli et al. 2005; Guerrero et al. 2012). Our

study not only confirmed the anti-inflammatory activity of

UA but also described the effect of MU in zymosan-

induced arthritis (Fig. 4). Pre-treatment with the latter

compound decreased the tibio-femoral junction edema and

the leukocyte influx (mainly neutrophils) provoked by

intra-articular injection of zymosan. Thus, MU presented

effects similar to UA on mouse experimental arthritis;

however, the ester inhibited leukocyte influx more effec-

tively than the free acid. This observation suggests better

solubility and bioavailability of the MU.

Zymosan-induced arthritis produces a severe and ero-

sive synovitis (Chaves et al. 2011). This response is

mediate by the TLR2 bind in macrophages, leading to the

induction of pro-inflammatory cytokines, arachidonate

mobilization, protein phosphorylation and also activates

complement via the alternative pathway (Asquith et al.

2009). In addition, according to Guerrero and collabora-

tors, zymosan-induced joint hypernociception depends on

activation of TLR2/MyD88 that results in the production of

TNF-a, IL-1b and CXCL1/KC which acts in a reciprocal/

self-stimulatory cytokine cascade (Guerrero et al. 2012).

CXCL-1/KC is known as the relevant chemokine for

neutrophil influx in the early phase of synovial inflammation

(Penido et al. 2006). Furthermore, the cytokine TNF-a is

considered to be a pivotal mediator in chronic arthritis.

Because TNF-a and IL-1b are expressed very early on knee

joints, they play a key role in leading the inflammatory state

(Guerrero et al. 2012; Ferraccioli et al. 2010). In addition, IL-

1b is an anti-apoptotic and pro-inflammatory cytokine that is

primarily produced in activated monocytes and macro-

phages, and it is known that in rheumatoid arthritis, macro-

phage-derived IL-1b is associated with the increasing

production of eicosanoids, collagenase and PGE2 (Ikeda

et al. 2008). In addition, Kang and collaborators (2008)

showed that treatment with UA (50 mg/kg) significantly

decreased PGE2 serum concentration in the complete Fre-

und’s adjuvant (CFA)-induced arthritis. In line, the IL-1binhibition by UA could be responsible for PGE2 decreasing

on CFA-induced arthritis.

This study compared the parallel responses to arthritis-

related inflammation upon the administration of UA and

MU in mice. We demonstrated that both of these com-

pounds decreased the levels of CXCL-1/KC, TNF-a and

IL-1b in synovial washes from arthritis induced by

zymosan. These cytokines are important to arthritis path-

ogenesis and their production depends on NF-jB translo-

cation in activated cells like phagocytes (Pinto et al. 2010;

Guerrero et al. 2012). The blocking of antigen presentation

to T-cells and the inhibition of NF-jB nuclear translocation

in lymphocytes by UA have been reported elsewhere

(Checker et al. 2012). Considering that the UA and MU

were administrated before zymosan stimulation, these

compounds could be interfere on TLR-2 zymosan recog-

nition by resident macrophages, suppressing its activation

and the NF-jB translocation. Thereby, the production of

inflammatory cytokines such as CXCL-1/KC, TNF-a and

IL-1b and consequently the neutrophil chemotaxis are

suppressed (Pinto et al. 2010; Guerrero et al. 2012).

Overall, MU exerted similar anti-inflammatory effects

as UA in different animal models and a better inhibitory

Fig. 5 Effect of the oral pre-treatment with ursolic acid (UA) and

methyl ursolate (MU) (50 mg/kg, p.o., 1 h) on inflammatory medi-

ators present on experimental arthritis induced by zymosan KC/

CXCL-1 levels (a), TNF-a levels (b), IL-1b levels (c). Analyses were

performed in synovial washes obtained 6 h after stimulation. The

reference drug used was dexamethasone *p \ 0.05 compared to

saline and ?p \ 0.05 compared to zymosan; one-way ANOVA

Potential use in rheumatoid arthritis

123

action on the experimental parameters of arthritis, sug-

gesting that the ester might have potential uses in thera-

peutics for this synovial inflammatory condition.

From a pharmaceutical point of view, by demonstrating the

pharmacological equivalence between MU and the free acid,

our study suggests that the ester should be used in formulations

of anti-inflammatory products. This could be useful consid-

ering the extremely low solubility of UA in aqueous and oily

solvents (Jager et al. 2007, 2008a) or alcoholic mixtures (Jin

et al. 1997). The solvation of UA is most likely hampered by

its chemical structure, which is characterized by a large

lipophilic moiety bearing a secondary hydroxy and a carboxy

group on opposite sides of the molecule. On the other hand, the

methyl ester derivative may decrease the inherent polarity of

the UA molecule without significantly changing the molecular

size, which is an observation that may broaden the range of

solvents that are applicable to pharmaceutical goals. It is

known that less polar triterpene C3-alcohols—either as pure

compounds or triterpene-enriched extracts—have been suc-

cessfully incorporated in oleogel formulations for parenteral

and topical applications (Laszczyk et al. 2006; Jager et al.

2008b). Moreover, some synthetic C3-derivatives of UA have

been synthesized to overcome the solubility issues to furnish

compounds for use against several therapeutic targets (Gno-

atto et al. 2008; Baglin et al. 2003). These issues may support

the incorporation of MU instead UA in topical formulations

from apple peel extracts obtained by organic solvent extrac-

tions. Once the efficacy of UA may be surrogate by MU, some

new perspectives arise by employing the latter when aiming at

product development. Additionally, a cost-effective process

would be involved by avoiding the hardly workable UA iso-

lation from the cutin-rich peel extracts, particularly concern-

ing for the scale-up viability of extraction processes.

Acknowledgments We are grateful for the financial support from

Conselho Nacional de Desenvolvimento Cientıfico e Tecnologico,

MCTI (CNPq/Proc. 475751/2009-0;CNPq/Proc. 304588/2010-5).

References

Ahmad, S.F., B. Khan, S. Bani, K.A. Suri, N.K. Satti, and G.N. Qazi.

2006. Amelioration of adjuvant-induced arthritis by ursolic acid

through altered Th1/Th2 cytokine production. Pharmacological

Research 53: 233–240.

Asquith, D.L., A.M. Miller, I.B. Mcinnes, and F.Y. Liew. 2009.

Animal models of rheumatoid arthritis. European Journal of

Immunology 39: 2040–2044.

Baglin, I., A.C. Mitaine-Offer, M. Nour, K. Tan, C. Cave, and M.A.

Lacaille-Dubois. 2003. A review of natural and modified

betulinic, ursolic and echinocystic acid derivatives as potential

antitumor and anti-HIV agents. Mini Reviews in Medicinal

Chemistry 3: 525–539.

Baricevic, D., S. Sosa, R. Della Loggia, A. Tubaro, B. Simonovska,

A. Krasna, and A. Zupancic. 2001. Topical anti-inflammatory

activity of Salvia officinalis L. leaves: the relevance of ursolic

acid. Journal of Ethnopharmacology 75: 125–132.

Bringe, K., C.F. Schumacher, M. Schmitz-Eiberger, U. Steiner, and

E.C. Oerke. 2006. Ontogenetic variation in chemical and

physical characteristics of adaxial apple leaf surfaces. Phyto-

chemistry 67: 161–170.

Chattopadhyay, D., G. Arunachalam, A.B. Mandal, T.K. Sur, S.C.

Mandal, and S.K. Bhattacharya. 2002. Antimicrobial and anti-

inflammatory activity of folklore: Mallotus peltatus leaf extract.

Journal of Ethnopharmacology 82: 229–237.

Chaves, H.V., R.E.A. Ribeiro, A.M. De Souza, A.A. Rodrigues E

Silva, A.S. Gomes, M.L. Vale, M.M. Bezerra, and G.A. Brito.

2011. Experimental model of zymosan-induced arthritis in the

rat temporomandibular joint: Role of nitric oxide and neutro-

phils. Journal of Biomedicine and Biotechnology 2011: 707985.

Checker, R., S.K. Sandur, D. Sharma, R.S. Patwardhan, S. Jayakumar,

V. Kohli, G. Sethi, B.B. Aggarwal, and K.B. Sainis. 2012. Potent

anti-inflammatory activity of ursolic acid, a triterpenoid antiox-

idant, is mediated through suppression of NF-jB, AP-1 and NF-

AT. PLoS One 7: e31318.

Ferraccioli, G., L. Bracci-Laudiero, S. Alivernini, E. Gremese, B.

Tolusso, and F. De Benedetti. 2010. Interleukin-1b and

interleukin-6 in arthritis animal models: roles in the early phase

of transition from acute to chronic inflammation and relevance

for human rheumatoid arthritis. Molecular Medicine 16:

552–557.

Frasnelli, M.E., D. Tarussio, V. Chobaz-Peclat, N. Busso, and A. So.

2005. TLR2 modulates inflammation in zymosan-induced

arthritis in mice. Arthritis Research & Therapy 7: R370–R379.

Gnoatto, S.C., L.D. Vechia, C.L. Lencina, A. Dassonville-Klimpt,

S.D. Nascimento, D. Mossalayi, J. Guillon, G. Gosmann, and P.

Sonnet. 2008. Synthesis and preliminary evaluation of new

ursolic and oleanolic acids derivatives as antileishmanial agents.

Journal of Enzyme Inhibition and Medicinal Chemistry 23:

604–610.

Guerrero, A.T., T.M. Cunha, W.A. Verri, R.T. Gazzinelli, M.M.

Teixeira, F.Q. Cunha, and S.H. Ferreira. 2012. Toll-like receptor

2/MyD88 signaling mediates zymosan-induced joint hypernoci-

ception in mice: participation of TNF-a, IL-1b and CXCL1/KC.

European Journal of Pharmacology 674: 51–57.

He, X., and R.H. Liu. 2007. Triterpenoids isolated from apple peels

have potent antiproliferative activity and may be partially

responsible for apple’s anticancer activity. Journal of Agricul-

ture and Food Chemistry 55: 4366–4370.

Henriques, M.G., P.M. Silva, M.A. Martins, C.A. Flores, F.Q. Cunha,

J. Assreuy-Filho, and R.S. Cordeiro. 1987. Mouse paw edema. A

new model for inflammation? Brazilian Journal of Medical and

Biological Research 20: 243–249.

Henriques, M.G.M.O., V.B. Weg, M.A. Martins, P.M.R. Silva, P.D.

Fernandes, R.S.B. Cordeiro, and B.B. Vargaftig. 1990. Differ-

ential inhibition by 2 hetrazepine Paf antagonists of acute-

inflammation in the mouse. British Journal of Pharmacology 99:

164–168.

Huelin, F.E. 1959. Studies in the natural coating of apples. Australian

Journal of Biological Sciences 12: 175–180.

Ikeda, Y., A. Murakami, and H. Ohigashi. 2008. Ursolic acid: An

anti- and pro-inflammatory triterpenoid. Molecular Nutrition &

Food Research 52: 26–42.

Ismaili, H., S. Sosa, D. Brkic, S. Fkih-Tetouani, A. Ilidrissi, D.

Touati, R.P. Aquino, and A. Tubaro. 2002. Topical anti-

inflammatory activity of extracts and compounds from Thymus

broussonettii. Journal of Pharmacy and Pharmacology 54:

1137–1140.

Jin, I.J., Y.I. Ko, Y.M. Kim, and S.K. Han. 1997. Solubilization of

oleanolic acid and ursolic acid by cosolvency. Archives of

Pharmacal Research 20: 269–274.

Jager, S., M.N. Laszczyk, and A. Scheffler. 2008a. A preliminary

pharmacokinetic study of betulin, the main pentacyclic triterpene

T. A. Padua et al.

123

from extract of outer bark of birch (Betulae alba cortex).

Molecules 13: 3224–3235.

Jager, S., M.N. Laszczyk, C.M. Struh, S.F. Martin, C.M. Schempp,

and A. Scheffler. 2008b. Triterpenes from natural sources for

parenteral and topical application. Planta Medica 74: 1119.

Jager, S., K. Winkler, U. Pfuller, and A. Scheffler. 2007. Solubility

studies of oleanolic acid and betulinic acid in aqueous solutions

and plant extracts of Viscum album L. Planta Medica 73:

157–162.

Kang, S.Y., S.Y. Yoon, D.H. Roh, M.J. Jeon, H.S. Seo, D.K. Uh, Y.B.

Kwon, H.W. Kim, H.J. Han, H.J. Lee, and J.H. Lee. 2008. The

anti-arthritic effect of ursolic acid on zymosan-induced acute

inflammation and adjuvant-induced chronic arthritis models.

Journal of Pharmacy and Pharmacology 60: 1347–1354.

Keystone, E.C., H.U. Schorlemmer, C. Pope, and A.C. Allison. 1977.

Zymosan-induced arthritis: A model of chronic proliferative

arthritis following activation of the alternative pathway of

complement. Arthritis and Rheumatism 20: 1396–1401.

Kolattukudy, P.E. 1980. Biopolyester membranes of plants: Cutin and

suberin. Science 208: 990–1000.

Laszczyk, M., S. Jager, B. Simon-Haarhaus, A. Scheffler, and C.M.

Schempp. 2006. Physical, chemical and pharmacological charac-

terization of a new oleogel-forming triterpene extract from the outer

bark of birch (betulae cortex). Planta Medica 72: 1389–1395.

Lee, J., S.T. Yee, J.J. Kim, M.S. Choi, E.Y. Kwon, K.I. Seo, and M.K.

Lee. 2010. Ursolic acid ameliorates thymic atrophy and hyper-

glycemia in streptozotocin–nicotinamide-induced diabetic mice.

Chemico-Biological Interactions 188: 635–642.

Liu, J. 2005. Oleanolic acid and ursolic acid: Research perspectives.

Journal of Ethnopharmacology 100: 92–94.

Ma, C.M., S.Q. Cai, J.R. Cui, R.Q. Wang, P.F. Tu, M. Hattori, and M.

Daneshtalab. 2005. The cytotoxic activity of ursolic acid deriv-

atives. European Journal of Medicinal Chemistry 40: 582–589.

Miceli, N., M.F. Taviano, D. Giuffrida, A. Trovato, O. Tzakou, and

E.M. Galati. 2005. Anti-inflammatory activity of extract and

fractions from Nepeta sibthorpii Bentham. Journal of Ethno-

pharmacology 97: 261–266.

Mitaine-Offer, A.C., W. Hornebeck, M. Sauvain, and M. Zeches-

Hanrot. 2002. Triterpenes and phytosterols as human leucocyte

elastase inhibitors. Planta Medica 68: 930–932.

Moncada, S., R.M. Palmer, and E.A. Higgs. 1991. Nitric oxide:

Physiology, pathophysiology, and pharmacology. Pharmacolog-

ical Reviews 43: 109–142.

Park, J.H., K.H. Son, S.W. Kim, H.W. Chang, K. Bae, S.S. Kang, and

H.P. Kim. 2004. Antiinflammatory activity of Synurus deltoides.

Phytotherapy Research 18: 930–933.

Penido, C., F.P. Conte, M.S. Chagas, C.A. Rodrigues, J.F. Pereira,

and M.G. Henriques. 2006. Antiinflammatory effects of natural

tetranortriterpenoids isolated from Carapa guianensis Aublet on

zymosan-induced arthritis in mice. Inflammation Research 55:

457–464.

Pinto, L.G., T.M. Cunha, S.M. Vieira, H.P. Lemos, W.A. Verri, F.Q.

Cunha, and S.H. Ferreira. 2010. IL-17 mediates articular

hypernociception in antigen-induced arthritis in mice. Pain

148: 247–256.

Seebacher, W., N. Simic, R. Wels, R. Saf, and O. Kunert. 2003.

Complete assignments of 1H and 13C NMR resonances of

oleanolic acid, 18-a-oleanolic acid, ursolic acid and their 11-oxo

derivatives. Magnetic Resonance in Chemistry 41: 636–638.

Song, W.C. 2012. Crosstalk between complement and toll-like

receptors. Toxicologic Pathology 40: 174–182.

Sultana, N., and Z.S. Saify. 2012. Naturally occurring and synthetic

agents as potential anti-inflammatory and immunomodulants.

Antiinflammatory & Antiallergy Agents in Medicinal Chemistry

11: 3–19.

Szakiel, A., C. Paczkowski, F. Pensec, and C. Bertsch. 2012. Fruit

cuticular waxes as a source of biologically active triterpenoids.

Phytochemistry Reviews 11: 263–284. doi:10.1007/s11101-012-

9241-9.

Takeoka, G., L. Dao, R. Teranishi, R. Wong, S. Flessa, L. Harden,

and R. Edwards. 2000. Identification of three triterpenoids in

almond hulls. Journal of Agricultural and Food Chemistry 48:

3437–3439.

Underhill, D.M., A. Ozinsky, A.M. Hajjar, A. Stevens, C.B. Wilson,

M. Bassetti, and A. Aderem. 1999. The Toll-like receptor 2 is

recruited to macrophage phagosomes and discriminates between

pathogens. Nature 401: 811–815.

Utsunomiya, I., M. Ito, and S. Oh-Ishi. 1998. Generation of

inflammatory cytokines in zymosan-induced pleurisy in rats:

TNF induces IL-6 and cytokine-induced neutrophil chemoat-

tractant (CINC) in vivo. Cytokine 10: 956–963.

Verri, W.A., T.M. Cunha, C.A. Parada, S. Poole, F.Q. Cunha, and

S.H. Ferreira. 2006. Hypernociceptive role of cytokines and

chemokines: Targets for analgesic drug development? Pharma-

cology & Therapeutics 112: 116–138.

Yamaguchi, H., T. Noshita, Y. Kidachi, H. Umetsu, M. Hayashi, K.

Komiyama, S. Funayama, and K. Ryoyama. 2008. Isolation of

ursolic acid from apple peels and its specific efficacy as a potent

antitumor agent. Journal of Health Science 54: 654–660.

Ying, Q.L., A.R. Rinehart, S.R. Simon, and J.C. Cheronis. 1991.

Inhibition of human leucocyte elastase by ursolic acid. Evidence

for a binding site for pentacyclic triterpenes. Biochemical

Journal 277(Pt 2): 521–526.

Zimmermann, M. 1983. Ethical guidelines for investigations of

experimental pain in conscious animals. Pain 16: 109–110.

Potential use in rheumatoid arthritis

123