Effect of pregnancy on nitrofurantoin disposition in mice

PHARMACOKINETICS, PHARMACODYNAMICS ANDDRUG METABOLISM

Effect of Pregnancy on Nitrofurantoin Disposition in Mice

YI ZHANG, LIN ZHOU, JASHVANT D. UNADKAT, QINGCHENG MAO

Department of Pharmaceutics, School of Pharmacy, University of Washington, Seattle, Washington 98195-7610

Received 18 August 2008; revised 14 December 2008; accepted 28 December 2008

Published online 6 May 2009 in Wiley InterScience (www.interscience.wiley.com). DOI 10.1002/jps.21698

Abbreviationtein; ABCG2, APBS, phosphatestandard; AUCcurve; CL, totahalf-life; Vss, steability.

Corresponden0355; Fax: 206-

Journal of Pharm

� 2009 Wiley-Liss

4306 JOURN

ABSTRACT: We investigated the effect of pregnancy on nitrofurantoin (NFT) disposi-tion in wild-type and Bcrp1�/� mice. Pregnant and non-pregnant mice were adminis-tered NFT intravenously (5 mg/kg) or orally (10 mg/kg). Blood samples were collected atvarious times (5–60 min) after drug administration, plasma NFT concentrations deter-mined by HPLC/UV, and pharmacokinetic parameters estimated. Dose-normalized areaunder the plasma concentration–time curve (AUC), terminal plasma half-life (T1/2), totalplasma clearance (CL), and steady-state volume of distribution (Vss) of intravenous NFTin wild-type or Bcrp1�/� mice were not altered by pregnancy. After oral administration,pregnancy did not affect dose-normalized AUC of NFT in wild-type mice; however, dose-normalized AUC in Bcrp1�/� mice was decreased by approximately 70% by pregnancy.In conclusion, since Bcrp1 plays a minor role in the systemic clearance of NFT in femalemice, pregnancy did not affect disposition of intravenous NFT despite the fact that Bcrp1expression in the liver and kidney of mice is significantly induced by pregnancy. On theother hand, pregnancy may affect expression and activity of certain intestinal effluxtransporters and/or metabolic enzymes in Bcrp1�/� mice, resulting in a drastic decreasein the systemic exposure of oral NFT in pregnant Bcrp1�/� mice. � 2009 Wiley-Liss, Inc.

and the American Pharmacists Association J Pharm Sci 98:4306–4315, 2009

Keywords: nitrofurantoin disposition;
pregnancy; Bcrp1/Abcg2; and mice
INTRODUCTION

A large number of pregnant women are subjectsof drug therapy for various diseases that requiretreatment during pregnancy, including viral,

s used: BCRP, breast cancer resistance pro-TP-binding cassette G2; NFT, nitrofurantoin;-buffered saline; gd, gestation day; IS, internal, the area under plasma concentration–timel plasma clearance; T1/2, terminal plasmaady-state volume of distribution; F, bioavail-

ce to: Qingcheng Mao (Telephone: 206-685-543-3204; E-mail: [email protected])

aceutical Sciences, Vol. 98, 4306–4315 (2009)

, Inc. and the American Pharmacists Association

AL OF PHARMACEUTICAL SCIENCES, VOL. 98, NO. 11, NO

fungal and bacteria infection, epilepsy, hyperten-sion, and gestational diabetes.1 Drug therapyduring pregnancy could be complicated by the factthat pregnancy may induce changes in pharma-cokinetics by altering factors that can influencethe absorption, distribution, and elimination ofdrugs.2 These factors include plasma albuminconcentration and thus plasma protein bindingof drugs,3 glomerular filtration rate,4 and theexpression and activity of drug metaboliz-ing enzymes5–10 and possibly transporters aswell.10–13 It is thus important to understand themechanisms by which drug disposition is alteredby pregnancy so that guidelines for drug useduring pregnancy may be optimized.

VEMBER 2009

EFFECT OF PREGNANCY ON NITROFURANTOIN DISPOSITION IN MICE 4307

The breast cancer resistance protein (BCRP/ABCG2) is an ATP-binding cassette (ABC) effluxtransporter belonging to the subfamily G ofthe large ABC transporter superfamily.14 BCRPtransports a wide variety of substrates, rang-ing from chemotherapeutic agents to organicanions.15,16 Owing to its high level expression inorgans important for drug disposition (e.g., theliver, small intestine, and the blood–brain andplacental barriers),14,17–19 the role of BCRP inabsorption, distribution, and elimination of drugshas been demonstrated in numerous studies.20–25

We have previously shown that the expressionof Bcrp1, the murine homolog of human BCRP, invarious organs of pregnant mice changes withgestational age.12 The protein levels of Bcrp1 inthe liver, kidney, and placenta of pregnant mice atgestation day (gd) 15 were significantly increasedthree to fourfold relative to those at gd 10, gd 19 orin non-pregnant mice (term in mice is approxi-mately 19–21 days). On the other hand, theprotein levels of Bcrp1 in the small intestine werenot significantly influenced by pregnancy.12

Since BCRP/Bcrp1 plays a significant role in thedisposition of substrate drugs, the pregnancy-induced change in Bcrp1 expression may lead toalterations in the pharmacokinetics of drugsduring pregnancy compared with non-pregnancy.

Therefore, in the present study, we system-atically investigated the effect of pregnancy onnitrofurantoin (NFT) disposition in mice andthe potential impact of Bcrp1. For several reasonsNFT was selected as a model BCRP/Bcrp1substrate. First, NFT is an antibiotic commonlyused by pregnant women to treat urinary tractinfection.26 Hence, the data obtained from ananimal model may have implications for NFTdisposition in pregnant women. Second, NFT isa selective substrate for BCRP/Bcrp1, but not forP-gp and MRP2.27 Third, NFT has been validatedas an in vivo probe for Bcrp1 function in mice andrats.20,27,28

MATERIALS AND METHODS

Materials

NFT and furanzolidone were purchased fromSigma (St. Louis, MO). Polyethylene glycol400 (PEG 400) was obtained from SpectrumLaboratory Products, Inc. (Gardena, CA).HPLC-grade acetonitrile and methylene chloridewere from Fisher Scientific Co. (Morris Plains,NJ). PBS was from Gibco (Carlsbad, CA). Micro-

DOI 10.1002/jps JOURNA

con Ultracel YM-10 centrifuge tubes were fromMillipore (Billerica, MA).

Animal Studies

FVB wild-type mice and the first generation ofBcrp1�/� mice with FVB genetic background werepurchased from Taconic (Hudson, NY). Under abreeding license purchased from Taconic, maleand female Bcrp1�/� mice were mated to generateoffspring, which were used in the subsequentanimal experiments. Pregnant and non-pregnantwild-type and Bcrp1�/� mice were cared for inaccordance with the United Stated Public HealthService policy for the Care and Use of LaboratoryAnimals. The animal studies were approved bythe Institutional Animal Care and Use Committeeat the University of Washington. The mice hadfree access to food (a standard diet) and water, andwere maintained on a 12:12 h automatically timedlight/dark cycle. Male mice of 7–9 weeks of agewere mated with female mice of the same age.Female mice that demonstrated sperm plugwere separated and housed in new cages. Gesta-tional age was calculated based on the estimatedtime of insemination (presence of sperm plug as gd0). Progress of pregnancy was regularly monitoredby visual inspection. Pregnant mice at gd 15 wereused, as we have previously shown that theprotein levels of Bcrp1 in the liver and kidney ofpregnant mice peak at gd 15.12

For intravenous administration, NFT wasdissolved in 10% (v/v) ethanol, 40% (v/v) PBS,and 50% (v/v) PEG 400 at a concentration of1.5 mg/mL. Under anesthesia (isoflurane), thepregnant and non-pregnant wild-type or Bcrp1�/�

mice (at gd 15) were administered NFT by retro-orbital injection (5 mg/kg body weight). For oralgavage administration (10 mg/kg body weight),NFT was dissolved in 50% (v/v) ethanol and 50%(v/v) PEG 400 at a concentration of 3.3 mg/mL.At various times (5, 10, 20, 30, 40, and 60 min)after drug administration (n¼ 3 per time point),animals were sacrificed under anesthesia (iso-flurane) by cardiac puncture. Blood samples werecollected in heparinized microtainer tubes (BectonDickinson, Franklin Lakes, NJ) and centrifuged.Blank mouse plasma samples were collected fromundosed mice. Plasma samples were harvestedand stored at �208C until analysis.

NFT HPLC/UV Assay

NFT concentrations in the plasma samples frompregnant and non-pregnant mice were deter-

L OF PHARMACEUTICAL SCIENCES, VOL. 98, NO. 11, NOVEMBER 2009

4308 ZHANG ET AL.

mined by a validated HPLC/UV assay as describedpreviously.20 Plasma concentrations of NFT inpregnant wild-type and pregnant Bcrp1�/� mice,after intravenous administration, were obtainedfrom our previously published study.20 This isindicated in the corresponding figure legend andtable.

Pharmacokinetic Data Analysis

Pharmacokinetic parameters were estimatedusing pseudoprofile-based bootstrap method aspreviously described.29 Briefly, in the firststep, plasma samples (one sample per mouse)obtained at various time points (5–60 min) wereused to generate 50 composite plasma NFTconcentration–time profiles for each mouse group(e.g., pregnant or non-pregnant wild-type mousegroup) by selection of one point from three pointsat each time, permitting random replacement,and construction of 50 pseudo-profiles. Thisresampling with random replacement was pro-grammed and performed in R language. Areaunder the plasma concentration–time curve(AUC) from time 0 to infinity (AUC0–1) and from0 to 60 min (AUC0–60 min), terminal plasma half-life (T1/2), and steady-state volume of distribution(Vss) after intravenous or oral administrationwere then estimated using non-compartmentalapproach (WinNonLin 3.2, Pharsight, MountView, CA) for each pseudo-profile. AUC0–60 min

was calculated using the linear trapezoidalmethod. This generated an estimate of thedistribution of the above pharmacokinetic para-meters. In the second step, resampling withrandom replacement of three samples each timefrom the ‘‘empirical’’ distribution was repeated 50times, and 50 estimates of the mean of eachparameter were calculated. Finally, the meansand standard deviations of the pharmacokineticparameters were calculated using the estimatesfrom the second resampling technique. Bioavail-ability (F) of orally administered NFT wasestimated as the ratio of (AUC0–1)oral/Doseoral

and (mean AUC0–1)intravenous/Doseintravenous.Pregnant or non-pregnant mice were adminis-

tered NFT based on their body weight (5 and 10mg/kg for intravenous and oral administration,respectively). Since pregnant mice usually havemultiple litters (6–14 fetuses), this results in thetotal pregnant body weight being nearly 1.5 timesthe non-pregnant body weight. Hence, pregnantmice received a larger amount of NFT than non-pregnant mice due to increased body weight,

JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 98, NO. 11, NOVEMBER 20

resulting in a larger AUC. Mean�SD of thebody weight of non-pregnant wild-type, pregnantwild-type, non-pregnant Bcrp1�/�, and pregnantBcrp1�/� mice were 21.2� 2.2, 32.3� 3.6,23.2� 2.6, and 31.9� 4.9 g, respectively. Theincrease in body weight of pregnant mice com-pared to non-pregnant mice is mainly due to theweight of fetuses and placenta. We previouslyshowed that NFT in the fetuses of pregnant miceonly accounted for a small fraction of the totalamount of NFT in the body,20 suggesting that thefetus is not a major site for NFT distribution.We estimated total plasma clearance (CL) anddose-normalized AUC using the followingequations

CL ¼ mean actual dose of the respective mouse group

AUC0�1

Dose � normalized AUC

¼ AUC

mean actual dose of the respective mouse group

Determination of Unbound Fractions of NFT inMouse Plasma

Unbound fractions of NFT in plasma samplesfrom pregnant and non-pregnant wild-type orBcrp1�/� mice were determined using ultrafiltra-tion. Briefly, 400 mL of blank mouse plasma werespiked with 5 mg/mL NFT and the mixtures wereincubated for 2 h at 378C. Blank human plasmasamples were also incubated in parallel for 2 h at378C. Then, the NFT-mouse plasma mixtures orblank human plasma samples were loaded intocentrifuge tubes (Microcon Ultracel YM-10, Milli-pore) and centrifuged at 1000g for 5 min at 378C.The filtrates were discarded and the remainingmixtures were centrifuged again for 5 min underthe same condition. The second filtrates werecollected for analysis. The internal standard (IS),furazolidone, was then added into 20 mL of thefiltrates from spiked mouse plasma samplesto achieve a final concentration of 1 mg/mL. Inparallel, IS was added into 20 mL to the filtratesfrom the blank human plasma samples to achievea final concentration of 1 mg/mL, and NFTwas also added to achieve the concentration range0.1–5 mg/mL to generate a calibration curve (eachcalibrator was assayed in triplicate). NFT con-centrations in the filtrates (Cu) of mouse plasmasamples were then determined by the HPLC/UVassay. Unbound fractions ( fu) of NFT in spiked

09 DOI 10.1002/jps


mouse plasma samples were calculated using thefollowing equation:

fu ¼ Cu

5 mg=mL� 100%

NFT concentration of 5 mg/mL was used for thefollowing reasons. First, the unbound fraction ofNFT in human plasma was not saturated withinthe 1–10mg/mL range (data not shown). As a weakacid, NFT mostly binds to albumin in plasma, andalbumin concentrations in human and miceplasma are similar.30 Thus, binding of NFT toprotein in mouse plasma at 5 mg/mL is expected tobe non-saturable. Second, 5 mg/mL NFT waswithin the range of mouse plasma concentrationsachievable in vivo.20

Statistical Analysis

All the estimated pharmacokinetic parameterswere presented as mean�SD. Statistical signifi-cance, when indicated, was analyzed using theStudent’s t-test, and differences with p-values<0.05 were considered significant.

RESULTS

Effect of Pregnancy on NFT Disposition in Mice afterIntravenous Administration

After intravenous administration, dose-normal-ized NFT plasma concentration–time profiles inwild-type or Bcrp1�/� pregnant and non-pregnantmice were comparable (Fig. 1). As a result, nomatter whether the mice were pregnant or non-pregnant, dose-normalized AUCs in wild-type orBcrp1�/� mice did not change much (Tab. 1).These results indicated that pregnancy did notaffect the systemic exposure of NFT in eitherwild-type or Bcrp1�/� mice after intravenousadministration. We also estimated T1/2, CL, andVss. As would be expected from the plasma con-centration–time profiles (Fig. 1), these pharma-cokinetic parameters were also comparable inpregnant and non-pregnant wild-type or Bcrp1�/�

mice after intravenous administration (Tab. 1).

Effect of Pregnancy on NFT Disposition in Miceafter Oral Administration

After oral administration, dose-normalized NFTplasma concentrations in non-pregnant Bcrp1�/�

mice were drastically increased compared to thosein non-pregnant wild-type mice (Fig. 2A). Dose-normalized AUCs of NFT in non-pregnant Bcrp1�/�


mice were approximately 2.5 times greater thanthose in non-pregnant wild-type mice (Tab. 2).Consequently, bioavailability (F) of NFT in non-pregnant Bcrp1�/� mice was increased to 71%from 41% in non-pregnant wild-type mice (Tab. 2).This suggests that Bcrp1 plays a significantrole in restricting oral bioavailability of NFT innon-pregnant mice.

Unlike the plasma concentrations of NFTin non-pregnant wild-type mice which peaked at10–20 min (Fig. 2A), the plasma concentrations ofNFT in non-pregnant Bcrp1�/� mice did notdecrease much until 40 min (Fig. 2A). Thus, theabsorption process of oral NFT seemed to besignificantly slowed in non-pregnant Bcrp1�/�

mice compared to wild-type mice. The plasmaconcentrations of NFT in pregnant Bcrp1�/� micewere low and did not change much over time(Fig. 2B), suggesting that a significant and clearelimination phase of oral NFT in pregnantBcrp1�/� mice did not exist during the course ofsampling. Therefore, except AUC0–60 min forpregnant Bcrp1�/� mice and AUC0–60 min andAUC0–1 for non-pregnant Bcrp1�/� mice, otherpharmacokinetic parameters of oral NFT such asCL could not be accurately estimated in thisstudy due to lack of sufficient data points inthe terminal elimination phase. We also onlyestimated AUC0–60 min and AUC0–1 values of oralNFT for pregnant and non-pregnant wild-typemice (Tab. 2).

Dose-normalized NFT plasma concentration–time profiles of pregnant and non-pregnantwild-type mice were similar (Fig. 2C), leading tocomparable dose-normalized AUCs in these mice(Tab. 2). Bioavailability of oral NFT in wild-typemice was around 40–55%, which was not drama-tically altered by pregnancy (Tab. 2). However,dose-normalized NFT plasma concentrationsin pregnant Bcrp1�/� mice were drasticallydecreased compared to those in non-pregnantBcrp1�/� mice (Fig. 2D), resulting in a 70% de-crease in dose-normalized AUC0–60 min in preg-nant Bcrp1�/� mice compared to non-pregnantBcrp1�/� mice (Tab. 2). Dose-normalized AUC0–60 min

in pregnant Bcrp1�/� mice was slightly lower thanthat in pregnant wild-type mice (Tab. 2).

Unbound Fractions of NFT in Pregnant andNon-Pregnant Mouse Plasma

Pregnancy may alter NFT disposition throughchanging plasma protein binding. We thereforemeasured plasma protein binding of NFT in


Table 1. Pharmacokinetic Parameters of NFT in Pregnant and Non-Pregnant Mice after IntravenousAdministration (5 mg/kg)

Wild-Type Mice Bcrp1�/� Mice

Pregnant Non-Pregnant Pregnant Non-Pregnant

Dose-normalized AUC0–60 min [(mg �min/mL)/mg] 0.70� 0.05 0.83� 0.13 0.95� 0.09 1.25� 0.20Dose-normalized AUC0–1 [(mg �min/mL)/mg] 0.70� 0.05 0.84� 0.08 1.03� 0.09 1.30� 0.14T1/2 (min) 9.90� 1.79 8.24� 0.86 17.1� 6.13 10.1� 1.7CL (mL/min) 1.44� 0.09 1.28� 0.11 0.99� 0.10 0.84� 0.12Vss (mL) 16.7� 1.6 13.82� 1.8 21.21� 3.81 11.9� 2.5

NFT was administered by intravenous (retro-orbital) injection. Plasma samples were collected, NFT concentrations determinedby HPLC/UV, and pharmacokinetic parameters estimated as described in Materials and Methods Section. Data shown aremean�SD. The plasma concentrations of NFT in pregnant wild-type and pregnant Bcrp1�/� mice after intravenous administrationwere taken from our published study.20

Figure 1. NFT disposition in mice after intravenous administration. Dose-normalizedplasma concentration–time profiles of NFT in non-pregnant wild-type and Bcrp1�/� mice(A), in pregnant wild-type and Bcrp1�/� mice (B), in pregnant and non-pregnant wild-type mice (C), and in pregnant and non-pregnant Bcrp1�/� mice (D). Mice wereadministered NFT (5 mg/kg body weight) intravenously by retro-orbital injection.Plasma samples were collected and NFT concentrations determined as described inMaterials and Methods Section. Data shown are mean�SD (n¼ 3 per time point). Dose-normalized NFT plasma concentrations were calculated by dividing the NFT plasmaconcentrations by the mean actual dose of the respective mouse group. The NFT plasmaconcentrations in pregnant wild-type and pregnant Bcrp1�/� mice after intravenousadministration were obtained from our previously published study.20 Plasma concen-trations of NFT in pregnant and non-pregnant mice are presented with dashed and solidlines, respectively.

JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 98, NO. 11, NOVEMBER 2009 DOI 10.1002/jps

4310 ZHANG ET AL.

Figure 2. NFT disposition in mice after oral administration. Dose-normalizedplasma concentration–time profiles of NFT in non-pregnant wild-type and Bcrp1�/�

mice (A), in pregnant wild-type and Bcrp1�/� mice (B), in pregnant and non-pregnantwild-type mice (C), and in pregnant and non-pregnant Bcrp1�/� mice (D). Mice wereadministered NFT (10 mg/kg body weight) by oral gavage. Plasma samples werecollected and NFT concentrations determined as described in Materials and MethodsSection. Data shown are mean�SD (n¼ 3 per time point). Dose-normalized NFT plasmaconcentrations were calculated as described in Figure 1. Plasma concentrations ofNFT in pregnant and non-pregnant mice are presented with dashed and solid lines,respectively.


pregnant and non-pregnant mouse plasma. Asshown in Table 3, unbound fractions of NFTin pregnant and non-pregnant Bcrp1�/� mouseplasma were 38.0� 4.7% and 46.4� 1.9%, respec-tively, and were not significantly different. The

Table 2. Pharmacokinetic Parameters of NFT in Pregnan(10 mg/kg)

W

Pregna

Dose-normalized AUC0–60 min [(mg �min/mL)/mg] 0.34� 0Dose-normalized AUC0–1 [(mg �min/mL)/mg] 0.42� 0F 0.55� 0

NFT was administered by oral gavage. Plasma samples werepharmacokinetic parameters estimated as described in Materials andCL, and Vss for all animal groups and AUC0–1 for pregnant Bcrp1�/�

sufficient data points in the terminal elimination phase. NE, not es


unbound fraction of NFT in pregnant wild-typemice (43.6� 3.7%) was slightly but significantlylower than that in non-pregnant wild-type mice(64.4� 6.3%). Unbound fraction of NFT in humanplasma was approximately 40%.31

t and Non-Pregnant Mice after Oral Administration

ild-Type Mice Bcrp1�/� Mice

nt Non-Pregnant Pregnant Non-Pregnant

.03 0.33� 0.03 0.23� 0.04 0.84� 0.05

.03 0.35� 0.02 NE 0.91� 0.05

.04 0.41� 0.04 NE 0.71� 0.04

collected, NFT concentrations determined by HPLC/UV, andMethods Section. Data shown are mean�SD. The data of T1/2,

mice after oral administration were not estimated due to lack oftimated.


Table 3. Unbound Fractions of NFT in the Plasma ofPregnant and Non-Pregnant Wild-Type or Bcrp1�/�

Mice

Wild-Type Bcrp1�/�

Pregnant mice 43.6� 3.7 38.0� 4.7Non pregnant mice 64.4� 6.3 46.4� 1.9p-value 0.013� 0.075

Unbound fractions of NFT were determined using ultrafil-tration as described in Materials and Methods Section. Datashown are mean�SD from three independent determinations.Significant difference between pregnant and non-pregnantwild-type mice: �p< 0.05 by the Student’s t-test.

4312 ZHANG ET AL.

DISCUSSION

While the bioavailability of NFT in humans wasapproximately 90%,31 the bioavailability of NFTin rodents and rabbits was relatively low (e.g.,30% in rabbits).32,33 In the present study, weshowed that the oral bioavailability of NFT inwild-type female mice was 40–55% (Tab. 2), whichis in the same range as previously reported.33 Thestudies using rat liver by Jonen34 suggested aminor role of hepatic metabolism in the elimina-tion of NFT in rats. Moreover, Watari et al.32

showed that the bioavailability of NFT in rabbitsfollowing intraduodenal administration and por-tal vein infusion was nearly unity. Thus, theseauthors concluded that the loss of dose of oral NFTin rabbits was likely due to incomplete absorptionand/or gastric degradation (e.g., through nitror-eduction by the intestinal flora), rather thanmetabolism in the intestine and/or the hepaticfirst-pass effect.32 NFT can readily cross tissuemembranes by passive diffusion,35 which cannotexplain the relatively low bioavailability of NFTin rodents either. Thus, efflux transportersexpressed in the apical membrane of the intestinalepithelium could play a role in limiting oralabsorption of NFT. BCRP/Bcrp1 is highlyexpressed in the apical membrane of the intestinalepithelium.17,36 Consistent with this tissue dis-tribution pattern, Merino et al.27,33 demonstratedthat AUC of orally administered NFT in Bcrp1�/�

mice was significantly increased compared withthat in wild-type mice. Likewise, we also showedthat the systemic exposure of oral NFT wasincreased 2.5-fold in non-pregnant Bcrp1�/� micecompared to non-pregnant wild-type mice, leadingto an increase in oral bioavailability of NFT(Fig. 2A and Tab. 2). These results suggestthat Bcrp1 restricts oral bioavailability of NFT


in non-pregnant mice, presumably through limit-ing intestinal absorption of the drug. A similarrole of Bcrp1 in reducing absorption of oral NFT inrats has also been demonstrated.28

Merino et al.27,33 reported a nearly twofoldincrease in AUC of intravenous NFT in Bcrp1�/�

male mice compared to wild-type male mice.However, we noted that the lack of Bcrp1increased the systemic exposure (dose-normalizedAUCs) of NFT by only approximately 50% in non-pregnant mice after intravenous administration(Fig. 1A and B, and Tab. 1). This suggeststhat Bcrp1 does not seem to play a major role inthe systemic clearance of NFT in female mice.Consistent with this conclusion, pregnancy had noeffect on the systemic clearance of NFT in miceafter intravenous administration (Fig. 1C and D,and Tab. 1), even though we have previouslyshown that the Bcrp1 protein levels in the liverand kidney of pregnant mice are induced three tofourfold compared to non-pregnant mice.12 Thisobservation may be explained as follows. First,Bcrp1 did not have a substantial effect on renalexcretion of NFT in mice even though it is highlyexpressed in mouse kidney.27 Second, hepato-biliary excretion of NFT in mice only accounted forless than 10% of the total dose of NFT admini-strated intravenously.27 The remaining fraction ofthe dose was mainly eliminated by nitroreductionoccurring in body tissues. Thus, as in humans,37,38

nitroreduction in tissues is likely a majorelimination route of NFT in mice. Third, Merinoet al.27,33 demonstrated a twofold increase inAUC of intravenous NFT in Bcrp1�/� male micecompared to wild-type male mice, which is likelyattributable to a remarkable contribution of Bcrp1(98%) to the hepatobiliary excretion of NFT.27

However, consistent with the observation of thisstudy, they also showed that the lack of Bcrp1only slightly increased the systemic exposure ofintravenous NFT in female mice.33 This sex-difference in the effect of Bcrp1 on systemicexposure of NFT is likely due to the sex-dependentexpression of Bcrp1 in mouse liver. Bcrp1expression in the liver of female mice has beenshown to be significantly lower than that in malemice, and the role of Bcrp1 in hepatobiliaryexcretion of NFT in wild-type female mice wasnearly completely diminished compared to thatin wild-type male mice.33 Thus, our data areconsistent with those published by Merino et al.33

AUCs of oral NFT in wild-type mice were alsonot altered by pregnancy (Tab. 2). Several factorsmay contribute to this lack of change. First,

09 DOI 10.1002/jps


although we showed that while Bcrp1 expressionin the liver and kidney of pregnant mice wasinduced by pregnancy, its expression in the smallintestine was not significantly altered.12 Thus,the influence of Bcrp1 on the absorption of oralNFT is not expected to be changed by pregnancy.The data from oral NFT study are thus consistentwith the conclusion drawn from intravenous NFTstudy that Bcrp1 is not a major player in thesystemic clearance of NFT in female mice. Second,gastric emptying time and intestinal transit timecan be increased by 30–50% during pregnancy.39

This could actually have occurred in our study,because the time of maximum concentrations ofNFT in pregnant wild-type mice appeared to besomewhat delayed compared to non-pregnantwild-type mice (Fig. 2C). Increase in residenttime in the gastric intestinal tract could increaseabsorption of NFT and compensate for theincrease in elimination due to increased Bcrp1expression in the liver during pregnancy. Anychange in free drug concentration could lead toaltered hepatic first-pass effect. However, thesmall change in plasma protein binding bypregnancy in wild-type mice (Tab. 3) is notexpected to alter systemic exposure of NFT to anoticeable extent, because as stated earlierhepatic first-pass effect does not play a major rolein the elimination of NFT.

Of particular interest of this study is the findingthat there was a drastic decrease (by 70%) insystemic exposure of oral NFT in pregnantBcrp1�/� mice compared with non-pregnantBcrp1�/� mice (Fig. 2D and Tab. 2). Such a drasticchange is unlikely caused by an alteration inplasma protein binding because unbound fractionof NFT in the plasma of Bcrp1�/� mice was notaffected by pregnancy (Tab. 3). Reasons for thisdrastic change in NFT exposure are not known,but might possibly be caused by the followingfactors. Expression and activity of certain effluxtransporters and/or metabolic enzymes in thegastric intestinal tract could be upregulated inpregnant Bcrp1�/� mice to compensate for thefunctional loss of Bcrp1. These efflux transportersand/or enzymes could be further induced bypregnancy, leading to decreased absorption,increased degradation, and/or metabolism ofNFT in pregnant Bcrp1�/� mice after oraladministration. As to why this does not happenin wild-type mice is not clear. Clearly, thereappears to be some interaction between Bcrp1knockout and pregnancy. Further studies areneeded in future work to elucidate the mecha-


nisms underlying this drastic change in thedisposition of oral NFT in pregnant Bcrp1�/�

mice. For example, microarray studies mayillustrate which genes’ expression in maternalorgans (e.g., the small intestine) of the Bcrp1�/�

mice is altered by pregnancy, and what is thefunctional consequence of this change for NFTdisposition which could be accessed by studiessuch as ex vivo small intestinal perfusion tomeasure the absorption of NFT or in situ deter-mination of the basolateral to apical transport ofNFT across the small intestine as well as in vitrometabolism studies of NFT using the mouse smallintestinal homogenates.

In summary, our data show that the systemicexposure and pharmacokinetic characteristics ofNFT in wild-type female mice, after intravenousand oral administration, are not significantlyaffected by pregnancy, even though Bcrp1 expres-sion in the liver and kidney of mice has beenshown to be significantly induced by pregnancy.This suggests that Bcrp1 plays a minor role in thesystemic clearance of NFT in female mice. Themost interesting and important finding of thisstudy is that the systemic exposure of oral NFTwas drastically reduced by pregnancy in Bcrp1�/�

mice, but not in wild-type mice. Such datamay have clinical implications. For example, ifpregnancy does not affect NFT disposition inhumans, dose adjustment may not be necessaryfor the use of NFT in pregnant women. It would bealso important to know if NFT disposition inwomen with low or no BCRP expression causedby genetic variations may be similarly affectedby pregnancy as in Bcrp1�/� mice. However,caution should be taken when extrapolating themouse data to humans, as the pharmacokineticcharacteristics of NFT in humans and mice arequite different.

ACKNOWLEDGMENTS

We gratefully acknowledge financial support fromNIH grant HD044404. Yi Zhang is the recipient ofthe Predoctoral Fellowship from Merck. Lin Zhouis the recipient of the William E. Bradley EndowedFellowship from the School of Pharmacy, Univer-sity of Washington.

REFERENCES

1. Andrade SE, Gurwitz JH, Davis RL, Chan KA,Finkelstein JA, Fortman K, McPhillips H, Raebel


4314 ZHANG ET AL.

MA, Roblin D, Smith DH, Yood MU, Morse AN,Platt R. 2004. Prescription drug use in pregnancy.Am J Obstet Gynecol 191:398–407.

2. Hodge LS, Tracy TS. 2007. Alterations in drugdisposition during pregnancy. Expert Opin DrugMetab Toxicol 3:557–571.

3. Dean M, Stock B, Patterson RJ, Levy G. 1980.Serum protein binding of drugs during and afterpregnancy in humans. Clin Pharmacol Ther 28:253–261.

4. Davison JM, Dunlop W, Ezimokhai M. 1980.Twenty-four hours creatinine clearance duringthe third trimester of normal pregnancy.Br J Obstet Gynaecol 87:106–109.

5. Unadkat JD, Wara DW, Hughes MD, Mathias AA,Holland DT, Paul ME, Connor J, Huang S, NguyenBY, Watts DH, Mofenson LM, Smith E, Deutsch P,Kaiser KA, Tuomala RE. 2007. Pharmacokineticsand safety of indinavir in human immunodeficiencyvirus-infected pregnant women. Antimicrob AgentsChemother 51:783–786.

6. Aldridge A, Bailey J, Neims AH. 1981. The disposi-tion of caffeine during and after pregnancy. SeminPerinatol 5:310–314.

7. Anderson GD. 2005. Pregnancy-induced changes inpharmacokinetics: A mechanistic-based approach.Clin Pharmacokinet 44:989–1008.

8. McGready R, Stepniewska K, Seaton E, Cho T, ChoD, Ginsberg A, Edstein MD, Ashley E, Looareesu-wan S, White NJ, Nosten F. 2003. Pregnancy anduse of oral contraceptives reduces the biotrans-formation of proguanil to cycloguanil. Eur J ClinPharmacol 59:553–557.

9. Acosta EP, Bardeguez A, Zorrilla CD, Van Dyke R,Hughes MD, Huang S, Pompeo L, Stek AM, Pitt J,Watts DH, Smith E, Jimenez E, Mofenson L. 2004.Pharmacokinetics of saquinavir plus low-dose rito-navir in human immunodeficiency virus-infectedpregnant women. Antimicrob Agents Chemother48:430–436.

10. Hebert M, Easterling T, Kirby B, Carr D, BuchananM, Rutherford T, Thummel K, Fishbein D, UnadkatJ. 2008. Effects of pregnancy on CYP3A and P-glycoprotein activities as measured by dispositionof Midazolam and Digoxin: A University ofWashington Specialized Center of Research Study.Clin Pharmacol Ther 84:248–253.

11. Cao J, Stieger B, Meier PJ, Vore M. 2002. Expres-sion of rat hepatic multidrug resistance-associatedproteins and organic anion transporters in preg-nancy. Am J Physiol Gastrointest Liver Physiol283:G757–G766.

12. Wang H, Wu X, Hudkins K, Mikheev A, Zhang H,Gupta A, Unadkat JD, Mao Q. 2006. Expression ofthe breast cancer resistance protein (Bcrp1/Abcg2)in tissues from pregnant mice: Effects of pregnancyand correlations with nuclear receptors. Am J Phy-siol Endocrinol Metab 291:E1295–1304.


13. Cao J, Huang L, Liu Y, Hoffman T, Stieger B, MeierPJ, Vore M. 2001. Differential regulation of hepaticbile salt and organic anion transporters in pregnantand postpartum rats and the role of prolactin.Hepatology 33:140–147.

14. Doyle LA, Yang W, Abruzzo LV, Krogmann T, GaoY, Rishi AK, Ross DD. 1998. A multidrug resistancetransporter from human MCF-7 breast cancer cells.Proc Natl Acad Sci USA 95:15665–15670.

15. Mao Q, Unadkat JD. 2005. Role of the breast cancerresistance protein (ABCG2) in drug transport.AAPS J 7:E118–133.

16. Krishnamurthy P, Schuetz JD. 2006. Role ofABCG2/BCRP in biology and medicine. Annu RevPharmacol Toxicol 46:381–410.

17. Maliepaard M, Scheffer GL, Faneyte IF, van Gas-telen MA, Pijnenborg AC, Schinkel AH, van DeVijver MJ, Scheper RJ, Schellens JH. 2001. Sub-cellular localization and distribution of the breastcancer resistance protein transporter in normalhuman tissues. Cancer Res 61:3458–3464.

18. Cooray HC, Blackmore CG, Maskell L, BarrandMA. 2002. Localisation of breast cancer resistanceprotein in microvessel endothelium of human brain.Neuroreport 13:2059–2063.

19. Mao Q. 2008. BCRP/ABCG2 in the placenta:Expression, function and regulation. Pharm Res25:1244–1255.

20. Zhang Y, Wang H, Unadkat JD, Mao Q. 2007.Breast cancer resistance protein 1 limits fetal dis-tribution of nitrofurantoin in the pregnant mouse.Drug Metab Dispos 35:2154–2158.

21. Kruijtzer CM, Beijnen JH, Rosing H, ten BokkelHuinink WW, Schot M, Jewell RC, Paul EM, Schel-lens JH. 2002. Increased oral bioavailability oftopotecan in combination with the breast cancerresistance protein and P-glycoprotein inhibitorGF120918. J Clin Oncol 20:2943–2950.

22. Yamasaki Y, Ieiri I, Kusuhara H, Sasaki T, KimuraM, Tabuchi H, Ando Y, Irie S, Ware J, Nakai Y,Higuchi S, Sugiyama Y. 2008. Pharmacogeneticcharacterization of Sulfasalazine disposition basedon NAT2 and ABCG2 (BCRP) gene polymorphismsin humans. Clin Pharmacol Ther 84:95–103.

23. Zhang W, Yu BN, He YJ, Fan L, Li Q, Liu ZQ, WangA, Liu YL, Tan ZR, Fen J, Huang YF, Zhou HH.2006. Role of BCRP 421C>A polymorphism onrosuvastatin pharmacokinetics in healthy Chinesemales. Clin Chim Acta 373:99–103.

24. Cisternino S, Mercier C, Bourasset F, Roux F,Scherrmann JM. 2004. Expression, up-regulation,and transport activity of the multidrug-resistanceprotein Abcg2 at the mouse blood–brain barrier.Cancer Res 64:3296–3301.

25. Zhou L, Naraharisetti SB, Wang H, Unadkat JD,Hebert MF, Mao Q. 2008. The Breast cancer resis-tance protein (Bcrp1/Abcg2) limits fetal distribu-tion of Glyburide in the pregnant mouse—An

09 DOI 10.1002/jps


Obstetric-Fetal Pharmacology Research Unit Net-work and University of Washington SpecializedCenter Of Research Study. Mol Pharmacol 73:949–959.

26. Le J, Briggs GG, McKeown A, Bustillo G. 2004.Urinary tract infections during pregnancy. AnnPharmacother 38:1692–1701.

27. Merino G, Jonker JW, Wagenaar E, van Herwaar-den AE, Schinkel AH. 2005. The breast cancerresistance protein (BCRP/ABCG2) affects pharma-cokinetics, hepatobiliary excretion, and milk secre-tion of the antibiotic nitrofurantoin. Mol Pharmacol67:1758–1764.

28. Wang X, Morris ME. 2007. Effects of the flavonoidchrysin on nitrofurantoin pharmacokinetics in rats:Potential involvement of ABCG2. Drug Metab Dis-pos 35:268–274.

29. Mager H, Goller G. 1998. Resampling methods insparse sampling situations in preclinical pharma-cokinetic studies. J Pharm Sci 87:372–378.

30. Brodfuehrer J, Valeriote F, Chan K, Heilbrun L,Corbett T. 1990. Flavone acetic acid and plasmaprotein binding. Cancer Chemother Pharmacol 27:27–32.

31. Hoener B, Patterson SE. 1981. Nitrofurantoin dis-position. Clin Pharmacol Ther 29:808–816.

32. Watari N, Funaki T, Aizawa K, Kaneniwa N. 1983.Nonlinear assessment of nitrofurantoin bioavail-ability in rabbits. J Pharmacokinet Biopharm 11:529–545.


33. Merino G, van Herwaarden AE, Wagenaar E,Jonker JW, Schinkel AH. 2005. Sex-dependentexpression and activity of the ATP-binding cassettetransporter breast cancer resistance protein(BCRP/ABCG2) in liver. Mol Pharmacol 67:1765–1771.

34. Jonen HG. 1980. Reductive and oxidativemetabolism of nitrofurantoin in rat liver. NaunynSchmiedebergs Arch Pharmacol 315:167–175.

35. Conklin JD. 1978. The pharmacokinetics of nitro-furantoin and its related bioavailability. AntibiotChemother 25:233–252.

36. Jonker JW, Buitelaar M, Wagenaar E, VanDer Valk MA, Scheffer GL, Scheper RJ, PloschT, Kuipers F, Elferink RP, Rosing H, BeijnenJH, Schinkel AH. 2002. The breast cancerresistance protein protects against a majorchlorophyll-derived dietary phototoxin and proto-porphyria. Proc Natl Acad Sci USA 99:15649–15654.

37. Conklin JD. 1972. Biopharmaceutics of nitrofuran-toin. Pharmacology 8:178–181.

38. Buzard JA, Conklin JD, O’Keefe E, Paul MF. 1961.Studies on the absorption, distribution, and elim-ination of nitrofurantoin in the rat. J PharmacolExp Ther 131:38–43.

39. Loebstein R, Lalkin A, Koren G. 1997. Pharmaco-kinetic changes during pregnancy and theirclinical relevance. Clin Pharmacokinet 33:328–343.


Documents

Effect of pregnancy on nitrofurantoin disposition in mice