Gli Co Protein A

7/28/2019 Gli Co Protein A

1/40

Role of P-Glycoproteinin PharmacokineticsClinical Implications

Jiunn H. Lin and Masayo Yamazaki

Department of Drug Metabolism, Merck Research Laboratories, West Point, Pennsylvania, USA

Contents

Abstract . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 591. Structure and Mechanism of Drug-Transporting P-Glycoprotein . . . . . . . . . . . . . . . . . . . . 61

1.1 Structure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 611.2 ATP- and Substrate-Binding Sites . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 621.3 Substrate Recognition . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 63

2. Polymorphisms of P-Glycoprotein . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 643. In Vitro/In VivoExtrapolation and Species Differences . . . . . . . . . . . . . . . . . . . . . . . . . 664. Role of P-Glycoprotein in Pharmacokinetics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 68

4.1 Drug Absorption . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 694.1.1 Distribution of Intestinal P-Glycoprotein . . . . . . . . . . . . . . . . . . . . . . . . . . . . 704.1.2 Interindividual Variability of Intestinal P-Glycoprotein . . . . . . . . . . . . . . . . . . . . 704.1.3 Evidence of Intestinal P-Glycoprotein Involvement in Drug Absorption . . . . . . . . . . 714.1.4 Saturable Efflux Transport by Intestinal P-Glycoprotein . . . . . . . . . . . . . . . . . . . 72

4.2 Drug Distribution . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 754.2.1 Blood-Brain Barrier . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 754.2.2 Placenta . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 79

4.3 Drug Metabolism . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 804.4 Drug Excretion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 81

4.4.1 Biliary Excretion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 824.4.2 Renal Excretion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 83

5. P-Glycoprotein-Mediated Drug-Drug Interactions . . . . . . . . . . . . . . . . . . . . . . . . . . . . 845.1 P-Glycoprotein Inhibition Does Not Follow Simple Kinetics . . . . . . . . . . . . . . . . . . . . 84

5.2 Drug Interactions Caused by P-Glycoprotein Inhibition . . . . . . . . . . . . . . . . . . . . . . 865.3 P-Glycoprotein Induction is a Complex Process . . . . . . . . . . . . . . . . . . . . . . . . . . 885.4 Drug Interactions Caused by P-Glycoprotein Induction . . . . . . . . . . . . . . . . . . . . . . 90

6. Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 91

Abstract P-glycoprotein, the most extensively studied ATP-binding cassette (ABC)transporter, functions as a biological barrier by extruding toxins and xenobioticsout of cells.In vitro and in vivo studies have demonstrated that P-glycoproteinplays a significant role in drug absorption and disposition. Because of its localisa-

tion, P-glycoprotein appears to have a greater impact on limiting cellular uptakeof drugs from blood circulation into brain and from intestinal lumen into epithelialcells than on enhancing the excretion of drugs out of hepatocytes and renal tubulesinto the adjacent luminal space. However, the relative contribution of intestinal

DRUG DISPOSITION Clin Pharmacokinet 2003; 42 (1): 59-980312-5963/03/0001-0059/$30.00/0 Adis International Limited. All rights reserved.


2/40

P-glycoprotein to overall drug absorption is unlikely to be quantitatively impor-tant unless a very small oral dose is given, or the dissolution and diffusion ratesof the drug are very slow. This is because P-glycoprotein transport activity be-comes saturated by high concentrations of drug in the intestinal lumen.

Because of its importance in pharmacokinetics, P-glycoprotein transportscreening has been incorporated into the drug discovery process, aided by theavailability of transgenic mdrknockout mice and in vitro cell systems. Whenapplying in vitro and in vivo screening models to study P-glycoprotein function,there are two fundamental questions: (i) can in vitro data be accurately extrapo-lated to the in vivo situation; and (ii) can animal data be directly scaled up tohumans? Current information from our laboratory suggests that in vivo P-glyco-protein activity for a given drug can be extrapolated reasonably well from in vitro

data. On the other hand, there are significant species differences in P-glycoproteintransport activity between humans and animals, and the species differences ap-pear to be substrate-dependent.

Inhibition and induction of P-glycoprotein have been reported as the causesof drug-drug interactions. The potential risk of P-glycoprotein-mediated druginteractions may be greatly underestimated if only plasma concentration is mon-itored. From animal studies, it is clear that P-glycoprotein inhibition always hasa much greater impact on tissue distribution, particularly with regard to the brain,than on plasma concentrations. Therefore, the potential risk of P-glycoprotein-mediated drug interactions should be assessed carefully. Because of overlapping

substrate specificity between cytochrome P450 (CYP) 3A4 and P-glycoprotein,and because of similarities in P-glycoprotein and CYP3A4 inhibitors and induc-ers, many drug interactions involve both P-glycoprotein and CYP3A4. Unlessthe relative contribution of P-glycoprotein and CYP3A4 to drug interactions canbe quantitatively estimated, care should be taken when exploring the underlyingmechanism of such interactions.

P-glycoprotein was first identified by Julianoand Ling as a surface phosphoglycoprotein ex-pressed in drug-resistant Chinese hamster ovary

cells.[1]

This discovery led to the finding that P-glycoprotein is an energy-dependent efflux trans-porter driven by ATP hydrolysis. In humans, twomembers of the P-glycoprotein gene family (MDR1and MDR3) exist, while three members of thisfamily (mdr1a, mdr1b and mdr2) are found inmice.[2,3] The P-glycoprotein encoded by the hu-manMDR1 and mouse mdr1a/1b genes functionsas a drug efflux transporter, whereas humanMDR3P-glycoprotein and mouse mdr2 P-glycoprotein

are believed to be functional in phospholipid trans-port.[4,5] However, the involvement of humanMDR3 P-glycoprotein in drug transport has been

recently reported. An increased directional trans-port of digoxin, paclitaxel and vinblastine acrosspolarised monolayers ofMDR3-transfected cells

has recently been reported by Smith et al.[6]

Theseresults suggest that MDR3 P-glycoprotein is alsoable to transport a range of drugs in addition tophospholipids.

In addition to the expression in tumour cells,human P-glycoprotein is also highly expressed innormal tissues. This transporter is localised on thecanalicular surface of hepatocytes in liver, the api-cal surface of epithelial cells of proximal tubulesin kidneys, columnar epithelial cells of intestine,

epithelial cells of placenta, and the luminal sur-face of capillary endothelial cells in brain in hu-mans.[7,8] The anatomical localisation of P-glyco-

60 Lin & Yamazaki

Adis International Limited. All rights reserved. Clin Pharmacokinet 2003; 42 (1)


3/40


4/40


5/40

reported that a highly conserved Lys residue

within the Walker A motif of histidine permease,an ABC transporter, is directly involved with thebinding of ATP, and a highly conserved Asp resi-due within the Walker B motif serves to bind theMg2+ ion. Mutations in either one of these residuesresult in nonfunctional activity of histidine perme-ase. These results of histidine permease suggestthat the ATP-binding sites may also be restrictedto the Walker A motifs of P-glycoprotein.

It is clear now that ATP binding and subsequenthydrolysis are essential for drug transport.[22]

Studies with photoactive analogues of ATP haveshown that these analogues bind to the ATP-bind-ing domains.[23] Based on the data from vanadatetrapping studies, Senior and Gadsby[24] haveproposed a so-called alternate ATP-binding sitemodel, which explains that although both ATP-binding sites are capable of binding ATP, only onesite participates in the catalysis at a given time, andconformation of this catalytic site precludes theother site from hydrolysing ATP. The stoichiome-try of ATP hydrolysis to drug transport has beenstudied, and the data indicate that, depending onsubstrate, 0.63 molecules of ATP are hydrolysedfor every molecule of drug transported out thecell.[25,26] The reason for the substrate-dependentATP stoichiometry is still unknown.

Unlike the ATP-binding sites that are restrictedto the Walker A motifs of ATP-binding domains,many substrate-binding sites have been identifiedthroughout the transmembrane domains (TM) ofP-glycoprotein. Two substrate-binding sites werefound in TM6 and TM12 by using photoaffinityprobes.[27,28] After complete digestion of the P-glycoprotein with trypsin, two major photo-labelled fragments (5 and 4kD) were mapped byimmunological analysis. These two fragments arelocated within, or immediately next to, the lasttransmembrane domain of each cassette, TM6 andTM12 of P-glycoprotein, respectively. The 5kDfragment includes amino acid residues from 311

456, extending a few residues beyond the WalkerA motif of the first ATP-binding site. In contrast,the 4kD fragment includes residues from 979

1048, but not including the Walker A motif of thesecond ATP-binding site. In addition to the regionsof TM6 and TM12, a region that includes TM7 andTM8 was also reported to be photolabelled specif-ically by an analogue of paclitaxel.[29]

Consistent with the studies of photoaffinityprobes, studies of many mutant P-glycoproteinmolecules suggest that the major drug-bindingsites reside in or near TM6 and TM12.[22] How-ever, mutational studies also suggest that aminoacid substitutions that affect substrate specificityare scattered throughout P-glycoprotein, including

TM1, TM4, TM6, TM10, TM11 and TM12.[30] Re-cent studies by Taguchi et al.[31,32] suggested thatthree amino acids (His61, Gly64 and Leu65) inTM1 are involved in the formation of a bindingpocket that plays a key role in determining the suit-able substrate sizes for P-glycoprotein. For exam-ple, substitution of His61 by an amino acid with ashort side-chain increased resistance to vinblastine(large molecular size; molecular weight 811),whereas substitution of an amino acid with a long

side chain increased resistance to colchicine (smallmolecular size; molecular weight 399).[31,32]

In addition to the transmembrane domains, mu-tational analyses have suggested that the intracel-lular linker loops of P-glycoprotein are also impor-tant for substrate recognition.[33,34] The systematicmutagenesis of 20 Gly residues in the cytoplasmicloops revealed that Gly141 and Gly187 betweenTM2 and TM3, Gly288 between TM4 and TM5,and Gly812 and Gly830 between TM8 and TM9

are important in determining substrate specific-ity.[33]

In summary, these data suggest that drug-bind-ing sites are scattered throughout the P-glycopro-tein molecule, including the transmembranedomains, intracellular loops and even the ATP-binding domains.

1.3 Substrate Recognition

One of the most intriguing aspects of P-glyco-protein is that a single integral membrane proteincan recognise and transport so many drugs with awide array of chemical structures, ranging from a

P-Glycoprotein 63



6/40

molecular weight of 250 (cimetidine) to 1202

(cyclosporin).

[35,36]

Although most of the drugstransported by P-glycoprotein are basic or un-charged, there are many exceptions. The only com-mon feature is that most of the P-glycoprotein sub-strates are hydrophobic in nature, suggesting thatpartitioning of the lipid membrane of cells is thefirst step for the interaction of a substrate with theactive sites of P-glycoprotein.

Various efforts have been made to establish thestructure-activity relationship for P-glycoproteinsubstrates. Originally, it was believed that a basicnitrogen atom was a prerequisite for the interactionof substrate and P-glycoprotein. For example,studies of colchicine and its analogues suggestedthat the nitrogen atom of the acetamido group atthe 7 position was essential for P-glycoprotein rec-ognition.[37] However, compounds lacking a nitro-gen atom, such as cortisol, aldosterone and dexa-methasone, are recognised to be good substratesfor P-glycoprotein.[38] Studies by Ecker et al.[39,40]

have shown that both the lipophilicity and numberof hydrogen bonds of compounds are probably themost important parameters in determining the af-finity of compounds to P-glycoprotein. The higherthe lipophilicity or the larger the number of hydro-gen bonds, the better the substrates are for the P-glycoprotein transporter.

Similarly, Seelig and Landwojtowicz[41] havealso suggested that both lipophilicity and numberof hydrogen bonds are important determinants forsubstrates and P-glycoprotein interaction. Theyconcluded that partitioning of the lipid membraneis the rate-limiting step for the interaction of a sub-strate with P-glycoprotein. Furthermore, they sug-gested that dissociation rate of the P-glycoprotein-substrate complex is controlled by the number ofhydrogen bonds. Based on structural analysis of100 P-glycoprotein substrates, Seelig[42] furtherproposed that in addition to lipophilicity and num-ber of hydrogen bonds, some essential structuralelements of substrates are required for an interac-

tion with P-glycoprotein. The recognition ele-ments of substrates are formed by two or three elec-tron donor (hydrogenbonding acceptor) groups

with a fixed spatial separation: 2.5 0.3 or 4.6 0.6. Additionally, the surface area and amphiphi-lic characteristic of the substrate also appear toplay a significant role in determining its P-glyco-protein activity.[43]

Although a wealth of information on the rela-tionship between physicochemical properties ofsubstrates and P-glycoprotein activity has beengenerated in recent years, a clear structure-activityrelationship for predicting P-glycoprotein sub-strates still cannot be established.[44] The lack ofclear structure-activity relationship for substrate

recognition is attributed mainly to the structuralcomplexity of P-glycoprotein.

2. Polymorphisms of P-Glycoprotein

Humans are not necessarily created equal interms of biological make-up. Because of evolu-tionary and environmental factors, there is a re-markable degree of genetic variability built into thepopulation. Like many cytochrome P450 (CYP)

isoenzymes,[45] genetic polymorphisms of P-gly-coprotein in animals and humans have been re-ported. Thus, the genetic polymorphisms of P-gly-coprotein may also represent a major source ofindividual variability in the potential toxicity andpharmacokinetics of drugs.

The genetic polymorphism of P-glycoproteinwas first reported in CF-1 mice by Lankas andUmbenhauer at Merck in 1997.[46,47] A subpopula-tion of CF-1 mice, approximately 25%, was very

sensitive to neurotoxicity following exposure toavermectin, an antiparasitic agent. The 50% lethaldose (LD50) values were 0.3 and 120 mg/kg for thesensitive and insensitive groups, respectively. Sub-sequently, it is now known that this avermectin-induced neurotoxicity is the result of a deficiencyin mdr1a P-glycoprotein that normally contributesto a functional blood-brain barrier (BBB). In nor-mal (wild-type) CF-1 mice the abundant P-glyco-protein in the BBB pumps avermectin efficientlyout of the brain, but in mdr1a-deficient mice, thisprotective function is absent, resulting in a morethan 80-fold higher accumulation of avermectin inthe brain. The P-glycoprotein-deficient CF-1 mice

64 Lin & Yamazaki



7/40

are also at higher risk of birth defects caused by

avermectin.

[48]

When female CF-1 mice weretreated with avermectin during pregnancy, fetusesdeficient in P-glycoprotein (/) were 100% sus-ceptible to cleft palate, while their heterozygotelitters (+/-) were less sensitive. The homozygousfetuses (+/+) with abundant P-glycoprotein weretotally insensitive at the dose tested. The birth de-fect is attributed to fetal exposure to avermectin.The above cases of neurotoxicity and teratogenesisdemonstrate the importance of P-glycoprotein inprotecting the brain and fetus against toxic xeno-biotics.

The molecular basis of mdr1a deficiency inCF-1 mice was further studied at the RNA and DNAlevel, using reverse transcription-polymerase chainreaction (RT-PCR) and long PCR with oligonucle-otides specific for mdr1a. Sequencing of the intronbetween exon 22 and 23 in P-glycoprotein-defi-cient CF-1 mice revealed an insertion of approxi-mately 8.35kb of DNA at the exon 23 intron-exonjunction. This insertion results in the aberrantsplicing of the mRNA and loss of exon 23 duringRNA processing.[49] A subpopulation of collie dogsis also known to be very sensitive to avermectinand it has been speculated that the avermectin-induced neurotoxicity in the dogs is due to P-gly-coprotein genetic polymorphisms.[50,51] Recently,RT-PCR studies revealed a deletion mutation ofthe MDR1 gene in avermectin-sensitive colliedogs.[52] The 4-bp deletion results in a frame shift,generating several stop codons that prematurelyterminate P-glycoprotein synthesis.

Genetic polymorphisms of human P-glycopro-tein were first reported from in vitro studies withcancer cells.[53,54] However, the kinetic impact ofpolymorphism on P-glycoprotein function in vivoremained unclear until the recent report by Hoff-meyer et al.,[55] who identified a single nucleotidepolymorphism (SNP) in exon 26 (C3435T) ofMDR1. There was a significant correlation of theSNP and the functional activity of P-glycoprotein.

The homozygous T-allele (mutant) is associatedwith more than 2-fold lower intestinal P-glycopro-tein expression levels compared with homozygous

C-allele (wild type). Individuals carrying homo-

zygous T-allele showed a lower duodenal P-glyco-protein level and consequently higher peak plasmaconcentrations (Cmax) of digoxin, a substrate of P-glycoprotein, probably through an increase in dig-oxin absorption as a result of decreased intestinalP-glycoprotein. This was the first example that in-dicated that P-glycoprotein polymorphism can di-rectly affect drug absorption in humans. However,a recent study by Sakaeda et al.[56] suggests thatthere are no statistically significant differences inthe C

maxand 24-hour area under the curve (AUC

24)

of digoxin between individuals carrying wild-typeC-allele or homozygous mutant T-allele. Simi-larly, conflicting results of the effect of C3534Tpolymorphism on the absorption of fexofenadinehave also been reported.Kim et al.[57] showed thatindividuals harbouring homozygous T-allele mu-tation tended to have lower plasma AUC24 of fexo-fenadine. In contrast, Drescher et al.[58] claimedthat there were no significant differences betweenT/T and C/T genotypes. Additionally, kinetic stud-ies in healthy subjects and renal transplant patientssuggested that C3534T polymorphism had littleeffect on the absorption of cyclosporin.[59,60]

Therefore, the impact of C3534T polymorphismon drug absorption is still not clear.

A strong association between C3435T alleleand G2677A/G2677T alleles was observed whenMDR1 polymorphisms were investigated in 100placentas from Japanese women.[61] Of 65 sampleswith a C3435T allele, 61 (93.8%) also had a mutantG2677T/G2677A allele. Interestingly, there ap-peared to be a correlation between the level of P-glycoprotein expression and G2677T/G2677A inexon 21, and between the P-glycoprotein level andC3435T in exon 26. The placental P-glycoproteinexpression levels for wild-type, heterozygotes andhomozygotes of G2677A/G2677T mutant allelewere 2.44, 1.97, and 1.45 (arbitrary units), respec-tively, while the corresponding mean P-glycopro-tein expression levels for the C/C, C/T and T/T

genotypes at position 3435 were 2.11, 1.84, and1.51 (arbitrary units). The frequency in exon 21occurred in 58% of the sample as heterozygosity

P-Glycoprotein 65



8/40

and 28% as homozygosity for the mutant allele,

while the frequency in exon 26 occurred in 46% ofthe sample as heterozygosity and 19% as homo-zygosity for the mutant allele. Given the highfrequency for the mutant allele G2677A/G2677Tin this study from 100 placentas, it is quite puzzlingwhy no G2677A/G2677T mutant allele was ob-served from a sample population of 188 in theaforementioned study by Hoffmeyer et al.,[55] eventhough a total of 15 different SNPs, includingC3435T, were identified in their study. Recently,Hoffmeyers group detected both C3435T andG2677A/G2677T polymorphisms with high fre-quencies in a relatively larger sample populationof 461.[62]

Similar to the ethnic variation in the polymor-phism of CYP,[63] interethnic differences inMDR1polymorphisms are observed. Using a polymerasechain reaction-restriction fragment length poly-morphism assay, 1280 subjects from 10 differentethnic groups were evaluated for the C3435T poly-morphism in exon 26.[64] Marked differences ingenotype and allele frequency were observed be-tween the African and the Caucasian/Asian popu-lations. The Ghanaian, Kenyan, African-Americanand Sudanese populations have frequencies of 83,83, 84 and 73%, respectively, for the C/C (wildtype) allele. In contrast, the British Caucasian, Por-tuguese, Southwest Asian, Chinese, Filipino andSaudi populations have lower frequencies of theC/C allele compared with the African groups, rang-ing from 3455%. Although the level of P-glyco-protein expression was not determined in thisstudy, these results suggest that P-glycoproteinexpression in African populations may be higherthan that in the Caucasian/Asian populations.

It is interesting to note that ivermectin, a potentanthelmintic agent used for the prevention andtreatment of river blindness (onchocerciasis) inAfrica, is a very safe drug, even though this drugcauses neurotoxicity in animals with low P-glyco-protein expression. To date, of more than 20 mil-

lion patients in Africa who had been treated withivermectin, not a single case of neurotoxicity hasbeen reported. The lack of neurotoxicity might be

attributed to the high P-glycoprotein expression inthe African population. On the other hand, the highexpression of P-glycoprotein might contribute tothe high incidence of drug resistance to cancertreatment in individuals of African origin.[65]

In summary, theMDR1 gene is highly polymor-phic. To date, at least 16 SNPs have been identifiedin theMDR1 gene, and it is anticipated that moreSNPs will be found in the future (table I). Althoughsome variants lead to amino acid changes, most ofthe detected polymorphisms are intronic or silent.So far, only three SNPs (T129C in exon 1b,

G2677A/G2677T in exon 21 and C3435T in exon26) have been demonstrated to be associated withvariation in P-glycoprotein expression. Undoubt-edly, variation in P-glycoprotein expression result-ing fromMDR1 polymorphism is one of the majorsources contributing to interindividual variabilityin drug absorption and disposition.

3. In Vitro/In VivoExtrapolationand Species Differences

As will be discussed in section 4, P-glycopro-tein plays an important role in absorption, distribu-tion, metabolism and excretion of many drugs. Be-cause of the importance of P-glycoprotein inpharmacokinetics, many pharmaceutical compa-nies have begun to incorporate P-glycoproteindrug transport screening into the drug discoveryprocess. The availability of transgenic mdrknock-out mice and in vitro cell systems has paved the

way for studies of the role of P-glycoprotein indrug absorption and disposition. When applying invitro and in vivo screening models to study P-gly-coprotein function, there are two fundamentalquestions that industrial drug metabolism scien-tists must confront daily: (i) can in vitro data beaccurately extrapolated to the in vivo situation, and(ii) can animal data be directly scaled to humans?

To test whether in vitro P-glycoprotein activityof drugs can be extrapolated to the in vivo situation,we have conducted a study to measure in vitro andin vivo P-glycoprotein activity of ten model com-pounds.[67] The in vitro P-glycoprotein activity ofthese compounds was determined using mdr1a-

66 Lin & Yamazaki



9/40

transfected LLC-PK1 cells, and expressed as theratio of basolateral-to-apical transport to apical-to-basolateral transport (B-to-A/A-to-B). On theother hand, the in vivo P-glycoprotein activity was

determined using CF-1 mdr1a (+/+) and mdr1a(/) mice. Following intravenous administration,the drug concentration in brain and plasma wasmeasured every 15 minutes up to 60 minutes. Theratio of brain AUC in mdr1a (/) mice to that inmdr1a (+/+) mice was used as an index ofin vivoP-glycoprotein activity. There was a strong posi-tive correlation (r2 = 0.93, p < 0.001) when the invitro B-to-A/A-to-B ratio was plotted against thein vivo brain AUC ratio for these ten com-

pounds.[67] A strong correlation was also observedwhen the brain concentration was normalised byplasma concentration. These results suggest that in

vivo P-glycoprotein activity of a given drug can bereasonably well extrapolated from in vitro data. Inour laboratory, the in vitro and in vivo correlationwas further evaluated with an additional 20 com-

pounds, and a strong correlation between in vitroB-to-A/A-to-B ratio and the in vivo brain AUC ra-tio was observed again (unpublished data).

The most obvious species differences in thedrug-transporting P-glycoprotein between miceand humans is that there are two members of drug-transporting P-glycoprotein (mdr1a and mdr1b)for mice, and just one in humans (MDR1).[2] Al-though the kinetics and substrate specificities aregenerally similar between mouse mdr1a and mdr1b

P-glycoprotein,[68-70] species differences in func-tional activity between human MDR1 P-glycopro-tein and these two mouse P-glycoproteins have

Table I. Single nucleotide polymorphisms (SNPs) in the MDR1 gene

SNPa Mutation Mutant allele frequency (%)

Hoffmeyer et al.[55] Ito et al.[66] Cascorbi et al.[62] Tanabe et al.[61] Mickley et al.[54]

5-flanking/-41 A/G 7.3 9.4

1a/145 C/G 1.0 1.0

1b/-129 T/C 5.9 8.3

2/1 G/A 5.6 9.0

2/61 A/G 9.3 11.2

5/25 G/T 16.5

5/35 G/C 0.6

5/307 T/C 0.6 0

6/+139 C/T 40.6 37.2

6/+145 C/T 1.2

11/1199 G/A 6.5 5.512/1236 C/T 37.8 38.5 (T>C) 41.0 35.4 (T>C)

12/+44 C/T 5.9 4.9

17/76 T/A 45.3 46.2

17/+137 A/G 0.6

21/2677 G 56.5 36.5 56.4

21/2677 G/T 0 41.6 41.7 43.6

21/2677 G/A 0 1.9 21.8 0

24/2956 A/G 0

24/2995 G/A 6.7

26/3220 A/C 0.2

26/3396 C/T 0.3

26/3435 C/T 48.1 53.9 49.0

28/4030 G/C 0

28/4036 A/G 25.0

a Exon/position.

P-Glycoprotein 67



10/40

been reported by Tang-Wai et al.[71] Stably trans-

fected cells were developed that expressed similaramounts of P-glycoprotein encoded by mdr1a,mdr1b andMDR1 genes. The three transfected celllines were tested for their cellular resistance (cellsurvival) to dactinomycin, doxorubicin, colchicineand vinblastine. The 50% inhibitory concentra-tions (IC50 values) for all of the drugs were muchlower (3- to 20-fold) in MDR1 transfected cellsthan in mdr1a cells.[71] Likewise, with the excep-tion of dactinomycin, the IC50 values were lowerin humanMDR1 cells than in mdr1b cells. Theseresults suggest that there are species differences infunctional capacity between human MDR1 P-gly-coprotein and murine P-glycoproteins.

In our laboratory, more than 640 compoundshave been evaluated for their P-glycoprotein activ-ity in mouse mdr1a and humanMDR1 transfectedLLC-PK1 cells (L-mdr1a and L-MDR1, respec-tively). As shown in figure 1, there is a poor corre-lation between mdr1a and MDR1 transcellulartransport activity as measured by B-to-A/A-to-Btransport ratio (r2 = 0.44). Approximately 35% ofthe compounds exhibited substantial differences(>3-fold) between mdr1a and MDR1 gene trans-fected cells. The observed species difference in theP-glycoprotein transport was not due to the differ-ences in the level of the P-glycoprotein expression.Western blotting data revealed that the P-glyco-protein level was similar in both mdr1a andMDR1transfected cells.

Species differences in transport activity werealso observed between human P-glycoprotein andP-glycoprotein of other animal species. We haverecently compared the P-glycoprotein activity ofmarker P-glycoprotein substrates using cell linesexpressing human, mouse, rat and canine P-glyco-protein. Again, significant species differences inP-glycoprotein activity were found among theseanimal species (unpublished data).

Although these in vitro studies strongly suggestthe possibility of species differences in P-glyco-

protein activity, there are still no in vivo data tosupport the potential of species differences. As willbe discussed in section 4, the mdr1a/1b knockout

mouse provides a very useful model for the studyof the role of P-glycoprotein in pharmacokineticsof drugs. However, because of the potential speciesdifferences in P-glycoprotein activity, extrapola-tion from knockout mice to humans should be car-ried out with discretion.

4. Role of P-Glycoproteinin Pharmacokinetics

Although the physiological function for P-gly-coprotein is still not fully understood, the role ofthis efflux transporter in pharmacokinetics is be-coming increasingly appreciated. In humans, P-glycoprotein is found on the apical surface ofcolumnar epithelial cells of small and large intes-tines, the biliary canalicular membrane of hepato-cytes, the apical surface of epithelial cells of theproximal tubules of kidney, the apical surface ofepithelial cells of placenta and the apical surfaceof endothelial cells in blood capillaries of thebrain.[7,8] Because of its strategic localisation, theP-glycoprotein transporter functionally can limit

0

20

40

60

80

100

120

B-to-A/A-to-BinL-MDR1(human)

B-to-A/A-to-B in L-mdr1a (mouse)

120100806040200

Fig. 1. Correlation of transcellular transport ratios for 642 struc-turally diverse compounds in monolayers of LLC-PK1 cellstransfected with mouse mdr1a or human MDR1 (L-mdr1a andL-MDR1, respectively). A-to-B = apical to basal; B-to-A = basal

to apical.

68 Lin & Yamazaki



11/40

cellular uptake of drugs from the blood circulation

into the brain and placenta, and from the gastroin-testinal lumen into the enterocyte. On the otherhand, this transporter can also enhance the elimi-nation of drugs from hepatocytes, renal tubulesand intestinal epithelial cells into the adjacent lu-minal space. Therefore, it is very important to dis-tinguish the localisation of P-glycoprotein in cellsin relation to drug movement either uptake ofdrugs into cells or excretion of drugs out of cells.As will be discussed later, there is a tendency forP-glycoprotein to have a greater impact on druguptake than on drug excretion.

Perhaps the most important milestone in P-gly-coprotein research was the development ofmdr-knockout mice. Since mdrknockout mice becameavailable, our understanding of the role of P-gly-coprotein in pharmacokinetics has increased expo-nentially. As discussed above, mice have two typesof drug-transporting P-glycoprotein (mdr1a andmdr1b), which are expressed in a tissue-specificmanner.[72] For example, only mdr1a P-glycopro-tein is expressed in the brain and intestine of mice,while both mdr1a and mdr1b P-glycoprotein areexpressed in the liver and kidney. Interestingly,mdr1a and mdr1b P-glycoprotein together appearto cover the same tissues as the single humanMDR1 P-glycoprotein, suggesting that mdr1a andmdr1b together fulfil the same function as the sin-gle P-glycoprotein in humans. Because of tissue-specific expression, it is expected that genetic dis-ruption of the mdr1a gene would have a greaterimpact on drug uptake into the brain and intestinethan drug excretion from the liver and kidney.

In addition, it should be noted that genetic dis-ruption of one or both of the mdrgenes might affectthe expression and function of other transportersystems or even drug-metabolising enzyme sys-tems in mice. For example, it is known that themdr1b gene is upregulated in mdr1a knockoutmice.[8] Recently, Schuetz et al.[73] have reportedthat both the protein expression and catalytic ac-

tivity of CYP is significantly increased in themdr1a, mdr1b and mdr1a/1b knockout micehoused in The Netherlands. However, for the ge-

netically identical mdr1a and mdr1a/1b knockoutmice housed in the US, there were no significantchanges in their protein expression and catalyticactivity of CYP. Because of the possible existenceof unrecognised factors that are associated with thegenetic disruption, data derived from mdr1a singleknockout mice or from mdr1a/1b double knockoutmice have to be interpreted with caution.

4.1 Drug Absorption

There are many factors that influence the

bioavailability of drugs, which can be broadlycategorised as physicochemical and biologicalfactors.[74,75] The former comprise the intrinsicproperties of the drug, such as pKa, molecular size,lipophilicity and solubility, and the latter includegastric and intestinal transit time, lumen pH, mem-brane permeability, mucosa blood flow rate andfirst-pass metabolism. After oral administration,drug absorption occurs predominantly within thesmall intestine, because of its large surface area

provided by epithelial folding and the villousstructures of epithelial cells. During oral absorp-tion, drugs can be transported by either the trans-cellular or paracellular pathway across the epithe-lial cells, or a combination of both. The relativecontribution of the transcellular pathway to overallabsorption is highly dependent on the lipophilicityof drugs. In an in vitro study with Caco-2 cells, therelative contribution of the transcellular pathwaywas determined to be 25, 45, 85 and 99% for chlo-

rothiazide, furosemide, cimetidine and proprano-lol, respectively. The values correlated fairly wellwith the lipophilicity of the drugs, the logP valuesof which were 0.2, 0.08, 0.4 and 3.6, respec-tively.[76] Most orally administered drugs enter thesystemic circulation by passive transcellular dif-fusion, because of their lipophilicity. Therefore,the intestinal absorption of a drug is often pre-dicted on the basis of its lipophilicity.

As noted earlier, the most common physico-chemical property for P-glycoprotein substratesidentified so far is that they are mostly lipophilic,which implies that lipophilic drugs are likely to beP-glycoprotein substrates. Therefore, absorption

P-Glycoprotein 69



12/40

of drugs is further complicated by the existence ofP-glycoprotein efflux transporter, which is highlyexpressed on the apical surface of epithelial cells.In the intestinal lumen, drugs that are P-glycopro-tein substrates will be absorbed and cross the epi-thelial cell membrane by simple diffusion. Onceinside the cells, a fraction of the drug moleculescontinues to diffuse along the concentration gradi-ent into capillary blood. However, a portion ofdrug molecules will be removed by the efflux P-glycoprotein transporter out of cells back into thelumen and another part of the drug molecules is

subject to intestinal metabolism. Consequently, thenet amount of drug absorbed into the mesentericblood circulation is the difference between theamount absorbed by the influx process and thesummation of the amount extruded by efflux trans-port together with the amount metabolised by en-zymes.

4.1.1 Distribution of Intestinal P-Glycoprotein

The distribution of intestinal P-glycoprotein is

not uniform among cells along the epithelial villi.Immunohistological studies with human jejunumand colon using MRK16 antibody revealed thathigh levels of P-glycoprotein were only observedin the apical surface of columnar epithelial cells,but not in crypt cells.[7] Unlike hepatocytes, whichregenerate only when untimely death occurs, intes-tinal epithelial cells have a programmed, limitedlife span. The villous epithelial cells are matureand nondividing, whereas the crypt cells continue

to mature as they ascend toward the villus and areextruded at its tip. The time required for migrationfrom the crypt base to the villous tip has been esti-mated to be 2 6 days.[77] Whether the programmedlife span and rapid migration will have impact onthe regulation of intestinal P-glycoprotein expres-sion is an open question, and requires further in-vestigation.

The distribution of P-glycoprotein is also notuniform along the length of intestine. Fojo et al.[78]

have measured the content ofMDR1 mRNA ex-pression over the total length of human gastroin-testinal tract. The levels of mRNA appear to in-crease progressively from the stomach to the colon

with a low level in the stomach (5 arbitrary units),

an intermediate level in the jejunum (20 arbitraryunits) and a high level in the colon (30 arbitraryunits). The uneven distribution of intestinal P-gly-coprotein is expected to have a significant impacton the absorption of P-glycoprotein substrates. Theinfluence of uneven distribution of P-glycoproteinin intestine was demonstrated in a clinical studywith cyclosporin.[79] Cyclosporin was given to tenhealthy volunteers at different parts of the gastro-intestinal tract (stomach, jejunum and colon). Theoral AUC of cyclosporin was in the rank orderstomach > jejunum > colon. There was a negativecorrelation betweenMDR1 mRNA expression andoral AUC of cyclosporin. Uneven distribution ofP-glycoprotein has also been observed in rats. Us-ing the rat intestinal loop technique, vinblastine,a well-known P-glycoprotein substrate, was ab-sorbed fairly well from ileal loops, ranging from3060% of the dose in 30 minutes, whereas absorp-tion of vinblastine from the jejunal loop was almostnegligible, suggesting a high level of P-glycopro-tein expression in rat jejunum.[80]

4.1.2 Interindividual Variability

of Intestinal P-Glycoprotein

Although interindividual variability in drug-metabolising enzymes is well documented,[45] onlya few papers deal with the issue of interindividualvariability of P-glycoprotein. In a clinical study of25 kidney transplant recipients, expression level ofintestinal P-glycoprotein was measured using im-munoblotting.[81] Biopsy specimens were obtainedfrom the second portion of the duodenum of eachpatient. Because P-glycoprotein is expressed ex-clusively in mature epithelial cells in the villous tipof intestinal mucosa, differences in the number oftotal mature cells in individual biopsies might con-tribute to the variability. To correct this practicalproblem, the investigators used villin, a constitu-tively expressed protein in mature epithelial cells,as an internal standard. The results from this studyindicate that there is a significant interindividualvariability in the intestinal P-glycoprotein expres-sion. More than 8-fold differences in the P-glyco-protein expression were observed in a small popu-

70 Lin & Yamazaki



13/40

lation of 25 patients; the intestinal P-glycoproteinlevel of duodenal biopsies ranged from 31263(arbitrary units).

As discussed earlier, theMDR1 gene is highlypolymorphic. At least 16 SNPs have been identi-fied (table I). Although complete P-glycoproteindeficiency has not been reported for these poly-morphisms, the SNP at 3435 in exon 26 does influ-ence the expression level of intestinal P-glycopro-tein. The mean values of intestinal P-glycoproteinexpression for the C/C homozygotes (wild type, n= 6), C/T heterozygotes (n = 10), and T/T homo-

zygotes (n = 5) at position 3435 were 1275, 956and 627 (arbitrary units), respectively.[55]

Interestingly, intraindividual variability in in-testinal P-glycoprotein expression has also beenreported. A profound intraindividual variability inintestinal P-glycoprotein expression (mRNA) wasobserved in a young patient during tacrolimus ther-apy after small bowel transplantation.[82] Both themRNA expression and plasma concentration oftacrolimus were measured periodically during the

immunosuppressant therapy. In a period of 120days, there was a 4-fold variation inMDR1 mRNAexpression level and a 2-fold variation in troughplasma concentration of tacrolimus. The variationin P-glycoprotein expression inversely relatedfairly well to the variation of tacrolimus concen-trations after oral administration.

Collectively, these results suggest that interin-dividual and intraindividual variability in intesti-nal P-glycoprotein expression may contribute to

variability of oral absorption of drugs that are P-glycoprotein substrates.

4.1.3 Evidence of Intestinal P-Glycoprotein

Involvement in Drug Absorption

Evidence of the involvement of intestinal P-glycoprotein in drug absorption was first demon-strated in vitro with Caco-2 cells in which P-gly-coprotein was highly expressed. Using Caco-2cells, the B-to-A transport of vinblastine anddocetaxel was 10- and 20-fold, respectively,greater than the A-to-B transport, and the A-to-Btransport was enhanced significantly in the pres-ence of verapamil and MRK16 by blocking the

P-glycoprotein efflux function.[83,84] Caco-2 cells

have also been used to study the intestinal transportof cyclosporin.[85] Similarly, the B-to-A transportof cyclosporin was much greater than the A-to-Btransport. The A-to-B cyclosporin transport in-creased and the B-to-A transport decreased aftertreatment with the P-glycoprotein inhibitors pro-gesterone and chlorpromazine. Furthermore, stud-ies with Caco-2 cells revealed that active B-to-Atransport of peptides was inhibited by verapamil,suggesting the involvement of P-glycoprotein inthe absorption of peptides.[86] These results fromin vitro studies clearly suggest that P-glycoproteinplays a significant role in drug absorption by lim-iting drug transport from intestinal lumen.

Direct evidence for the role of intestinal P-gly-coprotein in drug absorption was derived from invivo studies with mdr1a (/) knockout mice. Theoral absorption of paclitaxel was studied in mdr1a(/) and mdr1a(+/+) mice.[87] The plasma AUCof paclitaxel was 2- and 6-fold, respectively,higher in mdr1a (/) mice than mdr1a (+/+) miceafter intravenous and oral drug administration. Theincreased AUC of paclitaxel after intravenous ad-ministration in mdr1a (/) mice reflected a de-crease in elimination clearance, whereas the higherAUC after oral administration in mdr1a (/) miceresulted from a combination of a decrease in theelimination clearance and an increase in the extentof drug absorption from intestinal lumen. Based onthe AUC values after intravenous and oral admin-istration, the bioavailability of paclitaxel was cal-culated to be 11 and 35% for mdr1a (+/+) andmdr1a (/) mice, respectively. From this study, itis clear that intestinal P-glycoprotein does limitdrug absorption by extruding drugs from epithelialcells back into the intestinal lumen. Further evi-dence for the involvement of intestinal P-glyco-protein in drug absorption in humans is providedby the clinical study by Hoffmeyer et al.,[55] whoshowed a negative correlation between duodenalP-glycoprotein expression and plasma level of dig-

oxin.The efflux function of intestinal P-glycoprotein

is further supported by the observations that a sig-

P-Glycoprotein 71



14/40

nificant amount of paclitaxel was excreted directly

from blood circulation into intestinal lumen afterintravenous administration. The effect of P-glyco-protein on intestinal excretion can be experimen-tally determined in mice after interruption of bileflow by gallbladder cannulation. A fraction of 11%of the dose was excreted into the intestinal lumenwithin 90 minutes after intravenous administrationof paclitaxel to mdr1a (+/+) mice with a cannulatedgallbladder, but only 2.5% of the dose in the intes-tinal lumen ofmdr1a (/) mice.[87] Similarly, P-glycoprotein-mediated intestinal excretion in micewas also reported for digoxin. Approximately 16and 2% of the digoxin dose, respectively, was ex-creted into the intestinal lumen ofmdr1a (+/+) andmdr1a (/) mice with a cannulated gallbladderwithin 90 min after intravenous administration.[88]

Although the phenomenon of intestinal excretionof drugs has been known for more than two de-cades,[89] involvement of P-glycoprotein in intesti-nal excretion has only been recognised recently.From a mechanistic point of view, intestinal excre-tion should also be considered as an additionalpathway for the elimination clearance of P-glyco-protein substrates.

Unlike the direct evidence obtained from mdr1aknockout mice, the involvement of MDR1 P-gly-coprotein in drug absorption is difficult to provedirectly in humans. Thus, the role of intestinal P-glycoprotein in drug absorption in humans is oftenderived indirectly from inhibition studies. In a clini-cal study, five patients received a safe oral dose ofpaclitaxel 60 mg/m2, and nine other patients re-ceived the same oral dose of paclitaxel combinedwith one single oral dose of cyclosporin 15mg/kg.[90] The oral bioavailability of paclitaxelwhen given without cyclosporin was less than5%, and increased to 50% when cyclosporin wascoadministered. Similar results were found in an-other clinical study with docetaxel.[91] The bio-availability of docetaxel in cancer patients in-creased from 8% without cyclosporin to 88% in

combination with cyclosporin. Because cyclo-sporin is known to be a potent P-glycoprotein in-hibitor, these results suggest the involvement of

P-glycoprotein as an intestinal barrier in limiting

the absorption of paclitaxel and docetaxel.Since cyclosporin is also an inhibitor of CYP-3A4 and other CYP enzymes, it is likely that theobserved increase in oral bioavailability of thesedrugs could also be partly attributed to reduced me-tabolism by the inhibition of CYP enzymes bycyclosporin. Moreover, cyclosporin is a potent in-hibitor of other transporters, such as canalicularbile salt transporter and canalicular multispecificorganic anion transporter (cMOAT). The inhibi-

tory Ki values of cyclosporin for taurocholate (bilesalt transporter), leukotriene C4 (cMOAT) anddaunorubicin (P-glycoprotein) were 0.2, 3.4 and1.5 mol/L, respectively.[92] Clearly, the increasedbioavailability of paclitaxel and docetaxel causedby cyclosporin cannot be explained by P-glyco-protein inhibition alone. Therefore, the inhibitionstudy can only provide a qualitative assessmentas to whether P-glycoprotein is involved in drugabsorption. Due to the lack of a specific P-glyco-

protein inhibitor, it is difficult to estimate the quan-titative contribution of P-glycoprotein to drug ab-sorption by the inhibition approach.

Another way by which evidence can be shownfor intestinal P-glycoprotein involvement in drugabsorption is to establish the correlation betweenthe absorption profile (oral AUC) of drugs andthe expression of intestinal P-glycoprotein (ormRNA). In a clinical study, cyclosporin was givenby gavage to ten male volunteers at different parts

of the gastrointestinal tract (stomach, jejunum/-ileum and colon) in a crossover manner, with awashout period between the administrations of atleast 7 days.[79] There was a strong negative corre-lation between the plasma AUC of cyclosporin af-ter administration at different locations of the gas-trointestinal tract and the local mRNA expressionof intestinal P-glycoprotein. Similarly, based onthe observation that a highly significant correlationexists between enterocyte P-glycoprotein content

and cyclosporin absorption kinetics, Lown et al.[81]concluded that intestinal P-glycoprotein plays asignificant role in the absorption of cyclosporin.

72 Lin & Yamazaki



15/40

With several lines of supportive evidence, theconclusion from these two correlation studies thatP-glycoprotein transport is involved in the absorp-tion of cyclosporin appears valid and appropriate.However, it should be noted that a good correlationitself does not prove a causal relationship. There-fore, it is highly desirable to have other supportivein vitro and/or animal data when applying the cor-relation approach.

4.1.4 Saturable Efflux Transport

by Intestinal P-Glycoprotein

Like that of drug-metabolising enzymes, thefunctional activity of P-glycoprotein is saturable.In a recent kinetic study involving Caco-2 cells,[93]

the efflux of vinblastine and digoxin by P-glyco-protein have been demonstrated to be saturable;the Michaelis-Menten constants (Km values) forvinblastine and digoxin were 26 and 58 mol/L,respectively. A similar Km value for vinblastine(18 mol/L) was reported by other investigatorsusing Caco-2 cells.[83] By using the Ussing cham-

ber technique, the transport of digoxin was studiedin human and rat intestinal tissues. Saturable trans-port for digoxin was also observed in both humanand rat intestinal tissues. The Km values for dig-oxin were 81, 74, 51 and 59 mol/L, respectively,for rat jejunum, rat ileum, rat colon and humancolon.[86] Interestingly, the Km values for digoxinderived from human colon tissue and Caco-2 cellsare almost identical (59 vs 58 mol/L).[93] UsingCaco-2 cells, transport of cyclosporin was shown

to be saturable with a Km of 3.8 mol/L.

[79]

This Kmvalue for cyclosporin was comparable to thatfound in other cell lines expressing human P-gly-coprotein (8.4 mol/L).[94] Collectively, these re-sults strongly suggest that the efflux function ofintestinal P-glycoprotein may be saturated whendrug concentrations in the intestinal lumen exceedthe Km values after high oral doses.

As discussed previously, for drugs that are P-glycoprotein substrates the net amount of drugpassing through the intestinal epithelial cells is thedifference between the amount absorbed by influxprocesses (passive diffusion and/or active uptake)and the summation of the amount extruded by

efflux transport together with the amount meta-

bolised by enzymes. It is evident that the P-glyco-protein-mediated efflux and CYP-mediated meta-bolism are saturable processes. The saturableP-glycoprotein efflux may, at least in part, explainthe observed dose-dependent absorption of tali-nolol (a P-glycoprotein substrate) in healthy vol-unteers.[95] For both enantiomers, the dose-normalised AUC increased with increasing dosesafter oral administration. The dose-normalisedAUC of (S)-()-talinolol increased from 18 g

h/L at a 12.5mg dose to 36 g h/L at a 200mgdose. Similar results were observed for (R)-(+)-talinolol. Consistent with the in vivo observations,concentration-dependent permeability across Caco-2 cell monolayers was observed when the concen-tration of talinolol was increased from 0.12mmol/L. Similarly, saturable P-glycoprotein-mediated efflux has also been reported for cyclo-sporin in rats.[96] The extent of cyclosporin absorp-tion increased with increasing doses in rats; the

bioavailability increased from 13% at an oral doseof 6 mg/kg to 25% at 18 mg/kg. These results in-dicate that P-glycoprotein-mediated efflux trans-port can be saturated when higher oral doses aregiven.

There is a widespread misconception that theextent of oral absorption of a drug is always mark-edly limited by intestinal P-glycoprotein when thedrug is a P-glycoprotein substrate. This is only truefor a few P-glycoprotein substrate drugs that are

given at low doses. Absorption of digoxin is a goodexample. Digoxin, a well-known P-glycoproteinsubstrate, which undergoes minimal metabolism,is given orally at a very low oral dose of 0.5 to 1mg.At these doses, the concentration of digoxin in in-testinal lumen is estimated to be less than 10mol/L, which is well below the Km value (58mol/L) derived from Caco-2 cells or human co-lon.[93] Thus, P-glycoprotein plays a quantitativelysignificant role in the absorption of digoxin. The

reported low and variable absorption of digoxincan most probably be attributed to the efflux trans-port of intestinal P-glycoprotein.

P-Glycoprotein 73



16/40

However, the oral dose for most drugs is high

(>50mg) and drugs in the intestinal lumen can eas-ily reach the mmol/L concentration range. A de-tailed literature survey revealed that all of the re-ported Km values for P-glycoprotein drugs arerelatively low, ranging from 4213 mol/L (tableII). Given the low Km values for P-glycoproteindrugs, P-glycoprotein activity can readily be satu-rated when drugs are administered at high doses,and hence the role of intestinal P-glycoprotein indrug absorption becomes quantitatively less signif-icant. Indinavir, an HIV protease inhibitor, is a P-glycoprotein substrate and is given orally at a doseof 800mg. At this high dose, the indinavir concen-tration in the intestinal lumen is expected to begreater than 1 mmol/L, which is much higher thanthe Km value for P-glycoprotein transport (140mol/L) derived from Caco-2 cells.[97] Therefore,at high doses, the effect of intestinal P-glycopro-tein on indinavir absorption becomes quantita-tively less important. This can explain why in-dinavir has a reasonably good bioavailability

(>60%) in patients, even though it is a good P-gly-coprotein substrate.[98] Interestingly, in a recent lit-erature survey, Chiou et al.[99] have concluded that

the in vivo oral absorption of 13 drugs is not sig-

nificantly impeded by efflux transport, in spite ofbeing good P-glycoprotein substrates.However, there are exceptions that intestinal P-

glycoprotein still plays a significant role in absorp-tion for some drugs, even when they are given athigh doses. For example, the clinical oral dose is200700mg for cyclosporin and 100200mg forpaclitaxel, but clinical studies clearly indicatethat P-glycoprotein does play a significant role inlimiting their oral absorption.[79,82,87,101] This canbe explained by the fact that both cyclosporin andpaclitaxel have very poor water solubility, slowdissolution rate and large molecular weight (1202for cyclosporin and 854 for paclitaxel). The poorwater solubility and slow dissolution rate can resultin low drug concentration in the intestinal lumenin relation to their Km value for P-glycoproteintransport, and the large molecular size can impedethe rate of passive diffusion across the cell mem-branes. The notion of slow dissolution rate and/orslow membrane diffusion rate of cyclosporin in in-

testine is supported by the fact that peak concen-trations of the drug occur slowly at 34 hoursafter administration to patients.[101]

Table II. Apparent values of the Michaelis-Menten constant (Km) for P-glycoprotein substrates

Compound Material (flux evaluated) Apparent Km (mol/L) References

Cyclosporin Caco-2 (net B-to-A) 3.8 79

Digoxin Caco-2 (net B-to-A) 58 93

Stripped rat jejunum (net B-to-A) 81 93

Stripped rat ileum (net B-to-A) 74 93

Stripped rat colon (net B-to-A) 51 93

Stripped human colon (net B-to-A) 59 93

Etoposide Caco-2 (B-to-A) 213 100

Stripped rat jejunum (B-to-A) 94 100

Stripped rat colon (B-to-A) 119 100

Indinavir Caco-2a (net B-to-A) 140 97

Verapamil Stripped rat jejunum (B-to-A) 31 100

Stripped rat ileum (B-to-A) 29 100

Stripped rat colon (B-to-A) 4.4 100

Vinblastine Caco-2 (net B-to-A) 19 83,84

Caco-2 (net B-to-A) 27 93

Stripped rat ileum (net B-to-A) 48 93

Stripped rat colon (net B-to-A) ~100 93

a Treated with calcitriol.

B-to-A = basal to apical.

74 Lin & Yamazaki



17/40

In conclusion, it is clear that the effect of intes-tinal P-glycoprotein on drug absorption is unlikelyto be quantitatively important unless a very smalloral dose is given, or the dissolution and/or mem-brane diffusion rates of the drug are very slow.

4.2 Drug Distribution

Drugs are often administered at a location dis-tant from their intended site of action. To be effec-tive, the drug must be absorbed and transported

from the site of administration across severalbiomembranes to reach the target tissue and thesite of action. Penetrating cell membranes is acomplex process that is highly dependent on thenature of the membrane and the physicochemicalproperties of the drug. The physical and biochem-ical properties of membranes, such as lipid bilayerstructure and dynamics, play an important role indrug penetration. In addition, the physicochemicalproperties of drugs, such as hydrophobicity, ioni-

sation profile, molecular size and number of hy-drogen bonds, also play a significant role in mem-brane penetration.[102]

Although extensive efforts have been made tostudy the molecular mechanisms of the processesof drug absorption, metabolism and excretion,drug distribution has historically received muchless attention than the other processes, in spite ofits importance as a key factor in determining drugresponse. Because of this lack of attention, drug

distribution has been regarded as a forgotten rela-tive in clinical pharmacokinetics.[103] The lack ofattention stems partly from a lack of useful exper-imental tools in studying drug distribution. How-ever, the importance of drug distribution is becom-ing increasingly recognised. For example,Schinkel et al.[8,9] demonstrated large differencesin drug distribution into the brain and other tissuesbetween mdr1a (/) and mdr1a (+/+) mice. In ad-dition, with recent advances in the molecular biol-ogy and biochemistry of transporter systems, sev-eral in vitro systems have now been developed forstudying drug distribution.[104] These newly devel-oped tools and experimental methodologies will

certainly provide important insights into the mech-anisms of drug distribution in the very near future.

4.2.1 Blood-Brain Barrier

Lipophilicity and Brain Penetration

The brain is different from other organs of thebody in many aspects. One of the most importantfeatures is that the brain is anatomically separatedfrom the blood circulation by the BBB. All otherorgans are perfused by capillaries lined with endo-thelial cells that have small pores to allow formovement of drugs into the organ interstitial fluid

from the circulation. However, the endothelialcells in brain capillary blood vessels are closelyjoined to each other, leaving no space betweencells. Consequently, only lipophilic drugs cancross endothelial cells and enter the BBB by wayof passive diffusion. A strong positive correlationbetween lipophilicity and brain penetration ofdrugs has been reported by many investiga-tors.[105,106] In addition, factors other than lipophil-icity may also play an important role in the trans-

port of drugs across the BBB. For example, anegative correlation was found between the BBBpermeability of lipophilic compounds (steroid hor-mones and peptides) and the total number of hy-drogen bonds; the greater the total number of hy-drogen bonds, the lower the permeability.[105,107]

In addition, the molecular size of drugs is also animportant determinant for brain penetration.[108]

Although lipophilicity is an important factor indetermining the BBB penetration of drugs, many

lipophilic drugs have exhibited poor BBB penetra-tion. In a rat study, Levin[108] reported a good cor-relation between the in vivo BBB permeability co-efficient of 22 compounds and their lipophilicity.However, they found that vincristine and epipo-dophylotoxin displayed poor BBB permeability,despite relatively high lipophilicity (logP value of2.8). Additionally, many other lipophilic com-pounds also exhibit poor BBB penetration. For ex-ample, a potent CCKB receptor antagonist candi-date is a lipophilic compound with a logP value of3.6, and it has poor brain penetration.[109] The poorBBB penetration of these drugs cannot be ex-plained by the number of hydrogen bonds and

P-Glycoprotein 75



18/40

molecular weight. Twenty years ago, these com-

pounds were regarded as outlier compounds forbrain penetration without knowing the exact cause.The possible efflux function of P-glycoprotein

in the BBB was not connected with the observedpoor BBB permeability of lipophilic drugs until thefindings of P-glycoprotein in brain capillaries byThiebaut et al.[110] and Cordon-Cardo et al.[111] Us-ing monoclonal antibodies, they demonstrated thatP-glycoprotein is highly expressed on the apicalsurface of the endothelial cells of the brain capil-

laries. With these findings, it is now clear that theobserved poor BBB permeability of those lipo-philic drugs is due mainly to the efflux function ofP-glycoprotein.

Localisation of P-Glycoprotein in Brain

Localisation of P-glycoprotein in the brain is awidely disputed issue that has become the centreof much controversy in recent years. Beaulieu etal.[112] provided very convincing evidence for P-glycoprotein being predominantly localised in the

luminal membrane of endothelial cells of rat braincapillaries facing blood circulation. Using a noveltechnique with cationic colloidal silica, the inves-tigators were able to selectively isolate the luminalmembrane of endothelial cells of rat brain capillar-ies. The isolation procedures resulted in a mem-brane preparation with a 9.9-fold enrichment of theendothelial membrane marker protein GLUT1(glucose transporter 1) and a 17-fold enrichment ofP-glycoprotein relative to isolated brain capillar-

ies. Enrichment of GLUT1 and P-glycoprotein rel-ative to whole brain membrane preparations was280- and 500-fold, respectively. However, glialfibrillary acidic protein (GFAP), a specific markerfor astrocytes, was enriched only 1.4-fold relativeto brain capillary, suggesting a very minor contam-ination of astrocytes in the luminal membranepreparations. The co-enrichment of P-glycoprotein(500-fold) and GLUT1 (280-fold) in brain capil-lary luminal membranes compared with whole

brain membrane preparations strongly suggeststhat P-glycoprotein is expressed predominantly inthe luminal membrane of brain endothelial cells.

Consistent with Beaulieus results, Barrand et

al.,

[113]

using a dual immunostaining approach incombination with confocal microscopy, concludedthat endothelial marker C219 staining did not co-localise with astrocyte marker GFAP staining in ratbrain microvessels. Similarly, Matsuoka et al.[114]

have demonstrated that P-glycoprotein is localisedin the brain capillaries of rats, using P-glycoproteinand GFAP double-immunolabelling technique.Furthermore, a recent study by Decleves et al.[115]

showed that P-glycoprotein is expressed in bothcultured rat endothelial cells and astrocytes. Inthis study, RT-PCR analysis showed that mdr1bmRNA was preferentially expressed in astrocytes,whereas both mdr1a and mdr1b mRNA were de-tected in endothelial cells. In this study, Westernblotting analysis revealed much higher expressionlevel of P-glycoprotein in endothelial cells com-pared with astrocytes. Collectively, these resultsconsistently suggest that brain P-glycoprotein ispredominantly expressed on the apical surface ofendothelial cells of capillaries.

Although most immunohistochemical studiesindicate that P-glycoprotein is predominantlylocalised on the surface of endothelial cells facingthe luminal side, Pardridge et al.[116] claim that P-glycoprotein is localised mainly to astrocyte footprocesses. They conducted an immunochemicalstudy with human brain microvessels using (a)the MRK16 antibody to human P-glycoprotein, (b)an antiserum to GFAP and (c) an antiserum toGLUT1. They observed that the P-glycoprotein-specific antibody MRK16 bound to microvesselswith a similar, discontinuous staining pattern as anantiserum directed against GFAP. The apparentlydiscontinuous and abluminal localisation of MRK-16 staining in isolated brain capillaries and thesimilarity in immunostaining patterns by MRK-16 and anti-GFAP antibodies led the investigatorsto conclude that P-glycoprotein is expressed in as-trocyte foot processes, rather than in endothelialcells. The conclusion was further supported by

their subsequent findings that staining of the endo-thelial membrane marker protein GLUT1 was con-tinuous and showed only minimal overlap with

76 Lin & Yamazaki



19/40

MRK16 staining. With their observations, Golden

and Pardridge

[117]

have proposed a revised kineticmodel of P-glycoprotein in the brain in contrast tothe classic kinetic model.

According to the revised kinetic model of Gold-en and Pardridge,[117] a deficiency (or inhibition)of P-glycoprotein would have no effect on BBBpermeability, but would cause a decrease in thedrug concentration in the interstitial fluid (ISF) ofbrain. On the other hand, according to the classicmodel, a deficiency (or inhibition) of P-glycopro-tein would result in an increase in BBB permeabil-ity as well as an increase in drug concentration inthe ISF of brain. Therefore, comparison of brainuptake of P-glycoprotein substrates in normal andP-glycoprotein-deficient mice can be used to ad-dress the question of whether the P-glycoproteintransporter is localised in the capillary endothe-lial cells or in the astrocytes. Using an intracereb-ral microdialysis technique, de Lange et al.[118]

showed that a deficiency of P-glycoprotein in miceresulted in the same degree of increase in drug con-centrations of rhodamine-123 (a P-glycoproteinsubstrate) in the brain as well as ISF. After an in-travenous infusion of rhodamine-123, the totalbrain concentrations were about four times higherin the mdr1a (/) mice compared with wild-typemice. Similarly, the rhodamine-123 concentra-tions in ISF were also about four times higher inmdr1a (/) mice than in mdr1a (+/+) mice. Theseresults appear to be consistent with the classic con-cept that in mice the P-glycoprotein transporter isexpressed in the brain capillary endothelial cells.

Recently, Lee et al.[119] have demonstrated thatP-glycoprotein is expressed and functional in brainmicroglia. Using a continuous rat brain microgliacell line (MLS-9), immunocytochemistry studiesrevealed the location of P-glycoprotein along thenuclear envelope and plasma membrane of mi-croglia. Furthermore, the accumulation of digoxinby microglia was significantly enhanced byvalspodar (PSC-833), a potent P-glycoprotein in-

hibitor. Consistent with these findings, P-glyco-protein has also been detected in mixed glial cells,but not in primary cultured neurons, by other in-

vestigators.[114] The expression of P-glycoprotein

in glial cells may partly explain the intriguingpharmacokinetics and pharmacodynamics of [D-penicillamine2,5]enkephalin (DPDPE) in mdr1a(/) and mdr1a (+/+) mice demonstrated by Chenand Pollack.[120] Although the brain concentrationsof DPDPE were 2- to 4-fold higher in mdr1a (/)mice than in mdr1a (+/+) mice after intravenousadministration, the dose required to elicit compa-rable antinociception was more than 30-fold lowerin mdr1a (/) mice compared with mdr1a (+/+)mice. Based on brain concentrations, the EC50(concentration producing half-maximal antinoci-ception effect) of DPDPE was 13 times lower inmdr1a (/) mice compared with wild-type mice(12 versus 160 ng/g). These results suggest eitherdifferences in DPDPE distribution within the brainor differences in the intrinsic activity of-opioidreceptors between mdr1a (/) mice and mdr1a(+/+) mice. Subsequent pharmacokinetic and phar-macodynamic modelling suggested that the differ-ence in antinociception between mdr1a (/) andmdr1a (+/+) mice was due to the distribution ofDPDPE within the brain as well as between theblood and brain, but not due to differences in in-trinsic response. These results are consistent withthe notion that P-glycoprotein is expressed in othertype of brain cells in addition to the endothelialcells of capillaries.

Evidence of P-GlycoproteinInvolvement in Brain Uptake

The first experimental evidence that P-glyco-protein is involved in drug transport in the BBBcame from Tsuji and coworkers.[121] Immunostain-ing with a P-glycoprotein antibody (MRK16) dem-onstrated an exclusively apical localisation ofP-glycoprotein in the cultured bovine brain endo-thelial cells. Kinetic studies showed that effluxtransport of vincristine from the bovine brain en-dothelial cells was inhibited by verapamil, result-ing in a significant increase in intracellular drugconcentration. Similar results were also reportedby Tsuruo and colleagues[122] with mouse braincapillary endothelial cells. They showed by im-munochemical studies that P-glycoprotein was

P-Glycoprotein 77



20/40

localised on the apical surface of endothelial cells

of mouse brain. The unidirectional transport of vin-cristine from basolateral side to apical side wasdemonstrated in the polarised monolayer of mouseendothelial cells. Similarly, efflux transport ofcyclosporin was also observed in cultured endothe-lial cells of bovine and mouse brain capillar-ies.[123,124] Immunostaining with P-glycoproteinantibody also demonstrated an exclusively apicallocalisation of P-glycoprotein in human brain en-dothelial cells.[125] These in vitro studies provideevidence of functional involvement of P-glycopro-tein in the BBB penetration of drugs.

Although, in vitro studies have shown that ATPis essential for P-glycoprotein transport func-tion,[126,127] the involvement of ATP in P-glyco-protein-mediated transport was first demonstratedin vivo by Tsuji and colleagues.[128] They demon-strated that the BBB permeability coefficient ofdoxorubicin increased from 14 l/min/g brain incontrol rats to 243 l/min/g brain in rats ATP-depleted by occlusion of vertebral and commoncarotid arteries. The BBB permeability coefficientof doxorubicin was determined by using in situbrain perfusion technique. Under the experimentalconditions, the ATP content in the rats with tran-sient brain ischaemia was only 3% of that in normalrats (0.04 versus 1.43 mol/g brain). Similarly, anincrease in the BBB permeability coefficient ofcyclosporin in ATP-depleted rats was also reportedby Tsujis group.[129] These results from the ATP-depleted rat studies are consistent with the classicconcept that the P-glycoprotein is expressed in thebrain capillary endothelial cells, because the BBBpermeability decreases when P-glycoprotein func-tion is impaired.

The knockout animal model ofmdr1a(/) micealso provides a powerful tool for studying brainuptake of drugs. Oral administration of [3H]iver-mectin in mdr1a (/) and mdr1a (+/+) mice re-sulted in 87-fold higher levels of radioactivity inthe brain ofmdr1a (/) mice as compared with

wild-type mice, whereas the levels in liver, kidney,small intestine and plasma were increased by lessthan 4-fold.[9] In another study, when [3H]digoxin

and [3H]cyclosporin were given intravenously,

markedly higher brain levels of radioactivity (17-and 55-fold, respectively) were observed in mdr1a(/) mice than in mdr1a (+/+) mice.[130] Again,only a moderate increase in radioactivity levels (2-to 3-fold) of these two drugs was observed forliver, kidney, small intestine and plasma ofmdr1a(/) mice. The large differences in brain concen-tration were also observed between mdr1a/1b (/)double knockout and normal mice.[10] A 27-foldincrease in the brain concentration of digoxin wasobserved in mdr1a/1b (/) double knockout micecompared with the wild-type mice, with only a 2.5-fold increase in digoxin concentration in the liver,kidney and plasma of double knockout mice.

It is puzzling why the most marked increase indrug concentration of P-glycoprotein substrates inmdr1a (/) knockout mice is always observed inthe brain, while the increases in drug concentrationin other tissues in which P-glycoprotein is alsohighly expressed, such as liver and kidney, are rel-atively modest. As mentioned earlier, the mdrgenein mice is expressed in a tissue-specific manner.Both mdr1a and mdr1b genes are expressed in theliver and kidney, while only mdr1a gene is ex-pressed in the brain of mice.[131] At first glance, onemight assume that the less profound increases indrug concentration in the liver and kidney in mdr1a(/) mice are due to the extra protective functionsof mdr1b P-glycoprotein in these tissues. How-ever, even in mdr1a/1b (/) double knockoutmice, increases in drug concentrations in the liverand kidney are also much lower than that in thebrain.[10] The marked increases in drug concentra-tion in the brain of P-glycoprotein-deficient micealso cannot be explained by the loss of BBB integ-rity. Experiments performed using fluorescein andfluorescein-dextran-4000 as integrity markersshowed that there were no differences inbrain/plasma concentration ratio of these com-pounds between mdr1a (/) and wild-type mice,being 2.3 and 1.4%, respectively, indicating main-

tenance of BBB integrity in the absence of P-gly-coprotein.[118] The underlying mechanism for themarked increases in brain concentration remains

78 Lin & Yamazaki



21/40

unknown. Regardless of the underlying mecha-nisms, the data from the mdr1a and mdr1a/1b-knockout mice studies suggest that the brain ismore sensitive to changes in P-glycoprotein func-tion than other tissues. Therefore, P-glycoproteininhibitors should be used with caution to avoid po-tential neurotoxicity.

Another important observation from thesemdr1a and mdr1a/1b knockout mice studies is thatfor certain P-glycoprotein substrates, an apprecia-ble amount of drug is still observed in the brain ofwild-type mdr1a (+/+) mice. For example, the

brain concentrations of tacrolimus were about 2.7-fold higher than the corresponding plasma concen-trations at the same time point in mdr1a(+/+) miceafter intravenous administration.[132] Similarly,appreciable brain concentrations of cyclosporin,ranging from 3070% of the corresponding plasmaconcentration, were observed in mdr1a (+/+)mice.[130] Kinetically, the drug concentration in thebrain is determined by the difference between theamount of drug transported by influx processes

(passive diffusion and active uptake) and theamount of drug extruded by the P-glycoprotein-mediated efflux process. A fraction of drug mole-cules can reach the brain tissue if the influx diffu-sion rate is greater than the P-glycoprotein effluxrate. Nonspecific binding to brain tissue may alsobe a contributing factor in determining drug con-centration in the brain. Therefore, one should takethe rate of influx diffusion, rate of P-glycoproteinefflux transport and nonspecific binding of com-

pounds into consideration when predicting brainpenetration.

4.2.2 Placenta

As noted earlier, Lankas et al.[48] have clearlydemonstrated the protective role of placental P-glycoprotein in reducing fetal exposure toxenobiotics in CF-1 mice. The role of placentalP-glycoprotein in protection of the fetus has beenfurther evaluated using mdr1a/1b (/) doubleknockout mice.[133] Heterozygous mdr1a/1b (+/)female mice were mated with heterozygous malemice to produce fetuses of three genotypes:mdr1a/1b (/), mdr1a/1b (+/) and mdr1a/1b (+/+).

Following intravenous administration of digoxin,

saquinavir and paclitaxel to pregnant heterozy-gous dams, fetal drug exposure was much higherin the mdr1a/1b (/) fetus than the wild-typemdr1a/1b (+/+) fetus. The ratio of drug concentra-tion in the mdr1a/1b (/) fetus to that in the wild-type fetus was 2.5, 5 and 16, respectively, for dig-oxin, saquinavir and paclitaxel. On the other hand,the drug concentrations in the heterozygousmdr1a/1b (+/) fetus were similar to those in thewild-type fetus, suggesting that the P-glycoproteinlevel in the placenta of the heterozygous fetus isstill sufficient to protect the fetus. In this study, theinvestigators further demonstrated that the P-gly-coprotein inhibitors valspodar and FG-120918were able to completely block the placental P-glycoprotein function. The fetal drug concentrationsin the wild-type fetus were increased and werecomparable to those in the mdr1a/1b (/) fetusafter oral administration of the P-glycoprotein in-hibitors to heterozygous mothers. As in the brain,these results clearly demonstrate that placenta isvery sensitive to changes in P-glycoprotein func-tion. Because of potential function blockade, itis recommended that P-glycoprotein inhibitorsshould not be used in women during pregnancy toavoid excessive fetal exposure to xenobiotics.

P-glycoprotein is also highly expressed in hu-man placenta. Using monoclonal antibody C219,immunostaining showed a high expression of P-glycoprotein in trophoblasts of human placen-ta. [134] Functional P-glycoprotein activity hasbeen demonstrated in cultured human placentachoriocarcinoma epithelial cells (BeWo cells).[135]

Western blotting studies with monoclonal anti-body C219 or JSB-1 indicated that P-glycoproteinis highly expressed in BeWo cells. In the BeWomonolayer, the B-to-A transport of vinblastine,vincristine and digoxin was significantly greaterthan the A-to-B transport. Addition of cyclosporinresulted in an increase in the A-to-B transport ofthe drugs and a decrease in the B-to-A transport.

These results suggest that the placental P-glyco-protein acts as an efflux transporter by removingxenobiotics from cells. Since P-glycoprotein is ex-

P-Glycoprotein 79



22/40

pressed in human placental trophoblasts, it is likelythat placental P-glycoprotein also protects fetusesfrom xenobiotics in humans as well.

Genetic variation in the expression level of pla-cental P-glycoprotein was studied by Western blot-ting in 100 placentas obtained from Japanesewomen.[61] In this study, nine SNPs were identifiedwith an allelic frequency of 0.0050.42 (table I).Of these SNPs, G2677A (allelic frequency 0.18)and G2677T (0.39) in exon 21 were associatedwith an amino acid conversion from Ala to Thr andto Ser, respectively. Comparison of the MDR1-

genotyping and corresponding placental P-glyco-protein level revealed a correlation between the P-glycoprotein expression level and SNPs in exon 1b(T129C) and exon 21 (G2677A/G2677T). Indi-viduals with heterozygous T129C (T/C) had sig-nificantly lower levels of P-glycoprotein than thewild-type (T/T) individuals (1.07 vs 1.99, arbitraryunits). The expression levels of placental P-glyco-protein in homozygotes for wild-type allele, het-erozygotes and homozygotes of mutant allele in

exon 21 were 2.44, 1.97 and 1.45 (arbitrary units),respectively. Although the clinical implications ofinterindividual variability in placental P-glycopro-tein remains to be investigated, high placental lev-els of P-glycoprotein may provide a better protec-tion for the fetus against xenobiotics.

4.3 Drug Metabolism

It has been widely accepted that the liver is the

major site of drug metabolism because of its sizeand high content of drug-metabolising enzymes. Inaddition to the liver, the small intestine and kidneymay contribute significantly to overall metabolismin the body.[136] In humans, CYP3A4 is the princi-pal enzyme involved in the hepatic and intestinalmetabolism of drugs. Studies in rats by Debri etal.[137] suggest that the CYP3A enzymes in the liverare evenly distributed. The data of Watkins etal.,[138] who investigated the CYP3A content at tendifferent locations in a human liver, also indicatethat hepatic CYP3A is homogeneously distributed.Unlike the liver, the distribution of CYP3A4 is notuniform along the length of the small intestine, or

along the villi within a cross-section of mucosa.

Using a monoclonal antibody to CYP3A, the co-lumnar absorptive epithelial cells of the villi exhib-ited the strongest immunoreactivity, whereas noimmunostaining was detectable in the goblet andcrypt cells.[139] It has also been shown that CYP-3A4 expression varied along the length of smallintestine: median values of 31, 23 and 17 pmol/mgmicrosomal protein were found in human duode-num, distal jejunum, and distal ileum.[140] In con-trast to CYP3A4, the expression of P-glycoproteinappears to increase progressively along the lengthof intestine.[78]

There is a striking overlap between CYP3A4substrates and P-glycoprotein substrates. Becauseof overlapping substrate specificity, and becauseof coexpression of CYP3A enzymes and P-glyco-protein in the intestine, kidney and liver, it isconceivable that P-glycoprotein may play an im-portant role in drug metabolism. However, themagnitude of the effect of P-glycoprotein on me-tabolism appears to be dependent on the spatialrelationship between P-glycoprotein and CYP3Aenzymes. In the liver and kidney, P-glycoprotein islocalised on the luminal membrane of hepatic can-aliculi facing the bile duct lumen or on the luminalbrush-border membrane of renal proximal tubularcells facing the renal tubule lumen. This means thatP-glycoprotein is localised at the exit site of hepa-tocytes and renal epithelial cells. Therefore, P-gly-coprotein only sees drug molecules after cellularuptake, intracellular distribution and metabolismin both the liver and kidney.

In contrast to the situation in the liver and kid-ney, P-glycoprotein is localised at the entrance siteof epithelial cells of intestines. Drug molecules areexposed to P-glycoprotein prior to intracellulardistribution and metabolism. A large fraction ofdrug molecules is extruded by intestinal P-glyco-protein from the inside of the epithelial cells backinto the intestinal lumen after the drug moleculesgain access across the luminal surface of the epi-

thelial cells; however, a portion of the extrudeddrugs then can be reabsorbed into the epithelialcells. Through the repetitive processes of extrusion

80 Lin & Yamazaki



23/40

and reabsorption, P-glycoprotein prolongs the in-

tracellular residence time of drug molecules andincreases the probability of exposure to drug-me-tabolising enzymes. Consequently, P-glycoproteinmay enhance intestinal metabolism of drugs,whereas it has less of an effect on drug metabolismin liver and kidney.

The effect of P-glycoprotein on CYP3A4-mediated intestinal metabolism of indinavir, asubstrate for both CYP3A4 and P-glycoprotein,has been carefully evaluated in our laboratory us-ing calcitriol-treated Caco-2 cells expressing bothCYP3A4 and P-glycoprotein.[97,141] The formationof the major metabolite (M6), expressed as the ra-tio of the amount of the metabolite formed to theamount of the parent drug transported across themonolayer, was more than 6-fold greater when thedrug was applied at the apical side than when thedrug was applied at the basolateral side. Similarly,the metabolism of cyclosporin in Caco-2 cellswas higher from the apical side than from the baso-lateral side.[142] These results strongly suggest arole of P-glycoprotein in enhancement of CYP3A4-mediated intestinal metabolism of drugs.

The effect of P-glycoprotein on intestinal meta-bolism was also shown in vivo. The intestinal first-pass metabolism of indinavir increased from 6%in control rats compared with 34% in dexametha-sone-treated rats.[143] Pretreatment of rats with dex-amethasone (40 mg/kg orally for 3 days) resultedin a 2.5-fold increase in both CYP3A and P-glyco-protein levels in the intestine. The 6-fold increasein intestinal first-pass metabolism of indinavir can-not be explained by the 2.5-fold increase in intes-tinal CYP3A level alone. The increased intestinalfirst-pass metabolism is most probably due to acombination of increased intestinal CYP3A andP-glycoprotein levels, providing in vivo evidencethat P-glycoprotein enhances the intestinal first-pass metabolism of indinavir.

In this study, the effect of P-glycoprotein onhepatic metabolism was also investigated. Pre-

treatment of rats with dexamethasone also inducedhepatic first-pass metabolism of indinavir, and theincreased hepatic CYP3A enzyme activity alone

appeared to be able to explain the increased hepaticfirst-pass metabolism, suggesting that P-glycopro-tein plays a less significant role in hepatic metabo-lism as compared with intestinal metabolism.[143]

These results clearly suggest that P-glycopro-tein may play an important role in intestinalmetabolism of drugs. However, it should be re-emphasised that the effect of P-glycoprotein ondrug metabolism becomes quantitatively less sig-nificant when high doses are given.[144]

4.4 Drug Excretion

Drugs are generally eliminated from the bodyby metabolism and/or excretion. Both the liver andkidney play an important role in the excretion ofunchanged drugs and their metabolites. In princi-ple, biliary excretion and renal tubular excretionshare certain characteristics. For biliary excretion,a drug must first traverse the sinusoidal (basolat-eral) membrane of the hepatocytes by passive dif-

fusion and/or hepatic uptake transporters. The si-nusoidal membrane of the hepatocyte contains anumber of active transporters responsible for theuptake of cations, anions and endogenous sub-stances into hepatocytes from the circulation.[145]

Once in the hepatocytes, the drug molecules con-tinue to diffuse and reach the canalicular mem-brane, where P-glycoprotein and other effluxtransporter systems will pump the drug moleculesinto bile. Often, biotransformation occurs when

the drug molecules are passing through the hepa-tocytes. Therefore, hepatic uptake, intracellulardiffusion and metabolism, as well as other effluxtransporter systems, have to be taken into consid-eration when biliary excretion of drugs is evalu-ated.

Similarly, uptake of drugs across the basolateralmembrane of renal epithelial cells is the first stepin renal excretion, and biotransformation may oc-cur. The basolateral membrane contains a numberof active transporters responsible for drug uptake.The luminal brush-border membrane also containsnumerous active transporters, inc

Date post:	03-Apr-2018
Category:	Documents
Upload:	omar-cruz-badillo
View:	218 times
Download:	0 times

Gli Co Protein A

Documents