dissipative particle dynamics simulation of ginsenoside ro...

Delivered by Publishing Technology to INST FOR APPLIED SCIENCE amp ENGINEERING RESEARCHACADEMIA SINICAIP 140123241242 On Sat 05 Jul 2014 044816

Copyright American Scientific Publishers

RESEARCH

ARTIC

LE

Copyright copy 2014 American Scientific PublishersAll rights reservedPrinted in the United States of America

Journal ofComputational and Theoretical Nanoscience

Vol 11 2046ndash2054 2014

Dissipative Particle Dynamics Simulation of GinsenosideRo Vesicular Solubilization Systems

Xingxing Dai1 dagger Xinyuan Shi134lowast dagger Haiou Ding2 Qianqian Yin2 and Yanjiang Qiao134lowast1Beijing University of Chinese Medicine Beijing 100102 China

2School of Traditional Chinese Medicine Capital Medical University Beijing 100069 China3Key Laboratory of TCM-information Engineer of State Administration of TCM Beijing 100102 China

4Beijing Key Laboratory for Basic and Development Research on Chinese Medicine Beijing 100102 China

Ginsenoside Ro (Ro) a natural biosurfactant derived from ginseng has been proven to form vesi-cles in aqueous solutions that enhance the solubility of the insoluble compounds With the intentionof expanding the applications of Ro in the pharmaceutical industry we here examined the influenceof drug additives on the solubilization of Ro vesicles The effects of the compatibility between eachdrug and the Ro molecules and of the length of the hydrophobic and hydrophilic structure of drugmolecules on the solubilizing capacity and solubilization site were here studied using dissipativeparticle dynamics (DPD) The results showed that the simple hydrophobic drugs lacking hydrophilicgroups were mainly located on the hydrophobic layer of the vesicle and the drugs containing bothhydrophobic and hydrophilic structures were located on the palisade layer There was no single sim-ple relationship between solubilizing capacity and drug properties The solubility parameters of thehydrophobic structure molecular size changes in hydrophobic and hydrophilic properties attributedto changes in the length of the hydrophobic and hydrophilic structures and the solubilization sitewere found to affect the molar solubilization ratio (MSR) of the Ro vesicles These results provideinsight into the vesiclar solubilization system formed by saponin and may serve as guidances forthe further development and application of Ro and other saponins

Keywords Biosurfactant Ginsenoside Ro Solubilization Vesicle Mesoscopic Simulation

1 INTRODUCTIONAlthough many drugs have been shown to have strong bio-logical activity in vitro many cannot be used in clinicaltreatments due to poor solubility1ndash3 For this reason solu-bilization is important for the preparation and absorptionof insoluble drugs Saponins are a class of biosurfactantsthey are mainly derived from plants They are glycosidescontaining a hydrophobic triterpene or steroid aglyconeand one or more hydrophilic sugar chains As solubilizersthey offer advantages over synthetic surfactants in that theyare derived from renewable sources and are biodegradablehighly active and either non-toxic4 Therefore saponinshave been proposed as potential safe and effective adju-vants suitable for enhancing the absorption and dissolutionof pharmacologically active substances and drugs throughsolubilization56

Ginseng (roots of Panax ginseng C A Mey) is a pop-ular medicinal plant Its main active compounds are a

lowastAuthors to whom correspondence should be addresseddaggerThese two authors contributed equally to this work

complex mixture of saponins which have a variety ofbioactivities including anti-inflammatory antioxidant andanticancer effects78 Over 15000 of the formulae in theTraditional Chinese Medicine (TCM) Database System(httpcoworkcintcmcomenginewindex1jsp) containginseng9 In addition to its own pharmacodynamic efficacyginseng can also enhance the efficacy of other drugs Thesolubilizing effect of saponins on insoluble compoundshas been considered one possible mechanism of ginsengrsquossynergistic effects For example ginsenoside Ro (Ro) hasbeen shown to markedly increase the solubility of saikos-aponin a (SSa) which is the active ingredient in RadixBupleuri (roots of Bupleurum Chinese DC and Bupleurumscorzonerifolium Wild) which is sparingly soluble inwater10 Our previous studies have proven that Ro can formvesicles in aqueous solutions and solubilize SSa moleculesin palisade layers It can also form mixed vesicles with themolecules to enhance the solubility of SSa1112

Studies of solubilization mechanisms and laws govern-ing interactions between insoluble drugs and saponin vesi-cle are of great significance More information may expand

2046 J Comput Theor Nanosci 2014 Vol 11 No 9 1546-19552014112046009 doi101166jctn20143605



RESEARCH

ARTIC

LE

Dai et al Dissipative Particle Dynamics Simulation of Ginsenoside Ro Vesicular Solubilization Systems

the applications of ginsenoside and similar saponins inmedical situations and the pharmacological industry How-ever in these studies standard experimental methods canbe very costly in both time and money and it can bedifficult to illustrate the solubilization mechanisms andthe exact internal structures of the vesicles Fortunatelypowerful computer simulations can render the researcheshighly efficient Many theories have been developed tostudy the organization of quantum matter13ndash19 Dissipativeparticle dynamics (DPD) is one of the mesoscopic sim-ulation methods used to study drug loading systems20ndash28

This is a particle-based method which coarse-grains thefamiliar atomistic representation of the molecules to gainorders of magnitude in both length and time scale rela-tive to traditional atomistic scale29 DPD allows particlesto interact through a simple-wise potential and accuratelycapture their hydrodynamic behavior and the underlyinginteractions3031 It can also depict the movement of meso-molecules In this way it offers advantages in elucidatingthe solubilization processStudies have shown that the solubilization effects of

saponin are related to the structure and properties of

Fig 1 Models used in DPD simulation (a) drug models (b) sketchfor Ro molecule self-assembly process The vesicle was used as basisfor study of the solubilization behavior of Ro (c) mesostructure templateand the final coarse-grained model system containing vesicles drugs andwater for DPD simulation (water beads were hidden)

insoluble drugs32 We divided the insoluble drugs intotwo types according to their hydrophilic and hydrophobicstructures The first group of molecules has only hydropho-bic parts This group includes some free aglycones (suchas quercetin and coumarin) The molecules in the othergroup contain both hydrophobic and hydrophilic partsAccording to the number of hydrophilic parts they canbe subdivided further This group includes some glyco-sides (such as baicalin and saikosaponin) These drugmolecules are coarse-grained and their mesostructures areshown in Figure 1 By employing DPD simulation theeffects of hydrophilic and hydrophobic properties andthe length of the hydrophobic and hydrophilic groupsof drugs on solubilization are studied These results canact as guidance for the further use of Ro and othersaponins

2 DPD SIMULATION METHODDissipative particle dynamics (DPD) is a mesoscopic sim-ulation technique suitable to the study of the collectivebehavior of complex fluids A DPD bead represents a smallregion of fluid matter and its motion is assumed to begoverned by Newtonrsquos laws31

dridt

= vi mi

dvidt

= f i (1)

Here ri vi mi and fi denote the position vector velocitymass and total force acting on particle i respectivelyThe force fi between each pair of beads contains three

parts a harmonic conservative interaction force (FCij

which is a soft repulsion acting along the line of centersa dissipative force (FD

ij which represents the viscous dragbetween moving beads and a random force (FR

ij whichmaintains energy input into the system in opposition to thedissipation All forces are short-range with a fixed cutoffradius rc which is usually chosen as the reduced unit oflength rc equiv 1 They are given as follows

fi =sumj =i

FCij +FD

ij +FRij (2)

FCij =

⎧⎨⎩aij1minus rij rij rij lt 1

0 rij ge 1(3)

FDij ==minuswDrij rij middotvij rij (4)

FRij == wRrij ij

1radict

rij (5)

Here aij is the maximum repulsion between particles iand particle j rij = ri minus rj rij = rij rij = rijrij isthe dissipation strength is the noise strength wD andwR are r-dependent weight functions vanishing for r gt 1ij is a random number with zero mean and unit varianceand t is the time step of the simulation

J Comput Theor Nanosci 11 2046ndash2054 2014 2047



RESEARCH

ARTIC

LE

Dissipative Particle Dynamics Simulation of Ginsenoside Ro Vesicular Solubilization Systems Dai et al

Combined with the known compressibility of water andbinodal data the following expression can be obtainedbetween aij and the Flory-Huggins parameter

aij = aii+327ij (6)

Here aii = 25kBT at a density =NV = 3 and the valueof kBT is defined as 1The Flory-Huggins parameter ij is often used to esti-

mate the compatibility between drugs and solubilizerLower ij values indicate better compatibility It is thesame for the repulsion parameter a because it comes fromij In DPD simulations changes in ij or aij betweentwo specific particles are a common means of studying thelaws governing these interactions However this perspec-tive often ignores the interactions between these particlesand other particles The ij values can be calculated fromthe solubility parameters using the following equation

ij =iminusj

2V

RT(7)

itself can also be used to estimate the compatibilityof different species The similar the solubility parametersof the two species the smaller the interaction parametersbetween them and the stronger their compatibility We sug-gest that changing the ij or aij values by changing the of specific particles will produce more accurate resultsIn this way the complex interactions between all the par-ticles can be taken into consideration in simulations depends on the chemical nature of the species and can beobtained through experimentation or molecular dynamics(MD) simulation

3 SIMULATION DETAILSModels The DPD simulation system was composed ofhydrophobic drugs ginsenoside Ro (Ro) and water Threedifferent types of hydrophobic drugs were studied andtheir mesostructures are shown in Figure 1(a) Bead Xand bead G represent the hydrophobic and hydrophilicparticles of drug respectively Model 1 represents drugscontaining only hydrophobic beads X Model 2 representsglycosides with hydrophobic aglycone Xm and single sugarchain Gn Model 3 represents glycosides containing twosugar chains at the both end of aglycone One end containsseveral sugars and the other end contains only one sugarThe mesostructure of Ro is shown in Figure 1(b) Ro ispentacyclic triterpenoid glycoside with two sugar chainsThe aglycone is divided into five identical beads (M) Eachsugar is represented by one bead (G for glucose GA forglucuronic acid) One water molecule is represented byone bead (W for water) All these beads have the samevolumeIn our previous study Ro molecules were shown to

self-assemble into vesicles11 The radius of each vesicle isabout 5rc (where rc is the DPD length scale) To study

Table I Solubility parameters of beads

Bead type (Jcm312

M 1618G 3537GA 6723W 4837

Table II Repulsion parameters aij used in DPD simulation

M G GA W X1lowast X2lowast X3lowast X4lowast X5lowast

M 2500G 8338 2500GA 43340 17575 2500W 12371 3970 5532 2500X1lowast 3140 12700 53768 16498 2500X2lowast 2524 9076 45202 13087 2916 2500X3lowast 2741 6245 37418 10152 4164 2916 2500X4lowast 3793 4205 30416 7692 6245 4164 2916 2500X5lowast 5677 2957 24197 5708 9157 6245 4164 2916 2500

Notes lowastX1-5 represent X beads with different values of X1 = 10 X2 = 15X3 = 20 X4 = 25 and X5 = 30

the solubilization mechanism by which hydrophobic drugsare loaded into vesicles the Ro vesicle was placed in awatery environment to serve as the basis (Fig 1(c)) Thesize of the cubic simulation box was 30times30times30r3c Thedrug molecules were homogeneously distributed in the sol-vent The molar concentration ratio of the drugs and Romolecules was fixed at 21 The total simulation time was50000 steps which was long enough for the simulationsystem to reach equilibrium The spring constant was fixedat 40 which has been found to give reasonable resultsin this system The simulation temperature was fixed at25 C To show aggregate morphologies clearly the dis-play styles were chosen so that all water beads were hiddenin the simulationRepulsion parameters To calculate repulsion parameter

aij the solubility parameters of all beads were neededIn this study solubility parameters of Ro were calculatedusing the Amorphous Cell module in Materials Studio 55(Accelrys Inc supported by CHEMCLOUDCOMPUT-ING) with the COMPASS force field The results are inTable I To study the compatibility effects between thedrug and Ro we set the solubility parameter of X (X)to 10 15 20 25 and 30 The interaction parameters andrepulsion parameters can be calculated using equations 6and 7 and the results are shown in Table II

4 RESULTS AND DISCUSSION41 Effects of Interaction Parameter Between Drug

Beads and Ro Beads on 2 SolubilizationThe interaction parameter between drug beads and Robeads reflects the compatibility of drug and Ro moleculesLower parameter values between two beads indicatestronger attractive interactions between them and they are

2048 J Comput Theor Nanosci 11 2046ndash2054 2014



RESEARCH

ARTIC

LE


tend to aggregate with each other Therefore is veryimportant to the study of the solubilizing capacity and sol-ubilization sites of solubilization systems can be cal-culated from solubility parameter as in Eq (7) HereModel 1 (Xm) and Model 2 (XmGn) drugs (m = n = 1)were used to study the effects of interaction parameterson the solubilization of Ro The X value was allowed torange from 10 to 30 The results are shown in Figures 2and 3Figures 2(a) and (b) show section views of Ro vesicles

They directly indicate the location of different drugs inthese vesicles Model 1 drugs containing only hydropho-bic groups were mainly solubilized in the hydrophobiclayer With increased values of X the location within thehydrophobic layer changed slightly as indicated by thedistribution probability of X in Figure 2(c) (distributionprobability P = nN n is the number of X beads on a

Fig 2 Section views and the distribution probability of vesicles for different drugs (a) vesicles with Model 1 drugs (b) vesicles with Model 2 drugs(1) X = 10 (2) X = 15 (3) X = 20 (4) X = 25 (5) X = 30 (c) and (d) are the distribution probability of X beads of Model 1 and Model 2 drugsrespectively

certain distance away from the center of the vesicle N isthe total number of X beads in the vesicle) When X = 10which was smaller than M = 1618 the drug moleculestended to self-aggregate in the hydrophobic layer formedby the aglycones of Ro molecules which was far fromthe external water solution environment (Fig 2(a1)) Theprobability distribution curve formed a single sharp peakin the hydrophobic layer This was because at this timethe interaction parameter between X and M (XM ) wasmuch smaller than that between X and W (XW W =4837) but it was larger than the one between X andX itself (XX The campatibility between X beads wasthe stongest followed by M beads The strong repulsiveforces between X and W forced themaway from eachother When X increased to 15 and 20 which was similarto M the XM decreased and approached XX This indi-cated excellent compatibility between the drugs and Ro




RESEARCH

ARTIC

LE


molecules The former were distrubited uniformly in thehydrophobic layer formed by the latter (Figs 2(a2) (a3))The probability distribution curve became wider When X

increased to 25 the difference between X and M andbetween X and G (where G is the hydrophilic groupof Ro) were almost the same That meant the strength ofthe interactions between the drugs and the hydrophobicand hydrophilic groups of Ro were largely the same Thedrugs were mainly associated with the zone between thehydrophobic and hydrophilic layers The probability dis-tribution curve peaks appeared in both the inner and outerjuction zones of the hydrophobic and hydrophilic layersUnlike Model 1 drugs Model 2 drugs were solubi-

lized in the palisade layer of the vesicles The hydropho-bic groups of the drug molecules were inserted into thehydrophobic palisade layer and the hydrophilic groups ofthe drugs were located on the surface where they inter-acted with the sugars of Ro through hydrogen bondingand dipolar interactions There were no significant changesin the increases of X from 10 to 25 (Figs 2(b1)ndash(b4))This was because the attractive interactions between thehydrophilic groups of drugs and Ro molecules were verystrong However when X increased to 30 the repulsiveinteractions between X and M beads were so large thatthey prevented the drug molecules from coming close toeach other The drug molecules could not then enter thehydrophobic layer and only adsorbed onto the vesicle sur-face by hydrophilic interaction (Fig 2(b5))The solubilization capacity of each vesicle was also

affected by the type of drug The solubilization capacityof a surfact can be quantified by the molar solubilizationratio (MSR) which is often calculated as follows3334

MSR= SminusScmc

Cs minusCMC(8)

Here Cs is the surfactant concentration CMC is the crit-ical micelle concentration of the surfactant S is the totalapparent solubility of the additive and Scmc is the apparentsolubility of the additive at CMC The MSR is a ratio ofthe moles of the additive solubilized to the moles of sur-factant in micelle or vesicle In DPD simulation the molesof the solubilized additive and the surfactant forming thevesicle can be determined easily As shown in Figure 3when the solubility parameter was small (X = 10) theMSR for both Model 1 and Model 2 drugs was also smallThat might have been because of the strong repulsive inter-actions between X and G which hindered the passage ofdrugs through the hydrophilic layer of the vesicle Thedrug molecules tended to self-aggregated in the aqueousfraction and the MSR tended to be low When X increasedto 15 and 20 which approached M the attractive interac-tion between drugs and hydrophobic layer peaked show-ing the highest MSR of any of the drugs studied here Withthe X increased to 25 the repulsive interaction betweendrugs and hydrophobic layer increased It was not easy forthe drugs to enter the vesicle and the MSR decreased

Fig 3 Solubilization efficiencies of Ro vesicles for Model 1 (X) andModel 2 (XG) drugs with different solubility parameters of X

The MSR results of Model 1 and Model 2 drugs werecompared When X was small (X = 10 and 15) theMSR of Model 1 was higher than that of Model 2 Thismight have been because the molecules of the Model 1group were located in the hydrophobic layer and thoseof the Model 2 were located in the palisade layer whichwas smaller than the hydrophobic layer At large valuesof X (X = 20 and 25) the MSR of Model 2 was higherthan that of Model 1 That might have been because ofthe repulsive interactions between X and M beads whichprevented the Model 1 drug molecules from entering thevesicle However Model 2 drug molecules were found toexist stably in the palisade layer where their hydrophilicgroup G interacted with the hydrophilic layer of the vesi-cle through hydrogen bonding or dipolar interactions

42 Effects of Length of HydrophobicStructures of Drugs on Solubilization

The length of hydrophobic structure can affect the molecu-lar size and polarity of the drug which can greatly influcethe solubilizating capacity of the Ro vesicle A Model 2drug was used in further study (n = 1 m = 1 2 3 4and 5) X was fixed at 15 which produced peak MSRThe results are shown in Figure 4 Here the MSR of drugswere not found to increase or decrease monotonically withincreases in the legnth of the hydrophobic group Insteadtwo decreasing processes were observed When the num-ber of beads here given as m increased from 1 to 3 theMSR decreased from 1553 to 1000 As shown inFigures 4(a1)ndash(a3) the drug molecules were mainly locatedin the outer palisade layer (Few of entered the inner layer)Typical molecules are highlighted in the figure) Howeverlonger hydrophobic structures produced larger moleculesThis made it difficult for the molecules to incorporate intothe vesicle because both the superficial area and the vol-ume of the vesicle were limited When m increased to4 the MSR increased to 1367 Because the length ofthe hydrophobic of the drug was similar to that of theRo molecule (the length of Ro hydrophobic structure was




RESEARCH

ARTIC

LE


5 beads) the drug molecules had a chance to interact withthe inner palisade layer of the vesicle In this case thedrug molecules almost formed a mixed vesicle with the Romolecules considering that the molecules of this lengthmay have surface activity This phenomenon has beenobserved in interactions between Ro and saikosaponin amolecules whose hydrophobic structure length was also 5beads11 When m increased to 5 the drug molecules couldstill form mixed vesicles with Ro molecules (Fig 4(a5))However due to the limited vesicle space the drugs withlarger molecular sizes could not easily incorporate intothe vesicle The MSR decreased to 987 Although theefficiency of solubilization fluctuated with increases inthe length of the hydrophobic structure the vesicle sizeincreased smoothly This indicates that the overall volumeof solubilized drugs also increases smoothly

43 Effects of Length of Hydrophilic Structure ofDrugs on Solubilization

The length of the hydrophilic structure of the drug canalso affect the drug molecular size and polarity and soinfluence the solubilization behavior of Ro vesicles HereModel 2 and Model 3 (G1XmGn drugs were employed tostudy the effects (m = 3 n = 1 2 3) The X was stillfixed at 15 As shown in Figure 5 when the number ofG beads increased from 1 to 2 the MSR increased forboth Model 2 and Model 3 drugs This might because thesolubilization site of both types of drugs was the palisadelayer (Fig 6) where the hydrophobic and hydrophilic

Fig 4 Simulation results of Ro vesicle and Model 2 (XmG1) drugs (a) vesicles with drugs with hydrophobic structures of different lengths (a1) m= 1(a2) m = 2 (a3) m = 3 (a4) m = 4 (a5) m = 5 (b) typical view of aggregation mophology of Ro vesicle and Model 2 drugs (c) Solubilizationefficiencies of Ro vesicles for Model 1 drugs with hydrophobic structures of different lengths

Fig 5 Solubilization efficiencies of Ro vesicles for Model 2 (X3Gn)and Model 3 (G1X3Gn) drugs with hydrophilic structures of differentlengths

structures of the drugs interacted with the hydrophobic andhydrophilic layers of the vesicle respectively Because thehydrophobic structure of each drug remained unchangedthe hydrophilic interactions became the bigggest influ-ence on solubilization capacity Increases in the length ofthe hydrophilic structure caused increases in the strengthof the hydrophilic interactions between the drug and Romolecules and stabilized the drug in the vesicle How-ever when n increased to 3 the greater size of thedrug molecules prevented them from entering the vesi-cle Increases in the length of the hydrophilic structurealso increased the polarity of each drug molecule The




RESEARCH

ARTIC

LE


Fig 6 Aggregation mophologies and section views of vesicles with Model 2 (X3Gn) and Model 3 (G1X3Gn) drugs (a) Aggregation mophologies(b) section views of vesicles (1) Model 2 (n= 2) (2) Model 3 (n= 2) (3) Model 3 (n= 3)

hydrophilicity of these drugs increased and they began toshow surface activity The drug molecules were found toself-aggregate in aqueous solution to form stable micellesor vesicles This decreased solubilization efficiencyWhen n = 1 or 2 the MSR of two-suger-chain drugs

(Model 3) was higher than that of single-sugar-chain drugs(Model 2) The morphology and sectional views of the vesi-cles may provide us with explanations Vesicles containg-ing solubilized single-sugar-chain drugs continued to takethe shape of spheres (Fig 6(a1)) These drug moleculeswere mainly located on both the outer and inner pal-isade layers with the hydrophobic structures embeddedin the hydrophobic layer (see the highlighted moleculesin Fig 6(b1)) However the vesicles increased in sizeand became spheroid as they incorporated two-sugar-chaindrugs (Fig 6(a2)) The axial section view of shows that thedrug molecules interacted with vesicles in three differentways (see highlighted molecules in Fig 6(b2))(1) Solubilized in outer palisade layer The hydrophilicgroups at both ends of drug molecules interacted with thehydrophilic layer of the vesicle and the hydrophobic struc-ture bent in the hydrophobic layer(2) Solubilized in inner palisade layer The behavior ofmolecules were similar to those in (1) However they werelocated near the internal hydrophilic region(3) Solubilized throughout whole the layer of the vesicle

The hydrophobic sturctures were embedded in thehydrophobic layer while two sugar chain interacted withouter and inner hydrophilic layer respectively In thisway the drug molecules acted as composites with Romolecules and solubilized more drug molecules As a resultof these three different means of solubilization vesicle sizeincreased significantly and the solubilization was highlyefficient

However this was not the case when n = 3 TheMSR for both single- and especially two-sugar-chain drugsdecreased This was because the hydrophilicity of bothdrugs increasd significantly Only a small number ofmolecules interacted with the vesicle Most tended toself-assemble in aqueous solution Figures 6(a3) and (b3)showed the aggregate morphology and section view ofvesicles with two-sugar-chain drugs solubilized Only afew drug molecules were solubilized in the outer palisadelayer No molecules entered into the internal palisade layerof the vesicle Most of the molecules self-assembled inaqueous solution to form micelles and even vesicles Thatled to a very low solubilization efficiency of 547

44 Solubilization Processes of Drugs by Ro VesicleThe mechanism by which the drug molecules enterthe vesicle is also significant Evaluations of molecular

Fig 7 Snapshots and section view of typical simulation (Ro vesiclewith Model 2 drugs added) (a) t = 0 step drug molecules were ran-domly distributed in aqueous solution (b) t = 30000 steps some drugmolecules were solubilized into the vesicles and others self-assembledin aqueous solution to form micelles (c) Section view of the vesicle att = 30000 steps Drug aggregates are gradually incorporated into the Rovesicle




RESEARCH

ARTIC

LE


trajectory showed that different types of molecules enterthe vesicles through similar processes For example themeans by which Model 2 drugs entered the vesicles areshown in Figure 7 (m= 2 n= 1 X = 15) Starting fromthe initial state of randomly dispersal (Fig 7(a)) some ofthe drug molecules close to the vesicle entered it preferen-tially Other molecules self-assembled in aqueous solutionto form small aggregates (Fig 7(b)) The aggregate mightbe small agglomerates micelles or vesicle for differenttypes of drugs Then the aggregates incorporated into thevesicle as a whole This was followed by re-distribution ofthe drug molecules within the vesicle (Fig 7(c))

5 CONCLUSIONEvery given surfactant system is influenced by many fac-tors such as the size shape polarity and branching ofthe target molecules which can significantly influence thesolubilizing behavior of the surfactant35 Extensive exper-imental and theoretical work have been performed onmicellar and vesicular solubilization systems formed bysynthetic surfactants and their relationships with additivesHowever little work has been done on vesicular solubiliza-tion system of saponin Ginsenoside Ro has been found toform vesicles and to solubilize insoluble drugs11 This indi-cates that this kind of vesiclar solubilization system meritsfurther study and may have significant applications Hereby using DPD simulation method the effects of hydropho-bicity topological structure and the length of hydropho-bic and hydrophilic structures were studied Both the siteof solubilization and capacity were found to be signif-icantly influenced by these factors Simple hydrophobicdrugs without hydrophilic groups were mainly located onthe hydrophobic layer of the vesicle and drugs containingboth hydrophobic and hydrophilic structures were locatedin the palisade layer When the length of both hydrophobicand hydrophilic structures were short (m= 1 2 3 n= 1)the drug molecules were only located in the outer pal-isade layer However when the length of the drug moleculeapproached that of the Ro molecule some of them becamedistrubuted in both the inner and outer palisade layersto form mixed vesicles with Ro molecules The effectsof the study factors on MSR were found to be complexThe similarity between the hydrophobic structures ofdrugsand Ro (X = M could increase the MSR above thatof drugs with the same topological structure This maybe used to guide the structural modification of drugs toenhance their solubility in this solubilization system Fora given solubilization site lower MSR were observed onlarger molecules as the result of the limited space Moremolecules were found to enter the vesicle if there weremany possible sites at which they might do so The drugswere also found to incorporate into the vesicle throughboth single-molecule and aggregete morphology Howeverwhenever the hydrophobicity or hydrophilicity of the drugwas too extreme the molecules tended to self-aggregate in

aqueous solution instead of entering the vesicle The MSRvalues of these drugs were consequently very lowIn this study DPD simulation has been proved to be

an applicable tool to study the organization behavior ofnanoaggregations It can provide high-efficiency predic-tion to experiments in both structures and dynamic pro-cesses Our simulation results not only provide insight intothe vesicular solubilization system formed by saponin butmay also facilitate the consturction of guidelines for thesafe and effective use of Ro as a solubilizer and the furtherdevelopment and application of Ro in the pharmaceuticalindustry

Acknowledgment This work was financially supportedby the National Natural Science Foundation of China(81073058) and Innovation Team Foundation of BeijingUniversity of Chinese Medicine Beijing (2011-CXTD-11Research Center of TCM-information Engineering)

References1 G Gaucher P Satturwar M C Jones A Furtos and J C Leroux

Eur J Pharm Biopharm 76 147 (2010)2 S Kim Y Shi J Y Kim K Park and J X Cheng Expert Opin

Drug Del 7 49 (2010)3 P van Hoogevest X Liu and A Fahr Expert Opin Drug Del 8

1481 (2011)4 Ouml Guumlccedilluuml-Uumlstuumlndag and G Mazza Crit Rev Food Sci 47 231

(2007)5 O Tanaka and N Yata Promotion of absorption of drugs adminis-

tered through the alimentary system US Patent 4501734 (1985)6 S Tsuzaki S Wanezaki and H Araki In Google Patents (2008)7 D Kiefer and T Pantuso Am Fam Physician 68 1539 (2003)8 A S Attele J A Wu and C S Yuan Biochem Pharmacol 58

1685 (1999)9 Traditional Chinese Medicine Database System httpcowork

cintcmcomenginewindex1jsp10 H Kimata N Sumida N Matsufuji T Morita K Ito N Yata and

O Tanaka Chem Pharm Bull 33 2849 (1985)11 X X Dai X Y Shi Y G Wang and Y J Qiao J Colloid Interf

Sci 384 73 (2012)12 X X Dai X Y Shi Q Q Yin H O Ding and Y J Qiao J Col-

loid Interf Sci 396 165 (2013)13 K Eom G Yoon J I Kim and S Na J Comput Theor Nanos

7 1210 (2010)14 G Miller and D Trebotich J Comput Theor Nanos 4 797 (2007)15 A Eghdami and M Monajjemi Quantum Matter 2 324 (2013)16 M Hashemy and M Monajjemi Quantum Matter 2 214 (2013)17 K Ghatak P Bose S Bhattacharya A Bhattacharjee D De

S Ghosh S Debbarma and N Paitya Quantum Matter 2 83(2013)

18 T Ono Y Fujimoto and S Tsukamoto Quantum Matter 1 4(2012)

19 B Tuzun and S Erkoc Quantum Matter 1 136 (2012)20 Z Wang F Wang C Su and Y Zhang J Comput Theor Nanos

10 2323 (2013)21 Z Wang Y Chen J Li X Luo and H Tan J Comput Theor

Nanos 8 1910 (2011)22 X D Guo Y Qian C Y Zhang S Y Nie and L J Zhang Soft

Matter 9989 (2012)23 S Yamamoto Y Maruyama and S Hyodo J Chem Phys

116 5842 (2002)24 X D Guo L J Zhang Z M Wu and Y Qian Macromolecules

43 7839 (2010)




RESEARCH

ARTIC

LE


25 C Zhong and D Liu Macromol Theor Simul 16 141 (2007)26 A Maiti J Wescott and G Goldbeck-Wood Int J Nanotechnol

2 198 (2005)27 C Soto-Figueroa and L Vicente Soft Matter 7 8224 (2011)28 P Posocco M Fermeglia and S Pricl J Mater Chem 20 7742

(2010)29 S McGrother L Y Ming and G Goldbeck-Wood Mesoscale Phe-

nomena in Fluid Systems edited by C Fiona ACS PublicationsWashington DC (2003) Chap 14 pp 227ndash241

30 P Hoogerbrugge and J Koelman Europhys Lett 19 155(1992)

31 R D Groot and P B Warren J Chem Phys 107 4423 (1997)32 U Walthelm K Dittrich G Gelbrich and T Schoepke Planta Med

67 49 (2001)33 D A Edwards R G Luthy and Z Liu Environ Sci Technol 25

127 (1991)34 M R Porter Surface active agents Handbook of surfactants edited

by M R Porterin Chapman amp Hall Blackie London (1994)pp 55ndash59

35 D Myers Surface chemistry and surface active agents SurfactantScience and Technology edited by D Myers Wiley Online LibraryHoboken New Jersey (1988) Chap 6 pp 194ndash200

Received 4 August 2013 Accepted 23 August 2013




RESEARCH

ARTIC

LE


the applications of ginsenoside and similar saponins inmedical situations and the pharmacological industry How-ever in these studies standard experimental methods canbe very costly in both time and money and it can bedifficult to illustrate the solubilization mechanisms andthe exact internal structures of the vesicles Fortunatelypowerful computer simulations can render the researcheshighly efficient Many theories have been developed tostudy the organization of quantum matter13ndash19 Dissipativeparticle dynamics (DPD) is one of the mesoscopic sim-ulation methods used to study drug loading systems20ndash28

This is a particle-based method which coarse-grains thefamiliar atomistic representation of the molecules to gainorders of magnitude in both length and time scale rela-tive to traditional atomistic scale29 DPD allows particlesto interact through a simple-wise potential and accuratelycapture their hydrodynamic behavior and the underlyinginteractions3031 It can also depict the movement of meso-molecules In this way it offers advantages in elucidatingthe solubilization processStudies have shown that the solubilization effects of

saponin are related to the structure and properties of

Fig 1 Models used in DPD simulation (a) drug models (b) sketchfor Ro molecule self-assembly process The vesicle was used as basisfor study of the solubilization behavior of Ro (c) mesostructure templateand the final coarse-grained model system containing vesicles drugs andwater for DPD simulation (water beads were hidden)

insoluble drugs32 We divided the insoluble drugs intotwo types according to their hydrophilic and hydrophobicstructures The first group of molecules has only hydropho-bic parts This group includes some free aglycones (suchas quercetin and coumarin) The molecules in the othergroup contain both hydrophobic and hydrophilic partsAccording to the number of hydrophilic parts they canbe subdivided further This group includes some glyco-sides (such as baicalin and saikosaponin) These drugmolecules are coarse-grained and their mesostructures areshown in Figure 1 By employing DPD simulation theeffects of hydrophilic and hydrophobic properties andthe length of the hydrophobic and hydrophilic groupsof drugs on solubilization are studied These results canact as guidance for the further use of Ro and othersaponins

2 DPD SIMULATION METHODDissipative particle dynamics (DPD) is a mesoscopic sim-ulation technique suitable to the study of the collectivebehavior of complex fluids A DPD bead represents a smallregion of fluid matter and its motion is assumed to begoverned by Newtonrsquos laws31

dridt

= vi mi

dvidt

= f i (1)

Here ri vi mi and fi denote the position vector velocitymass and total force acting on particle i respectivelyThe force fi between each pair of beads contains three

parts a harmonic conservative interaction force (FCij

which is a soft repulsion acting along the line of centersa dissipative force (FD

ij which represents the viscous dragbetween moving beads and a random force (FR

ij whichmaintains energy input into the system in opposition to thedissipation All forces are short-range with a fixed cutoffradius rc which is usually chosen as the reduced unit oflength rc equiv 1 They are given as follows

fi =sumj =i

FCij +FD

ij +FRij (2)

FCij =

⎧⎨⎩aij1minus rij rij rij lt 1

0 rij ge 1(3)

FDij ==minuswDrij rij middotvij rij (4)

FRij == wRrij ij

1radict

rij (5)

Here aij is the maximum repulsion between particles iand particle j rij = ri minus rj rij = rij rij = rijrij isthe dissipation strength is the noise strength wD andwR are r-dependent weight functions vanishing for r gt 1ij is a random number with zero mean and unit varianceand t is the time step of the simulation




RESEARCH

ARTIC

LE



aij = aii+327ij (6)



ij =iminusj

2V

RT(7)





Bead type (Jcm312

M 1618G 3537GA 6723W 4837












RESEARCH

ARTIC

LE









RESEARCH

ARTIC

LE






MSR= SminusScmc

Cs minusCMC(8)









RESEARCH

ARTIC

LE











RESEARCH

ARTIC

LE












RESEARCH

ARTIC

LE


























43 7839 (2010)




RESEARCH

ARTIC

LE
















RESEARCH

ARTIC

LE



aij = aii+327ij (6)



ij =iminusj

2V

RT(7)





Bead type (Jcm312

M 1618G 3537GA 6723W 4837












RESEARCH

ARTIC

LE









RESEARCH

ARTIC

LE






MSR= SminusScmc

Cs minusCMC(8)









RESEARCH

ARTIC

LE











RESEARCH

ARTIC

LE












RESEARCH

ARTIC

LE


























43 7839 (2010)




RESEARCH

ARTIC

LE
















RESEARCH

ARTIC

LE









RESEARCH

ARTIC

LE






MSR= SminusScmc

Cs minusCMC(8)









RESEARCH

ARTIC

LE











RESEARCH

ARTIC

LE












RESEARCH

ARTIC

LE


























43 7839 (2010)




RESEARCH

ARTIC

LE
















RESEARCH

ARTIC

LE






MSR= SminusScmc

Cs minusCMC(8)









RESEARCH

ARTIC

LE











RESEARCH

ARTIC

LE












RESEARCH

ARTIC

LE


























43 7839 (2010)




RESEARCH

ARTIC

LE
















RESEARCH

ARTIC

LE











RESEARCH

ARTIC

LE












RESEARCH

ARTIC

LE


























43 7839 (2010)




RESEARCH

ARTIC

LE
















RESEARCH

ARTIC

LE












RESEARCH

ARTIC

LE


























43 7839 (2010)




RESEARCH

ARTIC

LE
















RESEARCH

ARTIC

LE


























43 7839 (2010)




RESEARCH

ARTIC

LE
















RESEARCH

ARTIC

LE














dissipative particle dynamics simulation of ginsenoside ro...

Documents