Indmn J PhYSIOI PhnrmaCQI 2002; 46 (2); 209-211

BIODEGRADABLE MlCROSPHERES OF CURCUMIN FORTREATMENT OF INFLAMMATION

VIRENDER KUMAR, SHAlLA ANGELA LEWIS, SRlNIVAS MUTALIK.DINESH B. SHENOY, VENKATESH AND N. UDUPA*

Department of Pharmaceutics,College of Pharmaceutical Sciences,Manipa.l - 576 119

(Received on August 7, 2000)

Abstract: Curcumin, a natural constituent of CurCllma Irmgfl (turmeric,CAS 458-37-7) was formulated as prolongcd release biodeg-radablemicrospheres for treatment of inflammation. Natural biodegradablepolymers, lI11mely, bovine serum albumin and chilosnn were used toencapsulate curcumin to form a depot forming drug delivery system.Microsphcres were prepared hy cmuhlion-solvent evaporation methodcoupled with chemical cross-linking of the natural polymers. Curcumincould be encapsulated into the biodegradable cnrriers upto an extcnt of19.49 and 39_66% respectively with albumin and chitosan. Differentdrug:polymer ratios did not affect the mean particle size or particle sizedistribution significantly. However, the concentration of the crosslin kingagent had remnrkable influcnce on the drug release. In-vitro release studiesindicated n biphasic drug release puttern, characterized by a typical burst­effect followed hy ;l slow relcalS!! which continued for several days.Evnluallon or untiinnllmmfltol'Y activity using Freund's adjuvant inducedarthritic lIlodel ill Wistar rats revenleu significant difference between boththe formulutions, albumin microspheres and chitosan microspheres liS wellas ng~linst contrul.

It wns evident from the present study that the cun;umin biodegrudnblemicrospheres could be successfully employed as prolonged release drugdelivery system for better therapeutic management of inflammation ascompared to oral or subcutaneous route.

Key words: curcuminchitosilll

biodegradnblemicrospheres

albuminantiinflammatory

INTRODUCTION

Free radical reactions participatein inflammatory processes by inducingchemotactic activity, by virtue of theirrelations with arachidonic acid metabolism

·Corresponding Author

and by inhibiting the serum inhibitor ofleukocytic protcases (1). Curcumin (difcruloylmcthanc), an important constituent ofCllrCllma Longa (turmeric CAS 458-37-7), is apotent antioxidant capable of scavengingoxygen fTee radicals like superoxide anion

210 KumAr et al

and hydroxyl radicals and the powerfulantiinflammatory activity of curcumin is aresult of its antiox.idant property (2).

During the past decade. considerableprogress has been made in the field ofcontrolled drug delivery. Microspheres, asused in drug delivery are discrete, micrometer­sized spherical particles containing anentrapped drug. They can be prepared from avariety of carrier materials of both naturaland synthetic origin. The main advantage ofcontrolled drug delivery systems seems to bemaintenance of therapeutically optimumdrug concentrations in the plasma throughzero-order release without significantfluctuations and eliminating tbe need forfrequent single dose administrations.Parenteral administration of a drug ina microencapsulated formulation into asubcutaneous tissue results in the formationof a depot at the site of injection. This depotacts as a drug reservoir which releases thedrug continuously at a rate determined to alarge extent by the characteristics of theformulation. lending to the sustainedabsorption of the drug thereby reducing1l1hel'ent disadvantages of cOlwcntionnlparentel'al drug administration.

In the prcsent invesl.igation, curcumin waschosen as a model nnti-inf1ammntory agent,as much emphasis is givcn to drugs of naturalorigin in recent years. Poorly water solublcand photoscnsitive nature of the curcuminmakes it worthwhile formulating as noveldelivery system. We have also chosen thebiocompntiblc polymers of natural origin(bovine serum albumin - a naturally occurringprotein and chitosan - a naturally availablepolysaccharide) to facilitate the usage ofnaturally available resources in drug

Indian J Phyaiol Pharmacol 2002: 46(21

delivery research. Both the chosen polymershave a long history in biomedical researchand variety of delivery systems have beendesigned employing these biopolymers(3-12).

METHODS

Curcumin was purchased from LobaChemie, Chenoai; bovine serum albumin(BSA) and Freund's complete adjuvant fromSigma Chemicals Co, USA. Chitosan was agift sample from Fisheries ResearchInstitute, Trivandrum. All the otherchemicals and reagents used were ofanalytical grade. For pharmacodynamicstudies, adult male Wistar rats (200-215 g).were obtained from Central Animal HousingFacility, Department of Pharmacology,KMC, Manipal.

Prepllrlltion of curcumin louded cbitosan

micro.pheres

Chitosnn microspheres were preparedas described earlier (10) with slightmodification. A 2% wlv solution of chitosanwas prepared in 5% wtv citric acid aqueoussolution with the assistance of heat. Aknown quantity of curcumin W3~ dispersedin polymer solution with the aid of nhomogenizer. This suspension was droppedinto the liquid paroffin bulk oil phasecontaining 1% vtv Span 80 under continuousstirring using l\ homogenizer at 2000 rpmto form a water-in-oil (w/o) emulsion. Fourml of the crosslinking agent (glutersldehydesaturated toluene, T I1Il ) was added drop wiseand the emulsion was 3gitated at constantrotation (1000-1500 rpm) for specified period(4-5 hours) at ambient temperature. Themicrospheres formed were separated from

Indinn J Physiol Pharm<lcol 2002; 46(2)

the reaction mixture by centrifugation (2500rpm), washed several times with n-hexaneuntil the oil phase was removed completelyand finally dried at room temperature in adesiccater. Three D:P (drug:polymer) ratiossuch as 1:4, 1:2, aod 1:1 were employed toprepare different formulations C1, C2 andC3 respectively.

Pl"cpnrntion or cUl"cumin londed BSA

microspheres

A techni4ue for manufacturing BSAmicrospheres has been developed bymodification of the emulsion cross-linkingmethod described by Gupta et aJ (4). BSA(200 mg) was dissolved in 1 ml of distilledwater at room temperature and a knownquantity of drug was dispersed bybomogenization. The dispersion wastransferred into a beaker containing liquidparaffin with 1% v/v Span 80. Stirring wasmaintained at a steady speed of 2000 rpmand 1 011 of Tns was added drop wise.Agitation was continued at roomtemperature until cross·linking completed(4-5 hours). The suspension was centrifugedand microspheres were collected; washedseveral times with n-hexane until the oilphase was removed completely and finallydried at room temperature in a desiccater.Three D:P (drug:polymer) ratios such as 1:4,1:2 and 1:1 were employed to preparedifferent formulations AI, A2 and A3respectively.

Physicochemical characteri:tation

The size of the microspheres wasdetermined by light microscopy usingcnlibrated eye piece micrometer. Three

Curcumin Microspheres ror Inflammation 211

hundred microspheres were counted andmean diameter and the size distributioncharacteristics were computed.

To determine the encapsulationefficiency and drug content, 100 mg ofcurcumin loaded BSA microspheres weredigested in 10 ml O.IN Hel for 48 hours ina shaker bath. The suspension wasthen transferred into a separating funneland 10 ml octanol was added to selectivelyextract curcumin. The octanol layer wasseparated, diluted suitably and curcuminconcentration was determined by measuringthe absorbance at 434 nrn againstsuitable blank prepared using microsphereswithout curcumin. Similarly, curcuminloaded chitosan micfospheres weredigested in 10 011 of 5% w/v citric acidsolution; curcumin was extracted into10 ml octanol and drug concentrationwas estimated spectrophotometrically at434 nm_

JII-vitru release studies

Microspheres containing known amountof the drug were taken in 5 viaJs containing10 ml of octanol maintained at a constantshaking' of 60 oscillations per minute andambient temperature. Samples werewithdrawn at periodic time intervals withreplacement of fresh release medium. Thl!drug concentration in the samples wasmeasured using UV-spectrophotometer CUV240, Graphicord, Shimadzu, Tokyo, Japan)at 434 nm after suitable dilution.

I,I-vivo studies

Pharmacodynamic evaluation of the drug

212 Kumar el al

in the formulation was carried out usingFreund's adjuvant induced arthritic model.Animals were divided in five groups of sixeach. Chronic inflammatory condition wasinduced in one of the hind paws of adult maleWistar rats (200-215 g) by an intradermalinjection (0.1 mt) of Freund's completeadjuvant into subplantar surface. The edemaproduced was regularly monitored andassessed by measuring the volume of theinjected paw using a mercury plethysmometer.Treatment schedule is briven in table 2. Oneighth day of inducing edema, paw volumeswere recorded as initial values and curcuminloaded chitosan and BSA microsphereswere given as single subcutaneous (s.c)injections after suspending in 0.3% w/vsodium carboxymethyl cellulose. Pawvolumes were recorded on 10th day post­treatment and antiinflammatory activitywas expressed as percentage inhibitionof inflammation (primary response) (13).Statistical analysis was performed by student'st-tesl.

Indilln J PhY.IIiol Pharmllcol 2002: 46(2)

RESULTS

Curcumin loaded BSA and chitosnnmicrospheres exhibited good morphologicalcharacteristics with smooth boundary andmostly spherical ones as observed by lightmicroscopy. However, occasional deformed anddiscrete particles were seen in both cases. Thesize distribution analysis revealed highestfrequency of the microspheres in the vicinityof 20·30 ~m. Chitosan microspheres exhibitedlarger diameter than their BSA counterpartsat corresponding D:P (drug and polymer)ratios (Table I). The percentage yield of theproduct ranged between 55 to 80% and wasalways higher for BSA than for chitosan. Fromthe table I, it is apparent that the content ofcurcumin in the chitosan microspheres issignificantly lower than in the equivalent BSAmicrospheres. BSA microspheres exhibitedhigher drug encapsulation efficiency rangingfrom 69 to 80% with chitosao microspheres inthe range of 30 to 40% at comparable D:Pratios.

TABLE I: PhY!licochemicsl characterization of curcumin loaded micro!lpheres.

Fvrmula/iall D:P ra/iD r:DcapsuJarionefficien.cy (%)

Al 1;4 79.49

A2 1:2 69.68

A3 I: 1 72.77

C1 1:4 32.96

C2 1:2 30.17

C3 1: 1 39.66

A = Albumin microsphere.ll. C = Chito!lsn microsphere.llD:P Ratio = Drug: Polymer ratio

Mean particlesize (pm)

22.83

23.92

28.51

29.23

25.01

33.33

Drug con/eM per100 mg micraspheres

15.89

23.21

36.35

6.59

10.05

19.83

Indian J Physiol Pharmacol 2002; 46(2) CUTcumin Microspberes Ior lnfiammation 213

In-vitro relea e tudi 100

_-1:1-.-1·2·· ... -1:4

2 : Til vitro Telease of curcumin from chitosl'lnmicrospheres.

Tn.-vitro release studies from chitosnnmicrospheres was done by placing microspher 5

containing k.nown amount of drug ill 5 vialscontaining 10 ml of oc anol maintajn d t aconstant shak.ing of 60 oscillations per minuteand ambient temperature. Values are expressedas mean ± S.D. of three trials.

150

Tlmli In haura

'00

9D

80

Fig.

...! 10

e fill

1 50.,.. '0~

:!i~ 30~u 20

10

The relea e pattern of cUTcurnin wasob erv d a biphasic (Fig. 1 and 2)characterized by an initial burst effect followedby a slow releas . A econdary bUT wasobserved in case of 1:1 ratio toward thtail end from albumin microspheres while 00

secondary burst effect was observed withchito an microspheres with 1:1 D:P ratio. Thetrend of drug release pattern was comparablefor both the polymer however, chitosanmicro pheres relea ed the loaded curcumin at

lower rate compared to the B Acounterparts. D:P ra io had a remarkableinfluence on drug releas with high rpolymer ratios retarding the drug relea e. Weexamined the polymers at single corsslinkingdensity as per their mol cular weights (BSA-47,000 Da; chitosan - 1 to 3 x 105 Da). 0

1: In vitro r lease of curcumin from B Amicrospheres.

In·vitro release studies from albuminmicrospheres was done by placing microspherescontaining known amount of drug in 5 vialscontaining 10 mJ of octanol maintained at aconstant shaking of 60 oscillations per minuteand ambient tempera ure. Values are expressedas mean ± S.D. of three l..rillls.

HlO

'" 90": 60

'"e 70

2 80

'" 50.0.. 40.~;; 30'3E 20~

I.) 10

Fig.

50 100 1~ 200

TIme in hours

__ 1;1 ...... 1:2 __ 1:4

250 300

flakes or aggregate were observed in bothca e ruling out possible bridging between themicrospheres formed and th products formedwere free flowing.

The result of pharmacodynamic activityof curcumin formulations in both th carrie-care depicted in the Table II. When curcuminwas given orally its anti-inflammatoryresponse was very minimal. Maximalinhibition of inflammatory response wasobserved in case of B A formula ionswhich wa significantly (P<0.05) highercompared to chito an formula ions as well ascurcumin given as injection. There wano ignifican duference observed betw enthe groups treated with curcumin 10 dedchitosan micro phere and plain curcumininj ction.

214 Kumar et al Indian J PhYlliol Pharmacol 2002; 46{2)

TABLE II: Anti-inflammatory activity of eurumin loaded mierollphel",·g.

Grflup TrCfllment Dl1se mglkg Change i,l paw v(J{llme (mL) % inhibition

Control (0.3% CMC Na) -0.136 -15.1341

2 Curumin (s.c. injection) 100 0.163 17.65

3 Curumin loral) 100 0.096 11.32

4 Albumin mierosphllrcs (A3) 100 0.396 35.86-(s.c. injection)

5 Chitosnn micrOllJlheres (C3) 100 0.17 15.93ls.c. injection)

Number of animals per group::: 6 rI\~

-Significant difference (P<O.051 compared to curcumin loaded microspheres and plain curcumin injectionCMC.Nll .. Carboxy melhyl cellulose sodium

DISCUSSION

Curcumin could be microencapsulated intoBSA :lnd chitosan even though the extent ofencapsulation varied significantly. Theproducts formed, when observed undermicroscope, exhibited discrete particles withsome microspheres showing deformed surfacelboundary and a wide size djstribution range(from 6 to 90 11m). Occasionally, microspheresobtained from natural polymers are notperfectly spherical because of variations inmolecular weight and other chemicalproperties of the polymer itself. Also, becaul:ieof the high viscosity of the liquid paraffin,the microspheres formed, experience greaterresistance (hence get deformed) while theinternal phase evaporates leaving behindspherical walls of the polymer within whichthe drug is encapsulated. As the internalphase is aqueous in both the cases, its rate ofevaporation is also very slow giving rise todeformed boundaries. Wide size distributionrange is also because of the higher viscosityof the external oil phase and higher rate of

shear. Discrete patches of precipitated drugon the surface of the microspheres werediscernible by their vivid yellow color underlight microscope.

We have used 5% w/v citric acid aqueoussolution for solubilizing chitosan as it formssoluble salt with acids. Even though most ofthe reported methods used for preparation ofchitosan microspheres employ acetic acid asthe solvent, its use in biomedicine is limitedby its toxicity aspect. Citric acid enjoys bettertolerance and safety index compared to othermineral acids, hence it was taken up. Theviscosity measurements of the solution ofchitosan in citric acid (120 cps for 1% w/vsolution in 5% w/v citric acid) showed optimumfluidity compared to solution in acetic acid(175 cps for 1% w/v solution in 6% v/v aceticacid). The higher viscosity of acetic acidsolution of the polymer reduced the ease ofhandling the solution for formulationprocedure. The citric acid as a solvent offeredoptimum viscosity for suspending the drug forpreparation of w/a emulsion and also

Indian J PhysioJ Pharmacol 2002: 46(2)

facilitated the formation of smooth walledmicrocapsules. We opted for chemicalcrosslinking procedure as control of thedegree of crossJjnking is easily variableand reproducible from batch to batch.Gluteraldehyde saturated toluene (TGs) wasemployed rather than aqueous solution ofgluteraldehyde as the former offer uniformdistribution of the crosslinking agentthroughout the continuous oil phase. Theaqueous solution of gluteraldehyde, dueto non·miscibility in the oil phase, resultsin non-uniformly crosslinked product leadingto erratic entrapment and drug releasekinetics. The curcumin was loaded into BSAmicrospheres as a concentrated suspensionranging from 5 to 20% w/v where as sameloaded into chitosan microspheres in the rangeof 0.5 to 2% w/v. Due to this discrepancy,there was significant difference in theencapsulation efficiency between the twoformulations. Also, the total polymer availablefor encapsulation was much less for chitosanformulation (1% w/v) compared to BSAformulation (20% w/v). It also could beassumed that, BSA having many reactivegroups distributed throughout the surface,may get involved in weak hydrophobicinteractions with curcumin, which could beone of the reasons for higher degree ofencapsulation.

During the in-vitro release studies, theinitial burst effect corresponds to the releaseof the drug located on or ncar the surface ofthe microspheres or release of poorlyentrapped drug. The washings with n-hexanedid not result in removal of the adsorbed drugparticles. The subsequent slow release periodis due to the release medium being diffused

Curcumin Microspheres fnr Inflammation 215

into the polymer matrix, whereby the drugdiffuses out of the microspheres. Thesecondary burst of albumin microspheres with1:} D:P ratio indicated the erosion of thepolymeric matrix due to continuous efflux ofthe drug followed by dissolution of the polymerinto the release medium. The rate of drugrelease i.ncreased with decrease in the polymerratio in the formulation. The chitosanformulations released drug pay load muchslower than the BSA counterparts because ofhigher molecular weight of the former andalso higher degree of crosslinking density(6.66% v/v for chitooan against 1.66% v/vfor BSA). Insignificant intra-batch and inter­batch variations with respect to in-vitro drugrelease pattern indicate uniform degree ofcrosslinking throughout the encapsulationprocedure.

We selected the formulation with highestrate of drug release (D:P =1:1) so as to projectthe pharmacodynamic profile at initial stageitself. We compared the activity with curcumingiven orally and as subcutaneous (s.c.)injection by suspending in 0.3% w/v sodiumcarboxymethyl cellulose at a dose level of 100mg/kg body weight, which is the ED50 ofcurcumin (14). The route of administrationwas s.c. which favours prolonged drug releasethan intravenous or intramuscular routesbecause of diffused vascularity of adiposetissue and is ideal for a depot formingformulation. The reasons for significantly(P<O.05) increased anti-inflammatory responsein case of BSA formulations can be explainedon the basis of total amount of drug releaseduring the period of study. As predicted fromthe in-vitro drug release data, 84% of theencapsulated curcumin was released from BSA

216 Kumar e~ al

formulation against 36% in case of chitosanmicrospheres. This percentage would be muchhigher in-vivo due to presence of enzymes andother tissue fluids at the injection site whichfavour faster extraction of the drug from theformulation. Oral route, which is not apreferred route of administration forcurcumin (as it is least absorbed from GIT)offered minimal protection against Freund'sadjuvant induced rat paw edema. The s.c.injection of curcumin also producedinsignificant reduction in inflammationthough the reduction in paw volume initiallywas greater than polymeric microsphereswhich could be due to greater amount ofcurcumin available in plain injection. Incase of microspheres, though the initialreduction (first two days post-treatment) ofpaw volume was not significant, the totalinhibition of inOammatory process at theend of 10 days was much higher than plaininjection. BSA microspheres, due to smallerdiameter, offered greater surface for drugextraction from the depot and hence greaterpharmacodynamic activity. Total amount ofcurcumin available was extended over thestudy period resulting in improved efficacy.With chitosan microspheres, the rate of

Indian J Physiol Pharmacol 2002; 46(21

drug release was slower as explained earlierand total amount available at the end of 10days could be comparable to that of plaininjection. Improved anti-inflammatoryactivity may be expected from chitosanformulation if the duration of study isfurther increased. We took 10th day post·treatment as the end point as extrapolatedfrom the in-vitro kinetics and formulationparameters.

Curcumin is a potent antiinOammatoryagent acting through scavenging oxygen freeradicals produced as a consequence ofreactions occurring due to inflammationprocess. Inhibition of pathological freeradical reactions has been supposed toaccount for part of the action of the knownantiinflammatory agents including NSAIDs.We have reported improved radioprotectionactivity for microencapsulated curcuminin synthetic biodegradable polymer earlier(I5). Incorporation of curcumin intonatural biodegradable polymers not onlyimproves the pharmacodynamic efficacybut also reduces the cost of therapy due toreduction in total dosing and lesser cost ofsuch polymers compared to synthetic ones.

REFERENCES

1. Feher- J, ClIomos G, Ver-eckei A. (ed) Free radicalreac~ions in medicine. Edn. L Springer--Verlag,Berlin. 1987:49-52.

2. Kuehamly R, Rao MNA. Oxygen radical scavengingactivity or cur-cumin. 1nt J PhlJrm 1990; 58:237-240.

3. Kramer PA. Albumin micr-ospheres ns vehicles rorachievinllspecificity in Drug deliver-y. J Phorm Sci1974; 63: 1646-1647.

4. Gupta PK. Hung CT, Perrier DG. Albumin

microspher-es. I: Release characteristics oradr-ioamycin. 1111 J Phnrm 1986;33:137-146.

5. Lewis CA, Field WN. Hayes K. Alper HO. The ulleor albumin microspheres in the treatment orcarrageenan-induced inflammation in the rnl.J Phnrm Pharmneol 1992; 44: 271-274.

6. Gi:eurarson S, Bechgaard E. Insulin-carryingmicrospheres: /!1-lJitro Studies. Chcm Pharm Bull1991; 39: 1892-1893.

7. Akbugn J. Durma:e G. Preparation and evaluation

Indian J Physiol Pharmacnl 2002; 46(2)

of cross-linked chitosan microsJlheres containingfurosemide. Int J Pharm 1994; Ill: 211-222.

8. Murata Y, Maeda T, Miyamoto E. Kawashima S.PrcparnLion of chiLosan . reinforced alginllLe gelbends·effects ofchitosEin on gel matrix erosion. hll

J Phtum 1963; 96: 139-145.

9. Thllnoo BC, Sunny Me, JaYllkrishnan A.Crosslinked chitosan microspheres : Preparationand evaluation 8S a matrix for the controlledrelease of pharmaceuticals. J Phurm Pharmacal1992; 44: 283-286.

10. Li VP, Machidll TV, Sannan T, Nagai T.PreparaLion of Chlto8an micro8phere8 containingfluorouracil using 'drug-in.oil' method andits release characteristics. STP Pharm Sci 199J; 1:363-368.

II. Akbuga J, Bergisadi N. 5-Pluorouracil-lotlded

Curcumin Microspheres for Inflammation 217

chitosan mjcro8pheres: preparfltion and releasechnracteristics. J Microencapsul 1996; 13: 161-168.

12. Singh UV, Udupa N. Methotre.xate. loaded chitosanand chitin microspheres _ in-vitro characterizationand pharmncokinetics in mice bearing Ehrlichascites carcinoma. J Microetlcopsul 1998; 15: 581­594.

13. Turner RA. ted) Screening Methodll inPharmacology. Edn. l. Acndemic prells, New Vork,1965; 152-162.

14. Srimal Re, Khnnna NM, Dhavan BN. A preliminnryreport on the antiinflammatory activity ofcurcumin. Ind J Pharmacol 1971; 3: 10.

15. D'8ou:r:a R, Singh UV, Kamath R, Uma Devi P,Udupa N. Curcumin loaded PLGA microsilhereswith enhanced radioprotection in mice. Pharm Sci1997; 3: 439-441.