Pharmacokinetics Studies of 12 Alkaloids in Rat Plasma after Oral Administration of Zuojin and Fan

18 Pages·2017·2.04 MB·English

molecules
Article
Pharmacokinetics Studies of 12 Alkaloids in Rat
Plasma after Oral Administration of Zuojin and
Fan-Zuojin Formulas
PingQian1,2,You-BoZhang1,Yan-FangYang1,WeiXu1andXiu-WeiYang1,*
1 StateKeyLaboratoryofNaturalandBiomimeticDrugs,DepartmentofNaturalMedicines,
SchoolofPharmaceuticalSciences,PekingUniversityHealthScienceCenter,PekingUniversity,
No.38,XueyuanRoad,HaidianDistrict,Beijing100191,China;[email protected](P.Q.);
[email protected](Y.-B.Z.);[email protected](Y.-F.Y.);[email protected](W.X.)
2 DepartmentofPharmacognosy,SchoolofPharmacy,ChinaMedicalUniversity,Shenyang110122,China
* Correspondence:[email protected];Tel.:[phone];Fax:[phone]
AcademicEditor:ThomasJ.Schmidt
Received:17December2016;Accepted:24January2017;Published:30January2017
Abstract: Zuojin formula (ZJ) is a traditional Chinese medicine (TCM) prescription consisted of
CoptidisRhizoma(CR)andEuodiaeFructus(EF),andhasbeenusedtotreatgastrointestinal(GI)
diseaseformorethan700years.Fan-Zuojinformula(FZJ)isarelatedTCMprescriptionalsoconsisted
ofCRandEFwiththeoppositeproportion. Inrecentyears,ZJwasgettingmoreattentionforits
antitumor potential, but the indeterminate pharmacokinetic (PK) behavior restricted its clinical
applications,andthePKdifferencesbetweenZJandFZJwerealsolargelyunknown. Consequentlyit
is necessary to carry out a full-scale PK study to demonstrate the physiological disposition of
ZJ, as well as the comparative PK study between ZJ and FZJ to illustrate the compatibility dose
effects. Therefore a liquid chromatographic–tandem mass spectrometry (LC–MS/MS) method was
established and validated for the determinations of coptisine, epiberberine, palmatine, berberine,
8-oxocoptisine,8-oxoepiberberine,noroxyhydrastinine,corydaldine,dehydroevodiamine,evodiamine,
wuchuyuamide-I,andevocarpineinratplasma.PKcharacteristicsof12alkaloidsafteroraladministration
ofZJandFZJwerecompared,andtheresultwasanalyzedanddiscussedwiththehelpofaninsilico
study.ThenanintegratedPKstudywascarriedoutwiththeAUC-basedweightingmethodandthetotal
drugconcentrationmethod. TheestablishedmethodhasbeensuccessfullyappliedtorevealthePK
profilesofthe12alkaloidsinratplasmaafteroraladministrationofZJandFZJ.Theresultsshowedthat:
(1)doublepeakswereobservedintheplasmaconcentration-time(C–T)curvesofthealkaloidsafter
ZJadministration;buttheC–Tcurvesapproximatelymatchedthetwo-compartmentmodelafterFZJ
administration;(2)Therewerewidevariationsintheabsorptionlevelsofthesealkaloids;andeven
foracertainalkaloid,thedosemodifiedsystemicexposurelevelsandeliminationratealsovaried
significantlyafteradministrationofZJandFZJextracts. Theresultscouldbeinterpretedasfollows:
firstly,inhibitioneffectonGImotilitycausedbythehighcontentCRalkaloids(especiallyberberine)
inZJcoulddelaytheT ,andincreasetheabsorptionandsystemicexposurelevelsoftheother
max
alkaloids,andalsoleadtothedoublepeakphenomenonofthesealkaloids. However,forquaternary
protoberberine alkaloids (QPA), double peaks were primarily caused by the different Ka value
in two intestinal absorption sites; Secondly, absorption was the major obstacle to the systemic
exposurelevelofthealkaloidsfromCRandEF.InsilicoandPKstudiessuggestedthattheabsorption
of these alkaloids, except QPAs, mainly depended on their solubility rather than permeability;
Thirdly,EFcouldpromotetheabsorptionandacceleratetheeliminationofQPAs,andhadagreater
influenceontheformerthanthelatter. AtlasttheintegratedPKanalysissuggestedthatberberine
and dehydroevodiamine could be regarded as the representative components to reflect the PK
behaviorsofCRandEFalkaloidsafteradministrationofZJandFZJ.Inconclusion,theabsorption,
eliminationandsystemicexposurelevelofthesealkaloidsweremainlyinfluencedbytheproportion
Molecules2017,22,214;doi:10.3390/molecules22020214 www.mdpi.com/journal/molecules
Molecules2017,22,214 2of18
ofEFandCR,thepharmacologicaleffectonGImotility,andthephysicochemicalpropertyofthese
alkaloids. Thesefindingswouldbehelpfulforabetterunderstandingoftheactivitiesandclinical
applicationsofZJ,FZJandotherrelatedTCMprescriptions.
Keywords: Zuojin formula; Fan-Zuojin formula; Coptidis Rhizoma; Euodiae Fructus; alkaloids;
pharmacokinetic
1. Introduction
Zuojinformula(ZJ)isatraditionalChinesemedicine(TCM)prescriptionconsistedofCoptidis
Rhizoma (CR, the rhizome of Coptis chinensis Franch.) and Euodiae Fructus (EF, the fruit of
Euodia rutaecarpa (Juss.) Benth.) at a ratio of 6:1. ZJ has been used to treat gastrointestinal (GI)
disease for more than 700 years in China. In modern clinics, it is mainly used to treat gastric
ulcer,gastroesophagealrefluxdisease,gastritis,pyloricobstruction,etc.[1]. Inrecentyears,ZJhas
beeninvestigatedforitsantitumoractivities[2–5],buttheambiguouspharmacokinetic(PK)study
restricteditsclinicalapplications. ThePKstudiesofthreequaternaryprotoberberinealkaloids(QPAs),
berberine, palmatine and jatrorrhizine in rat plasma after administration of the CR–EF powder
hadalreadybeenreported[6],Ourpreviousstudieshaddiscoveredthatcompoundsisolatedfrom
ZJ had multiple structural skeletons except QPAs [7–9], Consequently, it is necessary to carry out
a multi-components PK study of ZJ, and both structure and bioactivity should be concerned in
theselectingofquantitativemarkers.
Alkaloids were the major components both in CR and EF. Berberine, palmatine, coptisine and
epiberberine,allbelongingtoQPAs,werethemostabundantalkaloidsinCR[10],Studiesdemonstratedthat
theyhadwidespectrumofpharmacologicaleffects,includinganti-diarrheal[11],anti-inflammatory[12],
anti-cancer [13,14] activities, and therapeutic potential on diabetes, hyperlipemia, hypertension [15],
cardiovascular system [16,17] and central nervous system [18,19] diseases. Besides QPAs, tertiary
protoberberinealkaloid(TPA)andsimpleisoquinolinealkaloid(SIA)wereothertwotypesofalkaloid
in CR, and the major components were 8-oxocoptisine, 8-oxoepiberberine, noroxyhydrastinine and
corydaldine. TPAs and SIAs also had notable bioactivities. For example, 8-oxocoptisine had gastric
mucousmembrane-protectiveactivity[20],andcouldinhibitP-glycoprotein(P-gp)mediatedmultidrug
resistance[21].ThereforefourQPAs,twoTPAsandtwoSIAsfromCRwereselectedasthequantitative
markersinthePKstudy.Evodiamineanddehydroevodiamine,belongingtoindoloquinazolinealkaloids
(IQAs)withfiveunbrokenringstructures,weretwomajoralkaloidsinEF,andhadbeenreportedto
haveanti-inflammatory[22],anti-cancer[23],neuroprotective[24,25],andvasodilatory[26]activities.
EFalsocontainedIQAswithbrokenringstructures[7],andwuchuyuamide-Iwaschosenasoneof
the quantitative markers to reflect the PK behavior of this type of alkaloid. Besides IQAs, EF also
contained quinolone alkaloids (QLAs), which had anti-bacterial [27], anti-inflammatory [28] and
anti-cancer[29]activities.SinceevocarpinewasoneofthemostabundantQLAsinEF[30],itshouldalso
bemonitoredinthePKstudy.Chemicalstructuresoftheselected12alkaloidswereshowninFigure1.
In previous PK studies of ZJ and related TCM prescriptions, discussions about the simple
PK results were insufficient, such as the lacked unambiguous interpretation about the double
peaks observed on plasma drug concentration-time (C–T) curves. Fan-Zuojin formula (FZJ) was
a TCM prescription also consisted of CR and EF, but with the opposite proportion (1:6) [4];
accordingly,acomparativePKstudyofZJandFZJwouldbehelpfultoreflecthowtheproportionof
EFandCRinfluencedthePKbehaviorsofthealkaloidsfromthetwoherbs. Astheabsorptionlevel
ofcompoundsmainlyinfluencedbytheirsolubilityandpermeability[31,32],aninsilicoassessment
wasnecessarytointerpretthewidevariationsoftheabsorptionandsystemexposurelevelbetween
thealkaloidswithdifferentstructuralskeletons. Atlast,theintegratedPKstudycouldgiveamore
visualizedC–TcurvesandPKparametersofCRandEFalkaloids. Areaundercurve(AUC)-based
Molecules2017,22,214 3of18
Molecules 2017, 22, 214 3 of 18
weightingmethodwasthemostfrequentlyusedintegratedmethodintheTCMPKstudy[33–35],
weighting method was the most frequently used integrated method in the TCM PK study [33–35],
whereas the total drug concentration method was more concise, thus the results calculated by
whereas the total drug concentration method was more concise, thus the results calculated by the
thetwomethodsdeservedtobecomparedanddiscussed. Takingintoconsiderationtheprevious
two methods deserved to be compared and discussed. Taking into consideration the previous issues,
issues,themaingoalsofthepresentresearchwastodevelopareliableliquidchromatographic–tandem
the main goals of the present research was to develop a reliable liquid chromatographic–tandem
massspectrometry(LC–MS/MS)methodtoanalyzetheratplasmapharmacokineticsofthemain
mass spectrometry (LC–MS/MS) method to analyze the rat plasma pharmacokinetics of the main
alkaloids from CR and EF, as well as to provide useful information for better understanding
alkaloids from CR and EF, as well as to provide useful information for better understanding the
the physiological disposition, pharmacological effect and clinical application of ZJ, FZJ and other
physiological disposition, pharmacological effect and clinical application of ZJ, FZJ and other related
relatTeCdMT CpMrespcrriepstciornipst. ions.
O H3CO H3CO H
O
cOoptisine N O He3CpiOberberine N O H3CpOalmatine N OCH3 O ONH
m/z 320.1 O m/z 336.1 O m/z 352.1 OCH3 nmo/zro 1x9y2h.y2drastinine
O -CO O -CH3 -H O -CH3 -H -HOCN -CO -H2
O N H3CO N H3CO N
O OCH3 H3CO
m/z 292.1 OCH3 m/z 320.2 O m/z 336.3 OCH3 m/z 119.1
H
HO
O O H H3CO H N OO N O
bme/rzOb 3e3ri6n.e1 N OOCCHH33 8m-/ozOx 3o3c6o.p1tisine N OOO 8-Hox3mCoeO/zp i3b5e2rb.2erineN OOO wucmh/uzy 3Hu5aH2m3.C3ideN-I
-CH3 -H -CO -CO -CH3 -H -CO
O O O
O N O NH H3CO NH N O
O O H
m/z 292.1 O m/z 308.2 m/z 308.3 m/z 158.2
O O
H H
O
H H
H
HH33CCcmOOo/rzy 2d0a8ld.i2nOeNH dehydmr/oze 3v0o1d.iHN7am3CinNeN O emv/oz d3iHaNH0m43C.i2nNeN O emv/oz cNC3aH4rp03.iCn2e13H25 8, 9 caOrmba/Nzm N2aH3z72e.p1ine
-HOCN -CH3 -H C11H22 -HOCN
H3CO H N N O HN O O
H3CmO/z 165.1 m/z 28H6.2N m/z H133C4.2 NCH3 m/z NH1294.2
m/z 186.1
Figure 1. Chemical structures of the precursor and product ions of the 12 alkaloids and
Figure1.Chemicalstructuresoftheprecursorandproductionsofthe12alkaloidsandcarbamazepine
carbamazepine (internal standard, IS).
(internalstandard,IS).
2. Results and Discussion
2. ResultsandDiscussion
2.1. Optimization of Chromatographic and Mass Conditions
2.1. OptimizationofChromatographicandMassConditions
Mobile phase was optimized by comparing acetonitrile (ACN)–water (H2O) and methanol
Mobile phase was optimized by comparing acetonitrile (ACN)–water (H O) and methanol
(MeOH)–H2O solvent systems. The mobile phase consisted of ACN and H2O was co2nfirmed finally
(MeOH)–H Osolventsystems. ThemobilephaseconsistedofACNandH Owasconfirmedfinally
with res2ults of higher response, lower background noise and shorter anal2ysis time than that of
withMreeOsuHl–tsHo2Of. hMigohreerovreers, ptohne saed,dliotiwone robf aac kbgurffoeur nsdystneomis ecoannsidstesdh oorft e0r.1a%n afolyrmsiisc taimcide (tvh/avn) atnhda t of
MeO11H m–Hm2oOl/L. Mamomreoonvieurm, tfhoermaadtde ictoiounld oimf aprbouveff ethres pyesatkem symcomnestirsyte, dionoifza0t.i1o%n effofercmt aicnda csiedns(ivti/vvit)y and
11momf tohle/ aLnaamlymteso. niumformatecouldimprovethepeaksymmetry,ionizationeffectandsensitivityof
theanalyAtells t.he tested alkaloids and internal standard (IS) all had better responses in positive mode than
nAegllattihvee tmesotdeed ina lkEaSIlo oidptsimanizdatiionnte. rTnhael Qst1a nadndar dQ3( IoSf) tahlel h12a dalbkaeltoteidrsr eanspdo InSs ewserine dpeotseicttievde bmy ode
infusing the standard solutions (200 ng/mL) in scan and product ions mode, respectively (structures
thannegativemodeinESIoptimization. TheQ1andQ3ofthe12alkaloidsandISweredetectedby
see Figure 1). Key parameters, such as DP, CE and CXP were also optimized.
infusingthestandardsolutions(200ng/mL)inscanandproductionsmode,respectively(structures

seeFigure1). Keyparameters,suchasDP,CEandCXPwerealsooptimized.
Molecules2017,22,214 4of18
2.2.MSaolmecupllees 2P01r7e,p 2a2r, a2t1i4o n 4 of 18
2D.2.i fSfaemrepnlet Psroeplvareantitosn, such as MeOH, MeOH–chloroform (3:1), and MeOH–ACN (2:1) (v/v),
wereusedasproteinprecipitationreagents,andtheextractionefficiencieswerecompared. Theresults
Different solvents, such as MeOH, MeOH–chloroform (3:1), and MeOH–ACN (2:1) (v/v), were
showuseedd tahsa tpMroteeOinH plreedciptoitaetxitorne mreealgyelnotws, arencdo vtheer ieexstfroacrtaiollna enfafilcyietensc,ieasn dweMree OcoHm–pAarCeNd. (T2h:1e) rwesausltbse tter
thanshMoweOedH –thcahtl oMroefOorHm le(3d: 1t)o beexctareumseeloyf ltohwe lorewcoevrebraiecsk gforro uanll dannaoliystee.sL, iaqnudi dM-leiqOuHid–AexCtNra c(2ti:o1n) wmaest hod
wasbaeltstoere tvhaalnu MateeOdHbu–cthfilonraolfloyrnmo (t3a:1d)o bpetceadu,sbe eocfa tuhsee lotwheerp boalackrigtrieosunodf tnhoeisaen. aLliyqtueisd-vlaiqruieidd ewxtidraecltyioann dit
wasmdeitfhfiocdu lwttaos fianlsdo aenvaalupapterodp bruiat tfeinsaolllvye nnottt oadeoxpttreadc,t baellcaaunsael ythtees pwolealrli.tiTehs uosf ,tphlea samnaalystaesm vpalreisedw ere
treawteiddewlyit ahnMd eitO wHa–s AdCiffNicu(l2t: 1to) afisndd easnc raipbpedroipnriSaetec tsioolnve3n.6t .to extract all analytes well. Thus, plasma
samples were treated with MeOH–ACN (2:1) as described in Section 3.6.
2.3. MethodValidation
2.3. Method Validation
Typical multiple reactions monitoring (MRM) mode chromatograms of blank plasma,
Typical multiple reactions monitoring (MRM) mode chromatograms of blank plasma, plasma
plasma spiked with the 12 alkaloids at the lower limit of quantification (LLOQ) and IS,
spiked with the 12 alkaloids at the lower limit of quantification (LLOQ) and IS, and plasma samples
and plasma samples after oral administration of ZJ and FZJ extract were shown in Figure 2.
after oral administration of ZJ and FZJ extract were shown in Figure 2. Retention times (Rt) were 7.57
Retentiontimes(R)were7.57(corydaldine),7.73(epiberberine),7.90(dehydroevodiamine),7.92(coptisine),
t
(corydaldine), 7.73 (epiberberine), 7.90 (dehydroevodiamine), 7.92 (coptisine), 8.12 (palmatine), 8.25
8.12(palmatine),8.25(noroxyhydrastinine),8.31(berberine),9.07(wuchuyuamide-I),10.04(carbamazepine),
(noroxyhydrastinine), 8.31 (berberine), 9.07 (wuchuyuamide-I), 10.04 (carbamazepine), 11.43
11.43 (8-oxoepiberberine), 12.18 (8-oxocoptisine), 12.32 (evodiamine), and 20.11 (evocarpine) min,
(8-oxoepiberberine), 12.18 (8-oxocoptisine), 12.32 (evodiamine), and 20.11 (evocarpine) min,
respreecstpivecetliyv.elTy.h Tehreerwe wereeren nooe ennddooggeennoouuss iinntteerrfeferreenncecse sini ntheth MeRMMR aMnaalynsaisly osfi sthoef altkhaeloaildksa alonidd sISa inn d IS
inblbalnankkp plalasmsmaa..

FiguFriegu2r.eM 2R. MMRcMhr cohmroamtoagtroagmrasmosf obfl abnlaknkp lpalsamsmaa( 1()1;)p; plalsamsmaas sppikikeedd wwiitthh tthhee 1122 aallkkaallooiiddss aatt LLLLOOQQ (2);
plas(m2)a; pslaasmmpal esasmaptle1s hat a1 fhte arftoerra olraald amdminiinsitsrtaratitoionn ooff ZZJJ (3(3) )anadn dFZFJ Z(4J) (e4x)treaxctt.r Aac t(c.oAryd(caoldriynde)a,l dB ine),
B (e(pepibibeerrbbeerriinnee)),, CC ((ddeehhyyddroroeveovdoidaimaminein),e )D, D(co(pctoispinties)i,n Ee) ,(pEalm(paatlinmea),t iFn e()n,oFrox(nyhoyrodxrayshtiyndinrea)s, tiGn ine),
(berberine), H (wuchuyuamide-I), I (carbamazepine), J (8-oxoepiberberine), K (8-oxocoptisine), L
G (berberine), H (wuchuyuamide-I), I (carbamazepine), J (8-oxoepiberberine), K (8-oxocoptisine),
(evodiamine), M (evocarpine).
L(evodiamine),M(evocarpine).
Molecules2017,22,214 5of18
Calibration curves of the 12 alkaloids had good linearity, and the regression equations,
Molecules 2017, 22, 214 5 of 18
linear ranges, correlation coefficients and LLOQ are listed in Table S1. It should be noted that
becauseCtahleibrcaotniocne ncutrravteiso nof othfet h12e aalnkaalloyitdess hhaads gbooeedn linqeuaardityru, panledd thdeu rreignrgestshioen seaqmuaptiloenps,r elipneaarra tion
procraenssg,esth, ceorarcetluatailonlo cwoeefsfticidernutsg acnodn LcLenOtQra atiroen lisdteedte irnm Tianbelde Si1n. Ipt lsahsomualdw bea snoateqdu tahratte rbeocfauthsee tLheL OQ.
Thecionntrcae-ndtraaytiaonnd ofi nthteer a-dnaalyypterse hciassi obneesn( rqeulaatdivruepslteadn dduarrdindg ethveia staiomnp,lRe SpDre)paanrdatiaocnc uprraoccieesss,( RthEe) aocftuqaul ality
conltorowle(sQt Cdr)usga mcopnlceesntwraetrioens hdoetweremdininedT ainb lpelaSs2m.aT hweasp rae qciusaiortners owf etrhee leLsLsOtQha. nTh1e5 %intaran-ddaayc caunrda cies
werientweri-tdhainy p±r1ec5i%sio.nTsh (erelliamtiivtea tsitoanndraarndg desevwiaetiroen2, 0R%SDa)n adnd± a2c0c%urraecsiepse (cRtiEv)e olyf qfuoarlQityC csoanmtropll e(QsCof) low
concsaemntpraletsio wne.rTeh sehorewseudlt sini nTdaibclaet eSd2. tThhaet tphreecmiseiothnso dwwerae slersesl itahbalne a1n5%d raenpdr oadccuucriabclieesfo wrethree wanitahliyns is.
±15%. The limitation ranges were 20% and ±20% respectively for QC samples of low concentration.
Absoluterecoveriesofthe12alkaloidswereallabove69%(TableS2).Theresultsofmatrixeffectwere
The results indicated that the method was reliable and reproducible for the analysis.
listedinTableS2,whichshowedaslightionsuppressioneffectforfourQPAs(from78.16%to85.16%),
Absolute recoveries of the 12 alkaloids were all above 69% (Table S2). The results of matrix
andanionenhancementeffectforwuchuyuamide-I(from116.32%to117.64%).However,nosignificant
effect were listed in Table S2, which showed a slight ion suppression effect for four QPAs (from
matrixeffectwasobservedforIS(from96.47%to105.22%),andtheRSDvalueswerewithintheacceptable
78.16% to 85.16%), and an ion enhancement effect for wuchuyuamide-I (from 116.32% to 117.64%).
criteria,whichindicatedthattherewerenoseverevariationsintheanalysis.Stabilityresultswerepresented
However, no significant matrix effect was observed for IS (from 96.47% to 105.22%), and the RSD
inTvaballueeSs3 w,werhei cwhitdheinm tohne satrcacetepdtatbhlae tctrhiteer1i2a, awlkhailcohi dinsdwicearteeds ttahbalte tihnerrea twpelares mnoa suenvdereer vdaifrfieartieonntss itno rage
condthitei oannsalaynsdis.d Sutraibniglitsya mrespuleltsp rwoecrees spinregs.ented in Table S3, which demonstrated that the 12 alkaloids
were stable in rat plasma under different storage conditions and during sample processing.
2.4. PKStudies
2.4. PK Studies
ThevalidatedanalyticalmethodwassuccessfullyappliedtothePKstudyofthe12alkaloidsafter
oraladmTihnei svtraalitdioantedo faZnJalayntidcaFl ZmJeetxhtorda cwt.aTs hsuecCce–sTsfcuullryv eapspolfietdh eto1 2thael kPaKlo sitdusdwy eorfe thseh o1w2 nalkinalFoiidgsu re3,
andatfhteer coorrarle asdpmonindiisntrgatPioKn poaf rZamJ aentder FsZwJ eerxetrlaisctte. dThine TCa–bTl ecu1r.vBees coafu tshee d1o2u ablklealpoeidask swwereer eshoobwsenr vine din
theFCi–gTurceu 3r,v aensdo tfhme coosrtroesfptohnedainlkga PloKid psa,roabmseetrevrse dwmerea xliismteudm inc Toanbcleen 1t.r aBteicoanusaen ddocuobrlree pspeaoknsd winegret ime
observed in the C–T curves of most of the alkaloids, observed maximum concentration and
valueswereusedasC andT . AfteroraladministrationofFZJextract,plasmaconcentrations
max max
of8c-ooxrroecsoppontidsiinnge atinmde 8v-oalxuoeesp wibeerreb eursiende aast mCmoasx tatnimd eTmpaox.i nAtsftewr eorreabl ealdomwinthisetrLatLioOnQ o,ft hFeZrJe efoxrtreatcht,e PK
plasma concentrations of 8-oxocoptisine and 8-oxoepiberberine at most time points were below the
parameterscouldn’tbecalculated.
LLOQ, therefore the PK parameters couldn’t be calculated.
IndrugPKstudies,theC–TcurvemayrevealqualitativecharacteristicsofPKprofiles,whilePK
In drug PK studies, the C–T curve may reveal qualitative characteristics of PK profiles, while
parameters such as AUC, MRT, C and t , can reflect the systemic exposure level of the drug
PK parameters such as AUC, MRTm,a Cxmax and1 /t21/2, can reflect the systemic exposure level of the drug
quantitatively. The PK results of the alkaloids of ZJ and FZJ were consequently discussed from
quantitatively. The PK results of the alkaloids of ZJ and FZJ were consequently discussed from the
theaabboovvee ttwwoo aaspspecetcst.s .

●: :CC–T– Tcucurvrevse soof ftthhee aallkkaaloloididss aafftteerr oorraall aaddmmiinniissttrraatitoionn ooff ZZJJ.. (cid:4)■:: CC––TT ccuurrvveess ooff tthhee aalklkaallooiiddss aafftteerr oorraall
administrationofFZJ.
administration of FZJ.
FiguFirgeu3r.e C3.– CT–cTu cruvrevseos fotfh thee1 122a alklkaallooiiddss aafftteerr oorraall aaddmmininisitsrtartaiotino nofo ZfJZ aJnadn FdZJF eZxJtreaxcttr (anc t= (6n).= 6).
Molecules2017,22,214 6of18
Table1.PKparametersofthe12alkaloidsinratplasmaafteroraladministrationofZJandFZJextract(mean±SD,n=6).
Analytes Group AUC0→t(ng·h/mL) AUC0→∞(ng·h/mL) MRT0→t(min) MRT0→∞(min) t1/2(min) Tmax(min) Cmax(ng/mL) Tsec(min) Csec(ng/mL)
ZJ 68.73±10.82 71.59±10.72 641.65±62.14 735.50±91.37 480.47±115.11 300 14.13±7.75 90 4.96±2.29*
coptisine
FZJ 10.99±1.64 11.85±1.25 250.52±27.45 312.29±83.18 176.05±73.47 180 4.78±2.31 - -
ZJ 68.46±15.94 69.61±15.46 645.43±62.71 696.78±87.57 366.42±101.94 300 13.65±7.05 90 5.51±2.17*
epiberberine
FZJ 10.13±1.09 11.40±1.27 233.44±25.91 315.37±52.51 235.17±101.63 180 4.82±1.97 - -
ZJ 88.34±24.92 96.63±33.22 404.26±105.70 525.01±192.01 339.21±144.61 90 34.07±19.25 300 11.18±8.71*
palmatine
FZJ 18.19±4.98 18.90±5.50 231.35±21.68 269.46±38.99 150.78±103.54 180 9.74±3.53 30 6.25±2.89*
ZJ 277.48±50.15 299.84±55.27 387.56±67.86 514.76±164.64 378.28±194.53 90 109.40±48.27 300 44.68±21.28*
berberine FZJ 48.07±9.60 50.24±10.74 244.37±17.40 280.76±37.56 133.26±50.10 180 23.67±7.42 30 11.64±7.28*
ZJ 24.74±3.90 27.02±2.23 607.27±59.56 809.87±254.49 535.67±158.71 300 2.46±1.17 90 2.42±0.89
8-oxocoptisine
FZJ - - - - - - - - -
ZJ 14.30±2.32 16.63±3.22 450.77±59.37 740.15±138.92 592.65±274.26 90 2.38±1.14 300 1.85±0.99
8-oxoepiberberine
FZJ - - - - - - - - -
ZJ 133.19±18.49 140.06±17.83 529.35±75.19 662.05±153.25 644.66±169.16 90 29.64±15.15 300 16.29±8.38
noroxyhydrastinine
FZJ 4.95±0.65 5.37±0.73 181.73±16.42 314.79±172.07 243.81±175.97 30 3.17±1.00 - -
ZJ 45.14±7.46 47.27±7.36 402.95±56.32 513.14±109.81 595.01±123.69 90 10.34±5.45 300 9.03±4.31
corydaldine
FZJ 2.86±0.35 3.10±0.46 170.94±11.39 239.52±47.43 233.20±77.36 30 1.73±0.58 - -
ZJ 532.34±57.78 537.43±54.97 260.43±15.40 282.90±34.14 192.57±135.94 240 155.16±27.92 90 117.29±45.45
dehydroevodiamine
FZJ 274.77±23.19 285.60±24.08 279.59±56.33 357.86±114.71 301.96±172.11 90 85.27±13.37 - -
ZJ 31.71±2.94 31.84±3.11 315.19±44.82 359.80±45.66 141.31±71.63 60 9.59±4.22 300 5.54±3.14
evodiamine FZJ 10.83±1.77 11.72±2.06 178.31±22.86 254.67±42.11 227.88±62.98 30 9.10±4.79 180 2.21±0.85
ZJ 11.54±1.40 12.77±1.43 231.61±27.74 313.20±68.30 218.14±89.78 90 3.45±1.73 20 2.55±0.38
wuchuyuamide-I
FZJ 48.25±8.86 50.11±8.81 346.41±46.61 413.10±94.85 295.01±109.89 30 10.73±4.43 - -
ZJ 55.65±12.40 59.02±13.34 251.37±29.26 390.99±119.47 539.67±383.48 60 24.20±12.95 240 7.36±4.89
evocarpine
FZJ 17.84±3.99 19.57±4.79 422.27±69.79 557.88±117.71 414.88±60.07 30 4.07±0.44 - -
TsecandCsecwerethetimeanddrugconcentrationofthesecondarypeakinC–Tcurve.*p<0.05comparedwithCmax.
Molecules2017,22,214 7of18
2.4.1. C–TCurvesoftheAlkaloids
DoublepeaksinC–TcurveswereobservedrepeatedlyinthePKstudyofZJandrelatedTCM
prescriptions,butexplanationstothephenomenonwerestillambiguous. Generally,possiblereasons
forthedoublepeakphenomenoncouldbesummarizedasfollows:(a)enterohepaticcirculation[36,37];
(b) two different sites of drug absorption [38–40]; (c) irregular pattern of gastric emptying [41–44].
Inthepresentstudy,however,thefactorscontributetodoublepeaksofQPAsandtheotheralkaloids
maybedifferent.
(1) For the alkaloids except QPAs, double peaks were visibly observed in the C–T curves after
oraladministrationofZJextract,butthesecondarypeakswereattenuatedandtheC–Tcurves
approximately matched the two-compartment model after administration of FZJ (calculated
by two fitting methods using DAS 2.0 software, Table 2). Moreover, T of the alkaloids
max
afteradministrationofZJextractweredeferredfromthatofFZJadministration. Alkaloidsof
CR, especially berberine, have been reported to have inhibition effect on GI motility [45–49].
Itprobablyplayamajorroleinthedoublepeakphenomenon,sincetheGImotilitycanaffect
thedrugabsorption. ThecontentofberberineinZJextractwere20timesmorethanthatinFZJ
extract(TableS4),leadingtomoresignificantinfluenceontheGImotilityafteroraladministration.
Thus,thedoublepeakphenomenonandthedelayedT ofthealkaloidsafteradministration
max
ofZJextractcouldbeattributedtotheinhibitioneffectonGImotilitycausedbyCRalkaloids,
especiallyberberine.
Table2.Two-compartmentmodelfittingresultofthealkaloidsinratplasmaafteroraladministration
ofZJandFZJextract.
ZJ FZJ
Analytes
r2 AIC r2 AIC
noroxyhydrastinine 0.313±0.350 118.137±17.623 0.802±0.128 16.505±12.486
corydaldine 0.243±0.267 85.431±9.718 0.805±0.066 3.947±8.187
dehydroevodiamine 0.324±0.280 154.46±11.416 0.886±0.055 107.718±6.140
evodiamine 0.428±0.169 71.547±6.327 0.853±0.232 30.067±6.112
wuchuyuamide-I 0.544±0.140 35.057±11.851 0.519±0.407 66.844±7.076
evocarpine 0.734±0.201 78.232±11.393 0.740±0.167 27.61±10.527
r2:correlationcoefficientmethod;AIC:Akaike’sinformationcriterionmethod.
(2) InZJgroup,twoplasmaconcentrationpeakswereobservedat90and300minforallthefourQPAs;
but the primary plasma concentration peak was at 300 min for coptisine and epiberberine,
and 90 min for palmatine and berberine (Table 1, significant differences between the plasma
concentrationvaluesoftheprimaryandsecondarypeakswereexaminedwitht-testbyMicrosoft
Office Excel 2007). Previous researches had suggested that the primary elimination route of
berberine invivo was renal excretion [50], and its C–T curve matched the two-compartment
modelfollowinganintravenousadministration[50,51]. Therefore,theenterohepaticcirculation
couldn’t be the major reason of the double peak phenomenon of QPAs, since the plasma
concentrationvaluesofcoptisineandepiberberineat300minweregreaterthanthatat90min.
Inaddition,researchfindingsofthemainphaseImetabolismofberberine[52–54]certifiedthat
metabolismcouldn’tbethereasonofthedoublepeakphenomenonofQPAs. Theabsorptionrate
constant (Ka) of berberine and palmatine at jejunum had been reported to be greater than
that at ileum and colon, which means the upper part of the intestine was their dominant
absorptionsite[55–57]. Coptisine,incontrast,hadabetterabsorptionrateatcoloncompared
with jejunum [57]. Therefore two absorption sites with different Ka could give a reasonable
explanationaboutthedoublepeakphenomenonofthefourQPAs: plasmaconcentrationpeak
at90minwasmainlycausedbytheabsorptionattheupperpartoftheintestine, butmainly
Molecules2017,22,214 8of18
bytheabsorptionatileumandcolonwhilethatisat300min. Plasmaconcentrationpeaksof
QPAsinFZJgroupat30and180minwereearlierthanthatofZJgroup,andpeaksat30min
causedbytheabsorptionattheupperpartoftheintestinewereattenuated. Thepossiblereason
wasthatthelowerberberinecontentinFZJweakenedtheinhibitioneffectonGImotilityand
subsequentlyledtothedecreasedresidencetimeandabsorptionlevelofQPAsatupperpartof
theintestine.
2.4.2. SystemicExposureoftheAlkaloids
Because the alkaloids content in ZJ and FZJ were no longer linear with the proportion of
CR and EF after extraction process [58,59], AUC0→∞/D (ratio of AUC0→∞ and dose) and Cmax/D
(ratioofC anddose)werecalculatedtoestimatethedosemodifiedsystemicexposurelevelsof
max
the12alkaloids(Table3). MRT0→∞andt1/2valuessuggestedthatthe12alkaloidswerenoteliminated
rapidlyinvivo,whichindicatedthatabsorptionswerethemajorobstacletotheirsystemicexposure
level. Physicochemical properties of these alkaloids had been calculated in silico to predict their
humancolonadenocarcinomacell(Caco-2)permeabilityaccordingtothethree-propertybasedrule
(3PRule)[60].Polarsurfacearea(PSA),logarithmofthen-octanol/waterdistributioncoefficient(LogD),
molecularweight(MW),hydrogenbonddonors(HBD),hydrogenbondacceptors(HBA),numberof
freerotatablebonds(NRB)andmolarsolubility(S)ofthealkaloidswereretrievedfromSciFinder®
(AmericanChemicalSociety). DataoftheinsilicostudywerelistedinTableS5. Physicochemicaldata
of the four QPAs and dehydroevodiamine were not given because of the electric charge in their
structureswhichhinderedthecalculation.
Table3. DosemodifiedAUC0→∞, Cmax andAUCweightedfactors(Wi)ofthe12alkaloidsinrat
plasmaafteroraladministrationofZJandFZJextract.
ZJ FZJ
Analytes
AUC0→∞/D Cmax/D Wi* AUC0→∞/D Cmax/D Wi*
coptisine 0.40 0.08 0.09 1.80 0.72 0.12
epiberberine 0.62 0.12 0.09 1.91 0.81 0.11
palmatine 0.77 0.27 0.13 3.03 1.56 0.19
berberine 0.84 0.31 0.39 2.56 1.21 0.50
CRalkaloids
8-oxocoptisine 44.44 3.98 0.04 - - -
8-oxoepiberberine 97.25 13.92 0.02 - - -
noroxyhydrastinine 1687.47 357.11 0.18 268.50 158.50 0.05
corydaldine 1688.21 369.29 0.06 310.00 173.00 0.03
dehydroevodiamine 39.69 11.46 0.84 12.15 3.63 0.78
evodiamine 8.55 2.57 0.05 3.73 2.89 0.03
EFalkaloids
wuchuyuamide-I 172.57 46.62 0.02 278.39 59.61 0.14
evocarpine 35.17 14.42 0.09 13.16 2.74 0.05
Ddenotedthedosageofthe12alkaloids,seeTableS4;*CalculatedbyAUC-basedweightingmethod.
(1) The four QPAs from CR, especially berberine, had relatively high systemic exposure levels,
but their AUC0→∞/D and Cmax/D values were extremely low. Previous studies suggested
that absolute bioavailability of berberine was less than 1% [61,62], and it should be mainly
attributedtoitspoorabsorption. Transportexperimentshadconfirmedthatberberine,palmatine
andcoptisinehadpoorpermeabilityacrossCaco-2cellmonolayerwithapparentpermeability
coefficient(P )valuesbetween0.1and1.0×10−6cm/s[63]aspoorlyabsorbedcompounds[64].
app
TPAsandSIAs,incontrast,werecomponentswithlowcontentsinCRbuthigherAUC0→∞/D
andC /Dvaluesindicatingtheirbetterabsorptionproperties. 3PRulesuggestedthatthetwo
max
TPAsandtwoSIAshadfavorablephysicochemicalpropertiesforCaco-2permeability: PSA≤60,
MW≤400andLogD>−1. However,compoundsmustbedissolvedinwaterbeforepenetrating
theintestinalepithelialcells,andthesuitablevaluesofsolubility(LogS)rangedfrom0to−4[65].
Molecules2017,22,214 9of18
Therefore,systemicexposurelevelsoftwoTPAs,especially8-oxocoptisine,appearedtobelimited
bytheirlowersolubilitythanSIAs.
(2) EvodiamineanddehydroevodiamineweretwomajorIQAsinEF,buttheresultssuggestedthat
dehydroevodiaminehadahighersystemicexposurelevelregardlessofthedosemodification.
Our previous study demonstrated that both evodiamine and dehydroevodiamine had high
permeabilityacrossCaco-2cellmonolayerwithP valuesof2.32×10−5and1.26×10−5cm/s,
app
respectively[66],whichwereclosetothevaluespredictedby3PRule;butitwasalsofoundthat
thefeedingconcentrationofevodiaminewaslimitedduetoitspoorsolubilityinthetransport
experiment. Thus, its solubility should be the major obstacle to a higher systemic exposure
level for evodiamine (predicted LogS < −5), and the poor solubility can be attributed to its
flat, rigid, andunsaturatedstructure. Incontrast, thesolubilitiesofdehydroevodiamineand
wuchuyuamide-Iwereimprovedbecauseoftheinnersaltstructureorthebrokenringstructure.
(3) Effects on the absorption and elimination of the alkaloids from CR and EF were illustrated
bycomparingtheAUC0→∞/D,Cmax/Dandt1/2 ofthealkaloidsaftertheadministrationsof
ZJ and FZJ extract. Since the t values of some alkaloids, such as evodiamine, were not
1/2
accuratelycalculatedbecauseofthemulti-peakphenomenon,MRT0→∞wouldbemorereliablein
thecomparison. MRT0→∞ andt1/2valuesofthefourQPAshalvedofFZJgroupcomparedwith
thatofZJgroup,butCmax/DandAUC0→∞/Dincreased4–9timesand3–4times,respectively.
ThecomparisonsindicatedthatabsorptionsofQPAswereincreased,buttheireliminationswere
acceleratedasaresultofincreasedEFintake. TheQPAslikecoptisine,palmatineandberberine
wereP-gpsubstrates[63,67],andtheireffluxratiocouldbereducedbyEFontheCaco-2transport
model[68],soEFmaypossiblypromotetheabsorptionofQPAsbyinhibitingP-gp. Ontheother
hand, EF could also induce hepatic UDP-glucuronosyltransferase 1A1 and then accelerate
the elimination of QPAs [69]. In the present study, QPAs’ systemic exposure (AUC0→∞/D)
werefinallyincreasedwhentheproportionofEFincreased,suggestingthatEFhadagreater
influenceontheabsorptionthanontheeliminationofQPAsafterthecompatibilityofCRandEF.
ContrarytoQPAs,noroxyhydrastinine,corydaldine,dehydroevodiamineandevocarpinehad
a larger AUC0→∞/D and Cmax/D values in ZJ group than that in FZJ group. As argued above,
highcontentofberberineinZJextractcouldcauseinhibitioneffectonGImotility;anddrugabsorption
couldbesubsequentlyimprovedwhentheGImotilitywasslowedprobablybecauseofthelonger
residencetime[70–72],whichmightleadtotheincreasedabsorptionsandsystemicexposurelevelsof
noroxyhydrastinine,corydaldine,dehydroevodiamine.
2.4.3. IntegratedPKAnalysis
Among the 12 alkaloids, berberine and dehydroevodiamine had the most similar C–T curve
profileswiththatoftheintegratedC–TcurvesofCRandEFalkaloids(Figures3and4),indicatingthat
berberine and dehydroevodiamine could be regarded as the representative components to reflect
the PK behaviors of CR and EF alkaloids. The highest systemic exposure levels of berberine and
dehydroevodiamineamongthealkaloids,i.e.,thebiggestW values(Table3),maycontributetothis
i
result. However,theAUC0→∞/DandCmax/Dvaluesofberberinewereextremelylow,indicatingthat
large quantity of berberine was not absorbed. It has been reported that at least 56% of berberine
existedasprototypeduringtheperiodof0–36hafterintragastricadministration,andtherewas24%of
berberinestillremainedintheGItract evenat36h [62]. In recentyears, ZJhasbeeninvestigated
foritsantitumoractivities,anditsinvitroandinvivoantitumoractivitiesagainstdigestivesystem
cancerhavealreadybeenconfirmed[2–5]. Inaddition,ourpreviousstudiesshowedthatberberine,
coptisine, and evodiamine had the inhibitory activities against the proliferation of human gastric
carcinoma(NCI-N87)andCaco-2cellswiththehalfinhibitoryconcentrationof12.61–91.18µmol/L,
suggestingthattheywereprobablythemaineffectivecomponentsinZJagainstthedigestivesystem
cancer [9]. To epithelial tumor cells of digestive tract, oral drug could take effect directly without
the absorption process, though the unabsorbed drug has been considered as useless traditionally.
Molecules2017,22,214 10of18
Asforberberine,coptisine,andevodiamine,poorabsorptionwouldbeconductivetotheirlocaleffects
in digestive tract; besides, GI retention effect caused by CR alkaloids could extend the treatment
durationofthesealkaloids. Thereforethisresearchmaygiveasupporttotheanti-digestivesystem
cancerpotentialsofZJinthePKrespect.
Molecules 2017, 22, 214 10 of 18

FFiigguurree 44.. IInntteeggrraatteedd CC––TT ccuurrvveess ooff CCRR aanndd EEFF aallkkaallooiiddss ccaallccuullaatteedd bbyy AAUUCC--bbaasseedd wweeiigghhttiinngg mmeetthhoodd
((11)) aanndd ttoottaall ddrruugg ccoonncceennttrraattiioonn mmeetthhoodd ((22))a afftteerro orraalla addmmiinniissttrraattiioonno offZ ZJJa annddF FZZJJe exxttrraacctt..
TThhee ttwwoo iinntteeggrraatteedd mmeetthhooddss hhaadds simimilialarrC C––TTc cuurrvveessp prorofiflieles.s.H Hoowweevveer,rt, htheein intetgergartaetdedA AUUCC0→0→∞∞
and Cmax values of CR and EF alkaloids calculated by AUC-based weighting method were
andC valuesofCRandEFalkaloidscalculatedbyAUC-basedweightingmethodweresignificantly
max
significantly lower than total drug concentration method (examined using t-test by Microsoft Office
lowerthantotaldrugconcentrationmethod(examinedusingt-testbyMicrosoftOfficeExcel2007,
Excel 2007, Table 4), and even lower than the values of berberine and dehydroevodiamine before
Table 4), and even lower than the values of berberine and dehydroevodiamine before integrated.
integrated. Therefore the total drug concentration method was more concise and credible than the
ThereforethetotaldrugconcentrationmethodwasmoreconciseandcrediblethantheAUC-based
AUC-based weighting method to reflect the systemic exposure levels of CR and EF alkaloids.
weightingmethodtoreflectthesystemicexposurelevelsofCRandEFalkaloids.
Table 4. Integrated PK parameters of the CR and EF alkaloids in rat plasma after oral administration
Table4.IntegratedPKparametersoftheCRandEFalkaloidsinratplasmaafteroraladministrationof
of ZJ and FZJ extract (mean ± SD, n = 6).
ZJandFZJextract(mean±SD,n=6).
ZJ FZJ
PK Parameters Integrated Method
CR AlkaloidsZ J EF Alkaloids CR Alkaloids FZJEF Alkaloids
PKParameters IntegratedMethod
A CR16A5.l8k5a ±lo 2id5.s24 E4F54A.0lk2 a±l o4i9d.5s0 C3R1.0A1l k± a4l.o3i8d s 22E2.F33A ±lk 1a8l.o8i7d s
AUC0→t (ng·h/mL) AB 16754.805.5±8 ±2 59.62.429 465342.0.255± ±4 699.5.901 9371.2.041 ±± 124..0338 35222.27.43 3± ±311.982.8 7
AUC0→t(ng·h/mL) BA 74107.548.9±0 ±9 62.32.960 643528.5.550± ±6 496.9.918 9376..2942 ±± 81.26.10 3 23305.29.57 4± ±193.219.9 2
AUC0→∞ (ng·h/mL) B 790.50 ± 129.01 * 640.61 ± 64.83 * 113.78 ± 19.69 * 368.20 ± 33.08 *
A 174.90±23.60 458.50±46.98 36.92±8.61 230.95±19.29
AUC0→M∞RT(n0→g·t h(m/minL)) BAB 79044.569058..±44871 ±±2 965.2401..52*64 6422066.6021..58±35 ±±6 411.578..35149* 122154313...727843 ±±± 111979...669922 * 223896258..21.2920 ±±± 54593..303.620 8*
A 465.48±62.56 260.53±15.54 251.24±19.62 282.29±55.06
MRMTR0→Tt0→(m∞ (inm)in) BAB 4945861.4437..95±97 ±±5 465.532..41755 222689233.8..57632± ±±1 34740.1..87957 222574423...907593 ±±± 671027..96.9442 3357923.99.540.81 ± 2± 1 ±91014..7925.73 2
A 464.99±65.15 283.73±34.85 254.95±60.94 359.90±111.27
MRT0→t1/∞2 ((mmiinn)) BA 51533.75.702± ±5 430.785.47 219935..6022± ± 4103.67.785 124712..6029 ±± 3712.7.654 3033.7134.4 ±8 1±629.08.87 5
B 573.75 ± 447.62 218.36 ± 137.70 186.77 ± 91.91 334.42 ± 104.19
A 537.02±408.47 195.02±136.85 141.62±31.75 303.14±162.88
t1/T2m(max i(nm)in) BA 573.75±9404 7.62 218.36±24103 7.70 186.7178±0 91.91 334.4920± 104.19
B 90 240 180 90
A 90 240 180 90
TCmamxax( m(nign/)mL) BA 54.1940 ± 22.14 131.1284 0± 23.45 14.821 8±0 4.70 67.02 ± 9100.36
B 198.70 ± 76.91 * 167.10 ± 28.40 * 43.79 ± 14.12 * 92.367 ± 12.51 *
A 54.14±22.14 131.18±23.45 14.82±4.70 67.02±10.36
CmaxT(snegc/ (mmLin)) BA 198.70±30706 .91* 167.10±902 8.40* 43.793±0 14.12* 92.367- ±12.51*
B 300 90 30 -
Tsec(min) AA 24.93900 ± 10.08 99.909 0± 38.75 7.51 3±0 4.19 - -
Csec (ng/mL) B 300 90 30 -
B 113.27 ± 40.70 138.18 ± 54.42 25.35 ± 10.26 -
A 24.99±10.08 99.90±38.75 7.51±4.19 -
CAse cd(enngo/tmedL) the AUC-based weighting method, and B denoted the total drug concentration method.
B 113.27±40.70 138.18±54.42 25.35±10.26 -
* p < 0.05 compared with method A.
AdenotedtheAUC-basedweightingmethod,andBdenotedthetotaldrugconcentrationmethod. *p<0.05
comparedwithmethodA.
3. Experimental Section
3.1. Chemicals and Materials
The rhizomes of Coptis chinensis Franch. (CR) were collected from “The National GAP Base of
Chinese Materia Medica for Coptidis Rhizoma” at Shizhu County (Chongqing, China). The nearly
ripe fruits of Euodia rutaecarpa (Juss.) Benth. (EF) were collected from Xiangtan city (Hunan, China).
The two herbs were identified by Prof. Xiu-Wei Yang from School of Pharmaceutical Sciences,