the ayurvedic medicine salacia oblonga attenuates diabetic...

Original Article

The Ayurvedic medicine Salacia oblonga attenuates diabetic renalfibrosis in rats: suppression of angiotensin II/AT1 signaling

Lan He1, Yanfei Qi2, Xianglu Rong3, Jianmin Jiang1, Qinglin Yang4, Johji Yamahara5,Michael Murray2 and Yuhao Li2

1School of Pharmaceutical Sciences, Sun Yat-Sen University, Guangzhou, 510006, China, 2Faculty of Pharmacy,The University of Sydney, Sydney, 2006, Australia, 3Department of Pharmacology, Guangzhou University ofChinese medicine, Guangzhou, 510006, China, 4Department of Nutrition Sciences, University of Alabama,Birminham, 35294-3360, USA and 5Pharmafood Institute, Kyoto, 602-8136, Japan

In human diabetic nephropathy, the extent of tubulointerstitial fibrosis is the leading cause ofend-stage renal disease; fibrosis is closely correlated with renal dysfunction. Although a widearray of medicinal plants play a role in the prevention and treatment of diabetes, there are fewreports of the application of herbal medicines in amelioration of renal fibrosis, or the under-lying mechanisms by which such benefits are mediated. The efficacy of the Ayurvedic antidia-betic medicine Salacia oblonga (SO) root on rat renal fibrosis was investigated. An aqueousextract from SO (100mg/kg, p.o., 6 weeks) diminished renal glomerulosclerosis and interstitialfibrosis in Zucker diabetic fatty (ZDF) rats, as revealed by van Giesen-staining. SO alsoreduced renal salt-soluble, acid-soluble and salt-insoluble collagen contents. These changeswere accompanied by normalization of hypoalbuminemia and BUN. Gene profiling revealedthat the increase in transcripts encoding the glomerulosclerotic mediators collagen I, collagenIV, fibronectin, angiotensin II type 1 receptor (AT1), transforming growth factor (TGF)-b1,plasminogen activator inhibitor (PAI)-1 observed in ZDF rat kidney was suppressed by SO.In rat-derived mesangial cells, similar to the effect of the AT1 antagonist telmisartan, SO andits major component mangiferin suppressed the stimulatory effect of angiotensin II on prolif-eration and increased mRNA expression and/or activities of collagen I, collagen IV, fibronectin,AT1, TGF-b1 and PAI-1. Considered together the present findings demonstrate that SO attenu-ates diabetic renal fibrosis, at least in part by suppressing anigiotensin II/AT1 signaling.Further, it now emerges that mangiferin is an effective antifibrogenic agent.

Keywords: angiotensin II – diabetes – fibrosis – kidney – Salacia

Introduction

Fibroproliferative diseases, including cardiovasculardisease, progressive kidney disease and macular degen-

eration, are leading causes of morbidity and mortalityand can affect all tissues and organ systems.

Nevertheless, despite the enormous impact of these

conditions on human health, there are currently no

approved treatments that directly target the mechanism(s)

of fibrosis (1). Diabetic nephropathy is the leading cause

of end-stage renal disease worldwide and is an indepen-

dent risk factor for all-cause and cardiovascular mortal-

ities in diabetic patients. Diabetic nephropathy is

structurally characterized by an early thickening of tubu-

lar and glomerular basement membranes due to the

excessive accumulation of the extracellular matrix (2).

Excessive deposition of extracellular matrix in the

For reprints and all correspondence: Yuhao Li, Faculty of Pharmacy,The University of Sydney, NSW 2006 Australia. Tel: þ61 2 9351 8585;Fax: þ61 2 9351 8638; E-mail: [email protected]

eCAM 2009; of 13

doi:10.1093/ecam/nep095

� 2009 The Author(s).This is an Open Access article distributed under the terms of the Creative Commons Attribution Non-Commercial License (http://creativecommons.org/licenses/by-nc/2.0/uk/) which permits unrestricted non-commercial use, distribution, and reproduction in any medium, provided the original work isproperly cited.

eCAM Advance Access published August 25, 2009

http://creativecommons.org/

glomerular mesangium and tubulointerstitium is closelyassociated with a progressive decline in renal functionin diabetes (2,3). In human diabetic nephropathy, theextent of glomerulosclerosis and tubulointerstitial fibrosisare well correlated with renal dysfunction (4–6).Accordingly, prevention and treatment of renal fibrosismay improve renal dysfunction.It is well known that activation of the renin–

angiotensin–aldosterone system is critical for develop-ment of cardiovascular and renal fibrosis in hypertensionand diabetes. Although all major components of therenin–angiotensin–aldosterone system can exhibit profi-brotic activity, angiotensin II, the principal constituentpeptide of the renin–angiotensin–aldosterone system,plays an important role in the development of renalfibrosis (7). Chronic angiotensin II infusion inducesextensive renal fibrosis in rats (8). The angiotensin con-verting enzyme inhibitor enalapril inhibits glomerulargene expression of extracellular matrix proteins in dia-betic rats (9). In mesangial cell culture, angiotensin IIincreases the synthesis of matrix proteins such as collagentype I and fibronectin; this can be inhibited by the angio-tensin II type 1 receptor (AT1) antagonist losartan (10).It has been suggested recently that targeting the renin–angiotensin–aldosterone system might be an effectivestrategy to decrease fibrogenesis in progressive renaldisease (11).Recently, there has been increased global interest in

traditional medicines, including Ayurveda (traditionalIndian) and Chinese medicines (12–14). Although thesemedicines have been used for prevention and treatmentof various disorders for several thousand years, moreextensive scientific evaluation of their mechanisms ofaction and provision of a strong evidence base are stillrequired. Salacia species (e.g. S. oblonga, S. prinoides,S. reticulata), known as ‘Ponkoranti’ in Ayurvedic med-icine, are widely distributed in Sri Lanka, India, China,Vietnam, Malaysia, Indonesia and other countries.The roots and stems of these plants have been used inthe treatment of diabetes and obesity in the Ayurvedicsystem of Indian traditional medicine (15,16). Mangiferin,one of the main components in Salacia species, is alsopresent in relatively abundant amounts in some fruits(e.g. mango) and other traditional anti-diabetic herbsand teas (e.g. Anemarrhena asphodeloides, Mangiferaindica, Aspalathus linearis and Cyclopia intermedia)(16,17). Recently, Salacia species have been extensivelyconsumed in Japan, USA and other countries asfood supplements for the prevention of obesity anddiabetes and they have been the subject of broad studiesfor diabetes management. It has been demonstratedthat aqueous extracts of the root of S. oblonga Wall.(Celastraceae) (SO) activate the peroxisome proliferator-activated receptor (PPAR)-a (18), and ameliorate post-prandial hyperglycemia, hyperlipidemia, hepatic steatosis,cardiac lipid accumulation and fibrosis in Zucker diabetic

fatty (ZDF) rats (18–21). It has not yet been evaluatedwhether this traditional antidiabetic herbal medicine isalso beneficial in renal complications associated withdiabetes.The aims of the present study were to examine the

effects and underlying mechanism of SO and its majorcomponent mangiferin on rat renal fibrosis using ZDFrats and rat-derived primary mesangial cells.

Methods

Chemicals and Reagents

Kits for determination of collagen (Biocolor, Brisbane,Australia), albumin (BioAssay Systems, Sydney,Australia), BUN (Sigma, Sydney, Australia) and glucose(Wako, Osaka, Japan), and uric acid (Wako, Osaka,Japan) and mangiferin (MA, Sigma, Sydney, Australia)were purchased commercially.

Animals and Diet

Male Zucker lean (ZL) and ZDF rats aged 13–15weeks were obtained from Monash University AnimalServices (Clayton, Victoria, Australia). Male Sprague-Dawley rats weighing 130–150 g were purchased fromthe Experimental Animal Center, Sun Yat-SenUniversity (Guangzhou, China). The animals werehoused in an air-conditioned room at 23� 1�C with a12-h light/dark cycle and were provided ad libitum withwater and standard pelleted diets. Animals were allowedfree access to the food and water. All animal experimen-tal procedures were in accordance with guidelines set bythe National Health and Medical Research Council ofAustralia, and approved by the animal ethics committeesof the University of Sydney, Australia or Sun Yat-SenUniversity, China.

Preparation and Identification of Aqueous SO Extract

SO roots were collected from Tamil Nadu, India andtheir identity was confirmed using botanical and pharma-cognostic criteria. The voucher sample was depositedwith the Pharmafood Institute (Kyoto, Japan; voucherNo: PS0075). The aqueous SO extract was prepared aspreviously described (19). The yield of the extract fromthe dried root was 6.5%. SO extract was characterizedby HPLC (21) and the content of MA, a prominent com-ponent considered suitable for the quality assurance ofSalacia species and its products (22), was found to be1.4%, which is within the previously reported range(Japanese patent P2002-267655).

Treatment Protocol

Plasma glucose levels in non-fasted animals were deter-mined prior to any treatment. Animals were then

2 of 13 Salacia oblonga attenuates renal fibrosis

subdivided into experimental groups on the basis of

plasma glucose level and body weight: ZL control, ZL

SO, ZDF control and ZDF SO groups (five animals

per group). There were no differences in plasma glucose

concentrations or body weight between the two ZL

groups or between the two ZDF groups; these parameters

were substantially lower in ZL groups than in ZDF

groups (data not shown), which is consistent with the

previous reports (18–21). SO at a dose of 100mg/kg,

which was obtained from the previous experiments

(18–21), was suspended in 5% Gum Arabic and adminis-

tered to animals by oral gavage once daily for 6 weeks.

The same volume of vehicle (5% Gum Arabic) was admi-

nistered to animals in the control group. The animals

were weighed once after 3–4 days in order to determine

the volumes of the test sample to be administered.

Measurement of Blood Biochemical Parameters and

Kidney Weight

Non-fasting blood samples (anti-coagulated with heparin)

were obtained from the tail veins of rats at the age of

20–22 weeks using light halothane anesthesia. Plasma

samples were stored at �80�C for subsequent determina-

tion of albumin, BUN and uric acid (using the commer-

cially available kits, listed above). The animals were

euthanized by rapid dislocation of the neck vertebra.

The kidneys were rapidly excised and accurately weighed.

A segment of the right kidney was cut into slices, frozen

in liquid nitrogen and stored at �80�C for collagen assay

and gene expression analysis.

Renal Histology and Determination of Renal Fibrosis

The remaining kidney segment was immediately fixed in

10% neutral formalin and prepared for pathological

examination. Renal fibrosis was measured essentially as

described previously (19,23). To determine the degree of

collagen fiber accumulation, forty fields in three individ-

ual sections were randomly selected and the areas of

van Giesen-stained interstitial collagen deposit and the

total kidney area were determined by image analysis

(KS 400 Imaging System; Carl Zeiss Vision, Eching,

Germany). The ratio of the areas of interstitial collagen

deposit to the total kidney area was calculated. The

glomerular cross-sectional areas in hematoxylin-eosin

(HE)-stained slices were determined morphometrically.

The cross-sectional area was measured in glomeruli

that were cut transversely. The outer borders of the glo-

meruli were traced at 200� magnification, and the

glomerular areas were calculated. Fifty glomeruli per

kidney were counted, and the averaged value was used

for analysis.

Renal Collagen Assay

The composition of collagen in renal tissue was analyzedusing the commercially available Sircol Collagen Assay(Biocolor Ltd, Northern Ireland), a colorimetric methodin which Sirius red dye binds to the side chains of aminoacids found in collagen. The salt soluble collagen andcovalently cross-linked insoluble collagen fractions wereobtained by the successive treatment of rat kidney withsalt soluble collagen buffer (0.05M Tris buffer, pH 7.5,1.0M NaCl), acetic acid (0.5M, pH 3.0) and Milli-Qwater (with heating to 80�C, 1 h). Aliquots of extract(50 ml) were mixed gently with 1ml of Sircol dye reagent,incubated for 30min at room temperature, and then cen-trifuged at 10 000g to isolate the collagen-bound dye inprecipitated material. One milliliter of alkali reagent wasthen added to release the collagen-bound dye into thesupernatant. After vortex-mixing (5min) and centrifuga-tion at 10 000g (15min), 200ml of supernatant was trans-ferred to 96-well plate, and the absorbance at 550 nm wasmeasured in a micro-plate reader. The concentration wasdetermined from standard curves, and the collagen con-tent in the kidney tissue was expressed relative to thetotal dry weight.

Cell Culture

Primary cultures of rat mesangial cell were prepared aspreviously described (24–26). Briefly, rats were eutha-nased by prompt dislocation of the neck vertebra. Inorder to prevent contamination with interstitial fibro-blasts the bilateral kidney cortical tissues distant fromthe medulla were harvested. The tissues were minced inHanks’ balanced salt solution. The suspension was fil-tered using 100, 200 and 250 mesh sieves and centrifugedtwice at 400g for 2min to obtain purified renal glomeruli.The resultant cell pellets were dispersed and incubated inHBSS containing 375U/ml collagenase type 1 (Gibco,China) at 37�C for 15min. After recentrifugation thecell pellet was re-suspended in a mixture of 80% RPMI1640 medium (Gibco) and 20% fetal bovine serum con-taining 100 ng/ml insulin, 55 ng/ml transferrin, 50 pg/mlselenium acid and antibiotics (Gibco). The cells wereincubated in an environment of 5% CO2–95% O2 at37�C. Mesangial cells were identified by morphologicalcharacterization and immunocytochemistry; they stainedpositive for vimentin and a-smooth muscle actin and neg-ative for factor VIII and cytokeratin expression. Cells atpassage 1–4 were used in all experiments.

Cell Viability and Proliferation

Cell viability was determined by MTT assay. In brief,cells (1� 104 cells) were grown to 70% confluencein 48 cell culture wells and were then serum-starvedfor �16–18 h. The medium was replaced with fresh

eCAM 2009 3 of 13

medium containing various concentrations of SO (10, 20and 40 mg/ml), MA (25mM, Sigma, Guangzhou, China)or telmisartan (10mM, Sigma) 1 h before treatment withangiotensin II (10�6M). After 20 h, MTT stock solution(5mg/ml; Sigma) was added and 4 h later the media wasremoved and cells were kept in 200 ml of DMSO for15min. The optical density at 570 nm was measured ina Gemini Microplate reader.DNA synthesis was assayed by 3H-thymidine incor-

poration. Confluent mesangial cells in 24-well culturedishes were serum-deprived for 16–18 h. SO (10, 20 and40 mg/ml), MA (25mM) or telmisartan (10 mM) was addedto the medium with 10% FBS and 1 mCi/ml 3H-thymidine(Beijing Atom High-tech Co. Ltd, Beijing, China) 1 hbefore treatment with angiotensin II (10�6M) and 24 hlater, the cells were washed three times with cold PBS(Boster, AR0030, Wuhan, China). DNA was precipitatedat 4�C by treatment with 5% trichloroacetic acid for1 h and the pellet was dissolved in 0.1M NaOH andabsolute ethanol; radioactivity was determined by liquidscintillation counting (Beckman Counter, LS6000,Minnesota, USA).

ELISA Analysis of Fibronectin and Transforming Growth

Factor-b Protein Levels in Rat Mesangial Cells

Confluent rat primary mesangial cells in 6-well culturedishes were serum-deprived for �16–18 h. SO (10, 20and 40 mg/ml), MA (25mM) or telmisartan (10 mM) wereadded to the medium in 10% FBS 1 h prior to treatmentwith angiotensin II (10�6M). Twenty-four hours later,the medium was collected for determination of fibronec-tin and transforming growth factor (TGF-b) proteinlevels by ELISA (Boster) according to the manufacturer’sinstructions.

Gene Expression Analysis in Rat Kidney and

Rat-derived Mesangial Cells

Total RNA was prepared from the kidneys of individualrats or rat mesangial cells using TRIzol reagent(Invitrogen, Australia and Invitrogen, China). The rela-tive levels of specific mRNAs were determined by reverse

transcriptase polymerase chain reaction (RT–PCR),as described previously (18,20,21). Single-strandedcDNA was synthesized from 1 mg of total RNA usingSuperScript II RNase H Reverse Transcriptase, as perthe manufacturer’s instructions (Invitrogen, Australiaand Invitrogen, China). PCR was performed on a ther-mocycler (PTC-200 DNA engine, MJ Research Inc, MA,USA). The genes examined were collagen I, collagenIV, fibronectin, plasminogen activator inhibitor (PAI)-1,urokinase-type plasminogen activator (uPA), TGF-b1and AT1. The sequences of the sense and antisenseprimers used for amplification are shown in Table 1.PCR samples were electrophoresed on 3% agarose gelsand stained with ethidium bromide. The gel images weredigitally captured with a CCD camera and analyzedusing ImageJ 1.29x software (NIH, USA). RT–PCRvalues are presented as ratios of the specified genesignal in the selected linear amplification cycle relativeto the signal for b-actin (housekeeping gene).

Data Analysis

All results are expressed as means�SEM. Datawere analyzed by single-factor analysis of variance(ANOVA). If a statistically significant effect was found,the Newman–Keuls test was performed to isolate thedifference between the groups. P-values50.05 were con-sidered to indicate significance.

Results

Renal Histology and Collagen Accumulation

in ZL and ZDF Rats

SO has been demonstrated previously to ameliorate post-prandial hyperglycemia, hyperlipidemia, hepatic steatosis,cardiac lipid accumulation and fibrosis in ZDF rats(18–21). In the present study, the extent of renal fibrosisin ZL and ZDF rats was first assessed using thevan Giesen-staining method. Renal glomerulosclerosisand interstitial fibrosis with tubular atrophy were evidentin sections from ZDF rats (Fig. 1C). The stained area

Table 1. Sequences of primers used in the amplification of renal genes by RT–PCR

Gene Accession no.or reference

Forward primer Reverse primer Size(bp)

Temp.(�C)

Collagen I Li et al. (27) 50-CATAAAGGGTCATCGTGGCTTC-30 50-GTGATAGGTGATGTTCTGGGAG-30 �500 62Collagen IV Li et al. (27) 50-GCAGGTGTGCGGTTTGTGAAG-30 50-AGCTCCCCTGCCTTCAAGGTG-30 328 55Fibronectin NM019143 50-CAAGACCATACCTGCCGAAT-30 50-CCGTGTAAGGGTCAAAGCAT-30 417 63AT1 NM030985 50-GAGAGGATTCGTGGCTTGAG-30 50-TAAGTCAGCCAAGGCGAGAT-30 464 62TGF-b1 NM021578 50-GACCTGCTGGCAATAGCTTC-30 50-GACTGGCGAGCCTTAGTTTG-30 468 58PAI-1 M24067 50-CCCTTCCAGAGTCCCATACA-30 50-CAGGCGTGTCAGCTCATTTA-30 485 60uPA NM013085 50-CAGATCCGATGCTCTTAGCC-30 50-GCTGCTCCACCTCAAACTTC-30 458 61b-actin NM031144 50-AGCCATGTACGTAGCCATCC-30 50-CTCTCAGCTGTGGTGGTGAA-30 228 60

Temp, temperature.


(red color) was more extensive in the glomerular mesan-gium and tubulointerstitium of ZDF rats (Fig. 1C),compared with ZL rats (Fig. 1A). Thus, the area of col-lagen deposition as a proportion of total kidney tissuewas increased in ZDF rats to 2-fold of ZL controls(P50.05; Fig. 1E). The kidney weights of ZDF ratswere increased by 15% over those in ZL rats (P50.05;Fig. 1F). SO treatment over a 6-week period significantlydiminished the extent of van Gieson-staining by 40% inZDF rats (P50.05), but had minimal effect in ZL rats(Fig. 1B, D and E). This treatment did not affect kidneyweight in rats of either genotype (Fig. 1F).Glomerular cross-sectional areas in ZDF rats were

1.5-fold of those in ZL rats (P50.05; Fig. 2A, Cand E). However, SO treatment minimally affected theglomerular cross-sectional area in kidney of ZL andZDF rats (Fig. 2B, D and E).Furthermore, the effects of SO treatment on the nature

of the collagens present in rat kidneys were analyzed.It was found that the quantities of total salt soluble(Fig. 3A), acid soluble (Fig. 3B) and insoluble (Fig. 3C)collagens were elevated by 46, 22 and 72%, respectively,in the kidneys of ZDF rats compared with ZL control(P50.05). Although treatment with SO significantlydecreased renal total collagen contents in all animals,

the effects were more pronounced in ZDF rats than inZL rats.

Biochemical Parameters Relating to Renal Function in ZL

and ZDF rats

Glomerular and interstitial injury and fibrosis may inducerenal dysfunction. Consistent with previous reports(27,28), ZDF rats were found in the present study toexhibit elevated BUN (34% increase compared with ZLcontrol; P50.05) and plasma uric acid levels (to 2-foldof ZL control; P50.05) as well as decreased serum albu-min concentrations (to 75% of ZL control; P50.05)(Fig. 3D–F), which is due to prolonged urinary proteinexcretion. Importantly, SO treatment restored albuminconcentrations to those found in ZL rat plasma andalso normalized BUN in ZDF rats (Fig. 3D and E).However, SO treatment did not affect plasma uric acid(UA) concentrations in ZDF rats (Fig. 3F), or the corre-sponding parameters in ZL rats.

Cell Proliferation in Rat-derived Mesangial Cells

The effect of SO, MA and the AT1 antagonist telmisar-tan on angiotensin II-induced mesangial cell proliferation

DC

ZDF Con ZDF SO

B

ZL Con

A

ZL SO

E

0

1

2

3

Ren

al f

ibro

sis

(%)

* *

ZL ZDF

SO SOConCon

Kid

ney

Wei

ght

(g)

*

0

1

2

3

4F

ZL ZDF

SO SOConCon

Figure 1. Renal fibrosis and kidney weights 6 weeks after treatment of male ZL and ZDF rats with vehicle (con) or SO extract. The renal van

Giesen-stained collagen deposit area and the total renal tissue area were determined by image analysis. (A–D) Representative samples of renal fibrosis

(red) (magnification 200�); (E) the ratio of the area of collagen accumulation to the total renal tissue area; (F) kidney weight. All values are

means� SEM (n¼ 5, each group). *P50.05.

eCAM 2009 5 of 13

was assessed. Cell viability, as reflected by MTT assay,and proliferation, reflected by 3H-thymidine incorpora-tion, were significantly increased (by 18 and 70%, respec-tively) after stimulation with angiotensin II (Fig. 4Aand B). Treatment with SO at concentrations of 20and 40 mg/ml, as well as MA (25mM) and telmisartan(10 mM), all suppressed the angiotensin II-inducedincrease in cell viability (between 10 and 15%; allP50.05) and proliferation (by 18–26%; all P50.05),whereas the low concentration of SO (10 mg/ml) showedminimal effect.

Renal Gene Expression Profiles

Genes Implicated in Fibrogenesis

Deposition of extracellular matrix proteins includingcollagen I, collagen IV and fibronectin is an importantcomponent of the scarring observed during the evolutionof glomerulosclerosis and tubulointerstitial fibrosis (29).In the present study, renal expression of these fibrogenicgenes was determined. Consistent with increased renalfibrosis, ZDF rats showed increased renal expression ofcollagen I, collagen IV and fibronectin at the mRNAlevel (by 55, 32 and 32%, respectively) compared withZL-control (P50.05; Fig. 5A–C). SO treatment

normalized the expression of all three genes when admi-nistered to ZDF rats; interestingly, collagen I and IVmRNAs were also decreased in ZL rats after SO treat-ment. In cultured rat mesangial cells, SO (40mg/ml), MA(25 mM) and telmisartan (10 mM) all effectively suppressedthe angiotensin II-stimulated induction of collagen I(by 16–40%; all P50.05) and collagen IV (by 24–34%;P50.05) expression (Fig. 6A and B). Similarly, fibronec-tin expression at the mRNA and protein levels were alsonormalized (Fig. 6C and D).

AT1 and TGF-�1 Genes

TGF-b is a key factor in glomerulosclerosis and intersti-tial fibrosis. The synthesis and secretion of TGF-b1 areactivated by angiotensin II signaling through the AT1 (1)and leads in turn to enhance TGF-b1 signaling. AT1 andTGF-b1 mRNAs were both upregulated by 22 and 72%,respectively, in kidneys of ZDF rats; these increases wereattenuated by SO treatment to levels comparable withthose in ZL control rats (Fig. 7A and B). In accordwith these in vivo findings, angiotensin II treatment ofmesangial cells also significantly upregulated AT1mRNA expression by 27% (Fig. 7C) and TGF-b1mRNA and protein expression by 55 and 62% over con-trol, respectively (Fig. 7D and E). Treatment of cells with

B

ZL Con

A

ZL SO

DC

ZDF Con ZDF SO

Glu

mer

ular

are

a

0

1

2E

*

ZL

SOCon Con SO

ZDF

Figure 2. Renal tissue morphometry and glomerular cross-sectional areas 6 weeks after treatment with vehicle or SO extract in male ZL and ZDF

rats. The cross-sectional area was measured in glomeruli (hematoxylin-eosin staining). (A–D) Representative samples (magnification 200�); (E)

glomerular cross-sectional areas. All values are means� SEM (n¼ 5, each group). *P50.05.


SO (40mg/ml), MA (25mM) and telmisartan (10 mM)reversed these effects on AT1 and TGF-b1 and proteinlevel in mesangial cells (Fig. 7C–E).

Genes Implicated in Fibrinolysis

PAI-1 is the principal inhibitor of plasminogen activation(30) and uPA is a central activator of fibrinolysis. As partof the present study, the renal expression of these genesthat participate in the synthesis and degradation of theextracellular matrix was examined. Renal PAI-1 mRNAwas upregulated significantly by 43% in ZDF rats(Fig. 8A), although renal uPA mRNA expression wasunchanged (Fig. 8B), the ratio of renal uPA to PAI-1mRNAs was decreased by 29% (P50.05; Fig. 8C). SOtreatment of ZDF rats and, to a lesser extent, ZL ratssignificantly decreased PAI-1 mRNA by 46 and 15%,respectively, but was without effect on uPA mRNAexpression. Thus, there was a corresponding increase inthe uPA : PAI-1 ratio in SO-treated animals that wasmore pronounced in ZDF rats (88% increase) than inZL rats (18% increase) (both P50.05; Fig. 8C). Findingsin mesangial cells were consistent with in vivo findings.

Similar to telmisartan (10 mM) (decreased by 36.7%,P50.05), SO (40mg/ml) and MA (25 mM) significantlysuppressed the angiotensin II-mediated increase inPAI-1 mRNA (P50.05; Fig. 8D) and prevented theangiotensin II-induced decline in uPA mRNA expression

(P50.05; Fig. 8E) and the uPA : PAI-1 mRNA ratio(Fig. 8F).

Discussion

The present study clearly demonstrates that SO treatmentattenuates the renal interstitial fibrosis, glomerulosclerosis

and collagen deposition in ZDF rats, as revealed byvan Giesen-staining. Consistently, the elevated renalexpression of fibrogenic genes including collagen I, col-lagen IV and fibronectin in ZDF rats, was restituted bySO treatment. Moreover, SO treatment amelioratedthe increase in both insoluble and soluble collagens inthe kidneys of ZDF rats. These improvements wereaccompanied by the normalization of BUN and hypo-albuminemia, findings that are consistent with the atten-uation of diabetic nephropathy.

ZL ZDF

SO SOConCon

ZL ZDF

SO SOConCon

ZL ZDF

SO SOConCon

Inso

lubl

e co

llage

n (m

g/g)

* *Sa

lt S

olub

le c

olla

gen

(mg/

g)

Aci

d So

lubl

e co

llage

n (m

g/g)* *

*

* *

* *

0

2

4

6

8

0

1

2

3

4

5

0

1

2

3

4BA C

0

1

2

3

4

5

Pla

sma

UA

(m

g/dl

)

Pla

sma

Alb

umin

(g/

dl)

0

10

20

30P

lasm

a B

UN

(m

g/dl

)

0

1

2

3

4

ZL ZDF

SOCon SOCon

ZL ZDF

SO SOConCon

ZL ZDF

SO SOConCon

* *** * FED

Figure 3. Composition of renal collagens and serum biochemical parameters reflecting renal function in rats. The composition of collagen in kidney

tissue was analyzed using the colorimetric Sircol Assay. The collagen content in the kidney tissue was standardized to the total dry weight of renal

tissue. The biochemical parameters were quantified enzymatically. Renal salt-soluble collagen (A), acid-soluble collagen (B), insoluble collagen (C),

plasma albumin (D), BUN, (E) and plasma uric acid (F). All values are means� SEM (n¼ 5, each group). *P50.05.

eCAM 2009 7 of 13

0

0.5

1.0

1.5

2.0

Col

I/

-act

in

Col

IV

/-a

ctin

Fib

rone

ctin

/-a

ctin

0

0.5

1.0

1.5

2.0

0

0.5

1.0

1.5

2.0

FBN

-actin

ZL ZDF

SO SOConCon

ZL ZDF

SO SOConCon

ZL ZDF

SO SOConCon

* *

* ** *

**

Col IVCol I

-actin-actin

CBA

Figure 5. Renal expression of collagen (Col) I (A), Col IV (B), fibronectin (C) mRNAs in ZL and ZDF rats. Total RNA was extracted from renal

tissues using TRIzol. The relative levels of specific mRNAs were determined by RT–PCR. Results were normalized to b-actin. All values are

means� SEM (n¼ 5, each group). *P50.05.

SO10

Angiotensin IISaline

SO40TelConCon SO200

0.2

0.4

0.6M

TT

**

A 0.8

* *

*

dpm


AM04OS01OSAM TelConCon0

4000

8000

12000

16000

B 20000

**

* *

*

Figure 4. Angiotensin II-induced cell proliferation in rat-derived mesangial cells. SO (SO10: 10; SO20: 20 and SO40: 40 mg/ml), MA (25 mM) or

telmisartan (Tel, 10 mM) was added 1 h before mesangial cells were treated with angiotensin II (10�6M). Cell viability was determined by MTT assay

(A, n¼ 6, each group) and DNA synthesis was assessed by 3H-thymidine incorporation (B, n¼ 6, each group), as described in ‘Methods’ section. All

values are means� SEM. *P50.05.


Angiotensin II is a renal growth factor, inducing hyper-plasia and hypertrophy in a cell type-dependent fashion.Angiotensin II is a key factor in the inflammatory andfibrotic response in kidney diseases (31). This vasoactivepeptide activates mesangial and tubular cells and intersti-tial fibroblasts, increasing the expression and synthesis ofextracellular matrix proteins, such as collagen I andfibronectin (7,10,32,33); these effects are mediated byangiotensin II via the AT1 (10).Glucose has been shown to increase the production of

angiotensin II, thereby decreasing collagenase activityand increasing collagen deposition; the latter effect wasprevented in rat mesangial cells by the AT1 antagonistlosartan (34). In ZDF rats, AT1 mRNA and proteinexpression was increased along with the associatedglomerulopathy, upregulation of the extracellular matrixprotein fibronectin, renal inflammation and proteinuria;again, these effects were attenuated by losartan (28).Thus, angiotensin II signaling is stimulated in thisanimal model and leads to renal dysfunction and fibro-genesis. The present study demonstrated that the

overexpression of the AT1 gene in ZDF rat kidney wasnormalized by SO treatment. Furthermore, SO and itsmajor constituent MA suppressed a range of angiotensinII-induced cellular endpoints including hyperprolifera-tion, overexpression of collagen I, collagen IV and fibro-nectin mRNAs, and increased fibronectin protein level inisolated rat mesangial cells. Angiotensin II-dependentinduction of AT1 gene expression was also suppressed.These effects were similar to those elicited by the AT1antagonist telmisartan. Thus, the present findings assertthat SO modulates angiotensin II/AT1 signaling, which inturn attenuates cell proliferation and fibrogenesis in ZDFrat kidney, and that MA is a candidate active constituentpresent in SO extracts.TGF-b is a key factor in glomerulosclerosis and inter-

stitial fibrosis that acts directly, by stimulating synthesisof extracellular matrix components and reducing collage-nase production, and indirectly through other profibro-genic factors. Furthermore, TGF-b is important for theproliferation of intrarenal fibroblasts and the epithelial–mesenchymal transition through which tubular cells

COL I

-actin

COL IV

-actin

SO10Angiotensin IISaline

SO40 MATelConConSO10Angiotensin IISaline

SO40 MATelConCon

FBN

-actin

SO10


SO40 MATelConCon SO10


SO40 MATelConCon SO200

Fib

rone

ctin

/-a

ctin

0

Fib

rone

ctin

(ng/

ml)

0

1

2

0

1

2

Col

I/

-act

in

Col

IV

/-a

ctin

BA

DC

3 3

1

2

3

40

80

120

160

**

* **

*

**

**

** **

**

*

Figure 6. Expression of collagen (Col) I (A), Col IV (B), fibronectin (C) mRNAs and fibronectin protein (D) in rat mesangial cells. SO (SO10: 10;

SO20: 20 and SO40: 40 mg/ml), MA (25 mM) or telmisartan (Tel. 10 mM) was added 1 h before mesangial cells were treated with angiotensin II

(10�6M). Twenty-four hours later total RNA was extracted from the mesangial cells using TRIzol. The relative levels of specific mRNAs were

determined by RT–PCR. Results were normalized to b-actin. In a parallel experiment, the medium was collected after mesangial cells were treated

with the test agents in combination with angiotensin II. fibronectin protein was determined by commercial ELISA kit according to the manufac-

turer’s instructions. All values are means� SEM (n¼ 3, each group). *P50.05.

eCAM 2009 9 of 13

become fibroblasts (35). Numerous studies have demon-strated that TGF-b plays a pivotal role in experimentaldiabetic kidney disease and human diabetic nephropathy(32). Angiotensin II stimulates TGF-b expression in thekidney by several mechanisms and upregulates receptorsfor TGF-b (35). Apart from its effects on TGF-b1 secre-tion and activation, angiotensin II also directly enhancesTGF-b1 signaling (1). Glucose-induced TGF-b1 secretionwas suppressed by losartan in rat mesangial cells (34).Thus, angiotensin converting enzyme inhibitors andAT1 antagonists are two classes of drugs that have thepotential to inhibit angiotensin II-mediated TGF-bexpression (35). In the present study, SO treatmentreversed the increase in renal TGF-b1 expression inZDF rat kidney. Moreover, similar to telmisartan, bothSO and MA inhibited the increase in TGF-b secretionand angiotensin II-induced TGF-b1 gene expression inrat mesangial cells.

TGF-b1 activates the plasmin system by stimulatingPAI-1 gene expression (30). PAI-1 has a number ofimportant roles in pathological processes, including theinhibition of fibrinolysis, regulation of extracellularmatrix turnover and activation of proenzymes andlatent growth factors that promote tissue fibrosis andsclerosis (36). In progressive renal diseases, such as dia-betic nephropathy, PAI-1 has been identified as a criticalmediator of glomerulosclerosis and interstitial fibrosis(36,37). The altered uPA to PAI-1 ratio reflects achange from a profibrinolytic to an antifibrinolytic state(38). The shift toward the uPA-enriched profibrinolyticstate favors renal collagen degradation. The presentfinding that the increase in PAI-1 gene expression inZDF rat kidney was suppressed by SO treatment is con-sistent with the SO-mediated decrease in renal collagencontents. In mesangial cells, SO, MA and telmisartantreatment not only suppressed angiotensin II-induced

TGF- 1AT1

-actin-actin

ZL ZDFSO SOConCon

ZL ZDFSO SOConCon

30.2

0 0

0.5

1.0

1.5

AT

1/-a

ctin

* **

TG

F-

1/-a

ctin

1

2

* *

BA

AT1

-actin

TGF- 1

-actinEC

SO10


SO40

MA

Tel

Con

Con

SO10


SO40

MA

Tel

Con

Con

SO10


SO40

MA

Tel

Con

Con

SO20

2.0

TG

F-

(pg/

ml)

0

100

200

300

400

500

600

2.0

0

0.5

1.0

1.5

TG

F-

1/-a

ctin

0

0.5

1.0

1.5

AT

1/-a

ctin

D*

*

***

*

**

**

*

**

Figure 7. Renal expression of AT1 (A) and TGF-b1 (B) in ZL and ZDF rats, and levels of AT1 (C) and TGF-b1 (D) mRNAs, and TGF-b protein

(E) in rat mesangial cells. In cell culture, SO (SO10: 10; SO20: 20 and SO40: 40 mg/ml), MA (25mM) or telmisartan (Tel. 10 mM) was added 1 h prior

to treatment with angiotensin II (10�6M), followed by twenty-four hours incubation. Total RNA was extracted from renal tissues and the mesangial

cells using TRIzol, respectively. The relative levels of specific mRNAs were determined by RT–PCR. Results were normalized to b-actin. In a parallel

experiment, the medium was collected after mesangial cells were treated with the test agents in combination with angiotensin II. TGF-b protein level

was determined by commercial ELISA kit according to the manufacturer’s instructions. All values are means� SEM (n¼ 5, each group in vivo; n¼ 3,

each group in vitro). *P50.05.


overexpression of PAI-1, but also restituted uPA geneexpression that had been suppressed by angiotensin II;thus, the mRNA ratio of uPA to PAI-1 was normalized.These findings were consistent with analogous measure-ments in ZDF rats.Because glucose stimulates the production of angioten-

sin II (34), it is conceivable that the reported improve-ment of hyperglycemia by SO treatment (19) may alsocontribute to the attenuation of renal fibrosis in ZDFrats by decreasing renal angiotensin II. However, it isalso pertinent that recent studies have provided compel-ling evidence that other components of the renin–angiotensin–aldosterone system, including angiotensinIII, renin, and aldosterone also activate the TGF-bsystem (35). It is now important to investigate whetherother mechanisms apart from the angiotensin II/AT1signaling pathway may also contribute to the antifibro-genic activity of SO and its active constituent MA. In thisregard it is noteworthy that the SO extract also containsother active components, such as salacinol, kotalanol and

kotalagenin 16-acetate (17). The potential contributionsof these constituents to the biological actions of SO couldbe considered in future studies.

Conclusion

The present findings demonstrate for the first time that atraditional antidiabetic herbal medicine attenuates renalfibrosis in ZDF rats, at least in part, by suppressingangiotensin II/AT1 signaling, possibly as proposed inFig. 9. Furthermore, it now emerges that MA is a can-didate antifibrogenic agent. These studies provide newinsights into the potentially important renoprotectiveeffects and the underlying molecular mechanism ofaction of SO.

Funding

Pharmafood Institute, Kyoto, Japan.

PAI-1

-actin

ZL ZDFSO SOConCon

2.0

0

uPA

-actin

ZL ZDFSO SOConCon

ZL ZDFSO SOConCon

uPA

/PA

I-1

0

0.5

1.0

1.5

2.0

* *CBA

0.5

1.0

1.5

2.0P

AI-

1/-a

ctin

* *

uPA

/-a

ctin

0

0.5

1.0

1.5*

PAI-1

-actin

uPA

-actin

SO10

Saline

SO40

MA

Tel

Con

Con

SO10


SO40

MA

Tel

Con

Con

SO10


SO40

MA

Tel

Con

Con

Angiotensin II

1.6

uPA

/PA

I-1

0

0.4

0.8

1.2

2.0

0

0.5

1.0

1.5

PA

I-1/

-act

in

0

0.5

1.0

1.5

2.0

uPA

/-a

ctin

FED*

*

****

** **

***

*

Figure 8. Renal expression of PAI-1 (A), uPA (B) mRNAs, and the ratio of uPA to PAI-1 (C) in ZL and ZDF rats, and levels of PAI-1 (D) and uPA

(E) mRNAs, and the ratio of uPA to PAI-1 (F) in rat mesangial cells. In cell culture, SO (SO10: 10 and SO40: 40 mg/ml), MA (25 mM) or Tel (10 mM)

was added 1 h prior to treatment with angiotensin II (10�6M), followed by 24-h incubation. Total RNA was extracted from renal tissues and the

mesangial cells using TRIzol, respectively. The relative levels of specific mRNAs were determined by RT–PCR. Results were normalized to b-actin.All values are means� SEM (n¼ 5, each group in vivo; n¼ 3, each group in vitro). *P50.05.

eCAM 2009 11 of 13

References1. Wynn TA. Common and unique mechanisms regulate fibrosis in

various fibroproliferative diseases. J Clin Invest 2007;117:524–9.2. Gilbert RE, Cooper ME. The tubulointerstitium in progressive

diabetic kidney disease: more than an aftermath of glomerularinjury? Kidney Int 1999;56:1627–37.

3. Steffes MW, Osterby R, Chavers B, Mauer M. Mesangial expansionas a central mechanism for loss of kidney function in diabeticpatients. Diabetes 1984;38:1077–81.

4. Bader R, Bader H, Grund KE, Mackensen-Haen S, Christ H,Bohle A. Structure and function of the kidney in diabetic glomer-ulosclerosis. Correlations between morphological and functionalparameters. Pathol Res Pract 1980;167:204–16.

5. Lane PH, Steffes MW, Fioretto P, Mauer SM. Renal interstitialexpansion in insulin-dependent diabetes mellitus. Kidney Int1993;43:661–7.

6. Taft JL, Nolan CJ, Yeung SP, Hewitson TD, Martin FI. Clinicalhistological correlations of decline in renal function in diabeticpatients with proteinuria. Diabetes 1994;43:1046–51.

7. Mezzano SA, Ruiz-Ortega M, Egido J. Angiotensin II and renalfibrosis. Hypertension 2001;38:635–8.

8. Zhao W, Chen SS, Chen Y, Ahokas RA, Sun Y. Kidney fibrosis inhypertensive rats: role of oxidative stress. Am J Nephrol2008;28:548–54.

9. Nakamura T, Takahasi T, Fukui M, Ebihara I, Osada S,Tomino Y, Koide H. Enalapril attenuates increased gene expressionof extracellular matrix components in diabetic rats. J Am SocNephrol 1995;5:1492–7.

10. Ray PE, Bruggeman LA, Horikoshi A, Aguilera G, Klotman PE.Angiotensin II stimulates human fetal mesangial cell proliferationand fibronectin biosynthesis by binding to AT1 receptors. KidneyInt 1994;45:177–84.

11. Iwano M, Neilson EG. Mechanisms of tubulointerstitial fibrosis.Curr Opin Nephrol Hypertens 2004;13:279–84.

12. Patwardhan B, Warude D, Pushpangadan P, Bhatt N. Ayurvedaand traditional Chinese medicine: a comparative overview. eCAM2005;2:465–73.

13. Cooper EL. Ayurveda is embraced by eCAM. eCAM 2008;5:1–2.14. Cooper EL. Ayurveda and eCAM: a closer connection. eCAM

2008;5:121–2.15. Chandrasena JPC. The Chemistry and Pharmacology of Ceylon

and Indian Medicinal Plants. Colombo, Sri Lanka: H&C Press,1935.

16. Li Y, Huang TH, Yamahara J. Salacia root, a unique Ayurvedicmedicine, meets multiple targets in diabetes and obesity. Life Sci2008;82:1045–9.

17. McKay DL, Blumberg JB. A review of the bioactivity of SouthAfrican herbal teas: rooibos (Aspalathus linearis) and honeybush(Cyclopia intermedia). Phytother Res 2007;21:1–16.

18. Huang TH, Peng G, Li GQ, Yamahara J, Roufogalis BD, Li Y.Salacia oblonga root improves postprandial hyperlipidemia andhepatic steatosis in Zucker diabetic fatty rats: activation ofPPAR-a. Toxicol Appl Pharmacol 2006;211:225–35.

19. Li Y, Peng G, Li Q, Wen S, Huang TH, Roufogalis BD,Yamahara J. Salacia oblonga improves cardiac fibrosis and inhibitspostprandial hyperglycemia in obese Zucker rats. Life Sci2004;75:1735–46.

20. Huang TH, Yang Q, Harada M, Uberai J, Radford J, Li GQ,Yamahara J, Roufogalis BD, Li Y. Salacia oblonga root improvescardiac lipid metabolism in Zucker diabetic fatty rats: modulationof cardiac PPAR-alpha-mediated transcription of fatty acid meta-bolic genes. Toxicol Appl Pharmacol 2006;210:78–85.

21. Huang TH, He L, Qin Q, Yang Q, Peng G, Harada M, Qi Y,Yamahara J, Roufogalis BD, Li Y. Salacia oblonga root decreasescardiac hypertrophy in Zucker diabetic fatty rats: inhibition of car-diac expression of angiotensin II type 1 receptor. Diab Obes Metab2008;10:574–85.

22. Yoshikawa M, Nishida N, Shimoda H, Takada M, Kawahara Y,Matsuda H. Polyphenol constituents from Salacia species:quantitative analysis of mangiferin with alpha-glucosidase andaldose reductase inhibitory activities. Yakugaku Zasshi 2001;121:371–8.

23. Li Y, Kishimoto I, Saito Y, Harada M, Kuwahara K, Izumi T,Takahashi N, Kawakami R, Tanimoto K, Nakagawa Y,Nakanishi M, Adachi Y, Garbers DL, Fukamizu A, Nakao K.Guanylyl cyclase-A inhibits angiotensin II type 1A receptor-mediated cardiac remodeling, an endogenous protection mechanismin the heart. Circulation 2002;106:1722–8.

24. Singh R, Alavi N, Singh AK, Leehey DJ. Role of angiotensin II inglucose-induced inhibition of mesangial matrix degradation.Diabetes 1999;48:2066–73.

25. Chouinard RF, Meek RL, Cooney SK, Tuttle KR. Effects of aminoacids and glucose on mesangial cell aminopeptidase a and angioten-sin receptors. Kidney Int 2002;61:S106–9.

26. Tsuchiya T, Shimizu H, Shimomura K, Mori M. Troglitazoneinhibits isolated cell proliferation, and induces apoptosis in isolatedrat mesangial cells. Am J Nephrol 2003;23:222–8.

27. Li Y, Qi Y, Kim MS, Xu KZ, Huang TH, Rong X, Murray M,Yamahara J. Increased renal collagen cross-linking and lipid accu-mulation in nephropathy of Zucker diabetic fatty rats. Diab MetabRes 2008;24:498–506.

28. Xu ZG, Lanting L, Vaziri ND, Li Z, Sepassi L, Rodriguez-Iturbe B,Natarajan R. Upregulation of angiotensin II type 1 receptor,inflammatory mediators, and enzymes of arachidonate metabolismin obese Zucker rat kidney reversal by angiotensin II type 1 receptorblockade. Circulation 2005;111:1962–9.

29. Remuzzi G, Benigni A, Remuzzi A. Mechanisms of progression andregression of renal lesions of chronic nephropathies and diabetes.J Clin Invest 2006;116:288–96.

30. Lund LR, Riccio A, Andreasen PA, Nielsen LS, Kristensen P,Laiho M, Saksela O, Blasi F, Danø K. Transforming growthfactor-beta is a strong and fast acting positive regulator of thelevel of type-1 plasminogen-activator inhibitor messenger-RNAinWi-38 human lung fibroblasts. EMBO J 1987;6:1281–6.

31. Ruiz-Ortega M, Ruperez M, Esteban V, Rodrıguez-Vita J, Sanchez-Lopez E, Carvajal G, Egido J. Angiotensin II: a key factor in theinflammatory and fibrotic response in kidney diseases. Nephrol DialTransplant 2006;21:16–20.

DIABETES MELLITUS

DIABETIC NEPHROPATHY

Matrix Protein AccumulationGlomeruloscerosis and

Tubulointerstitail Fibrosis

TGF- b 1

Collagen I, IV and Fibronectin

AT1signaling

PAI-1

Angiotensin II

AT2signaling?

SO(MA, etc.)

?

Figure 9. Proposed model showing SO- and MA-dependent effects on

the angiotensin II/AT1 signaling pathway and fibrogenic mediators in

diabetic rat kidneys.


32. Wolf G, Haberstroh U, Nielson EG. Angiotensin II stimulates theproliferation and biosynthesis of type I collagen in cultured murinemesangial cells. Am J Pathol 1992;140:95–107.

33. Kagami S, Border WA, Miller DE, Noble NA. Angiotensin IIstimulated extracellular matrix protein synthesis through inductionof transforming growth factor-b expression in rat glomerular cells.J Clin Invest 1994;93:2431–7.

34. Singh R, Alavi N, Singh AK, Leehey DJ. Role of angiotensin II inglucose-induced inhibition of mesangial matrix degradation.Diabetes 1999;48:2066–73.

35. Wolf G. Renal injury due to renin-angiotensin-aldosterone systemactivation of the transforming growth factor-beta pathway. KidneyInt 2006;70:1914–19.

36. Eddy AA. Plasminogen activator inhibitor-1 and the kidney. Am JPhysiol Renal Physiol 2002;283:F209–20.

37. Fogo AB. Renal fibrosis: not just PAI-1 in the sky. J Clin Invest2003;112:326–8.

38. Deatrick KB, Eliason JL, Lynch EM, Moore AJ, Dewyer NA,Varma MR, et al. Vein wall remodeling after deep vein thrombosisinvolves matrix metalloproteinases and late fibrosis in a mousemodel. J Vasc Surg 2005;42:140–8.

Received February 4, 2009; accepted June 26, 2009

eCAM 2009 13 of 13

the ayurvedic medicine salacia oblonga attenuates diabetic...

Documents