21
Alternative Measures of Lipophilicity: From Octanol–Water Partitioning to IAM Retention COSTAS GIAGINIS, ANNA TSANTILI-KAKOULIDOU Department of Pharmaceutical Chemistry, School of Pharmacy, University of Athens, Panepistimiopolis, Zografou, Athens 157 71, Greece Received 13 June 2007; revised 8 October 2007; accepted 8 October 2007 Published online in Wiley InterScience (www.interscience.wiley.com). DOI 10.1002/jps.21244 ABSTRACT: This review describes lipophilicity parameters currently used in drug design and QSAR studies. After a short historical overview, the complex nature of lipophilicity as the outcome of polar/nonpolar inter- and intramolecular interactions is analysed and considered as the background for the discussion of the different lipophilicity descriptors. The first part focuses on octanol–water partitioning of neutral and ionisable compounds, evaluates the efficiency of predictions and provides a short description of the experimental methods for the determination of distribution coefficients. A next part is dedicated to reversed-phase chromatographic techniques, HPLC and TLC in lipophilicity assessment. The two methods are evaluated for their efficiency to simulate octanol–water and the progress achieved in the refinement of suitable chromatographic conditions, in particular in the field of HPLC, is outlined. Liposomes as direct models of biological membranes are examined and phospolipo- philicity is compared to the traditional lipophilicity concept. Difficulties associated with liposome–water partitioning are discussed. The last part focuses on Immobilis- ed Artificial Membrane (IAM) chromatography as an alternative which combines membrane simulation with rapid measurements. IAM chromatographic retention is compared to octanol–water and liposome–water partitioning as well as to reversed- phase retention and its potential to predict biopartitioning and biological activities is discussed. ß 2008 Wiley-Liss, Inc. and the American Pharmacists Association J Pharm Sci 97:2984–3004, 2008 Keywords: lipophilicity; octanol–water system; reversed phase chromatographic retention; extrapolated retention factors; liposome–water partitioning; IAM chromato- graphy INTRODUCTION Lipophilicity, expressed by the logarithm of octanol– water partition coefficient log P or distribution coefficient log D, if ionised molecular species are present, constitutes a physicochemical property of paramount importance in medicinal chemistry. It plays an essential role in absorption, distribution, metabolism and elimination (ADME) character- istics of drugs while affecting also their pharma- codynamic and toxicological profile. 1–4 Historically, the importance of lipid solubility in biological activity was demonstrated already at the dawn of 20th century, when Meyer 5 and Overton 6 independently published their work on narcosis. They were the first who used oil–water partition Correspondence to: Anna Tsantili-Kakoulidou (Telephone: þ30-210-7274530; Fax: þ30-210-7274747; E-mail: [email protected]) Journal of Pharmaceutical Sciences, Vol. 97, 2984–3004 (2008) ß 2008 Wiley-Liss, Inc and the American Pharmacists Association. 2984 JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 97, NO. 8, AUGUST 2008

log p

Embed Size (px)

DESCRIPTION

log p

Citation preview

Page 1: log p

Alternative Measures of Lipophilicity:From Octanol–Water Partitioning to IAM Retention

COSTAS GIAGINIS, ANNA TSANTILI-KAKOULIDOU

Department of Pharmaceutical Chemistry, School of Pharmacy, University of Athens, Panepistimiopolis, Zografou,Athens 157 71, Greece

Received 13 June 2007; revised 8 October 2007; accepted 8 October 2007

Published online in Wiley InterScience (www.interscience.wiley.com). DOI 10.1002/jps.21244

Correspondeþ30-210-727453E-mail: tsantili@

Journal of Pharm

� 2008 Wiley-Liss

2984 JOURN

ABSTRACT: This review describes lipophilicity parameters currently used in drugdesign and QSAR studies. After a short historical overview, the complex nature oflipophilicity as the outcome of polar/nonpolar inter- and intramolecular interactionsis analysed and considered as the background for the discussion of the differentlipophilicity descriptors. The first part focuses on octanol–water partitioning ofneutral and ionisable compounds, evaluates the efficiency of predictions and providesa short description of the experimental methods for the determination of distributioncoefficients. A next part is dedicated to reversed-phase chromatographic techniques,HPLC and TLC in lipophilicity assessment. The two methods are evaluated fortheir efficiency to simulate octanol–water and the progress achieved in the refinementof suitable chromatographic conditions, in particular in the field of HPLC, is outlined.Liposomes as direct models of biological membranes are examined and phospolipo-philicity is compared to the traditional lipophilicity concept. Difficulties associatedwith liposome–water partitioning are discussed. The last part focuses on Immobilis-ed Artificial Membrane (IAM) chromatography as an alternative which combinesmembrane simulation with rapid measurements. IAM chromatographic retention iscompared to octanol–water and liposome–water partitioning as well as to reversed-phase retention and its potential to predict biopartitioning and biological activitiesis discussed. � 2008 Wiley-Liss, Inc. and the American Pharmacists Association J Pharm Sci

97:2984–3004, 2008

Keywords: lipophilicity; octanol–water

system; reversed phase chromatographicretention; extrapolated retention factors; liposome–water partitioning; IAM chromato-graphy

INTRODUCTION

Lipophilicity, expressed by the logarithm of octanol–water partition coefficient log P or distributioncoefficient log D, if ionised molecular species

nce to: Anna Tsantili-Kakoulidou (Telephone:0; Fax: þ30-210-7274747;pharm.uoa.gr)

aceutical Sciences, Vol. 97, 2984–3004 (2008)

, Inc and the American Pharmacists Association.

AL OF PHARMACEUTICAL SCIENCES, VOL. 97, NO. 8, AUG

are present, constitutes a physicochemical propertyof paramount importance in medicinal chemistry. Itplays an essential role in absorption, distribution,metabolism and elimination (ADME) character-istics of drugs while affecting also their pharma-codynamic and toxicological profile.1–4 Historically,the importance of lipid solubility in biologicalactivity was demonstrated already at the dawnof 20th century, when Meyer5 and Overton6

independently published their work on narcosis.They were the first who used oil–water partition

UST 2008

Page 2: log p

ALTERNATIVE MEASURES OF LIPOPHILICITY 2985

coefficients in their studies, stimulating theinterest for further investigation, such as the workof Gaudette and Brodie7 who found a parallelismbetween heptane–water partition coefficients ofcertain drugs and their penetration rate throughthe blood brain barrier. However, it was Hansch andFujita8 and Leo et al.9 who organised the systematicresearch on lipophilicity, introducing the octanol–water system as a reference to construct anempirical scale for this property which could beused in the newly developing at that time QSARanalysis, known also as Hansch analysis. Sincethen, octanol–water partition coefficient in itslogarithmic form, log P, became a characteristicconstant for a chemical substance. Log P valueshave been compiled in commercially availabledatabases.10–12 Nowadays, Pomona College Med-Chem Database contains more than 53000 log Pvalues, among them 11000 are characterised ashigh quality belonging to the ‘star’ list.12 More to thepoint, Hansch and his coworkers created the basisfor the calculation of lipophilicity, recognizing thepartially additive character of log P. Up to now agreat number of QSAR models have been published,the majority of which include lipophilicity asdescribed by octanol–water partition coefficient,while optimum log P values (log Po) for the penetra-tion through certain biological barriers have beenestablished.13–16 The dominant role of octanol–water partition coefficients in the pharmacokineticprocesses, as well as in ligand–macromolecule inter-actions triggered further research on the develop-ment of rapid methods for their experimentalassessment, among them reversed-phase chromato-graphic techniques, as well as of suitable calculationmethods, so that screening of compound libraries isnow possible. On the other hand, guidelinesreferring to druglike characteristics include upperlimits for log P values,17,18 since there is strongevidence that high lipophilicity is associated withundesired drug features, like extensive and un-predictable metabolism, high plasma protein bind-ing or accumulation to tissues.19 The recognitionof the imperative role of lipophilicity in drugdesign led to the organisation of three successfulinternational Symposia focused especially on thisproperty.20–22

Despite its successful application in drug design,the octanol–water system has received a lot ofcriticism throughout these years, as an isotropicmedium with only a superficial similarity tobiomembranes. Partitioning into liposomes andImmobilised Artificial Membrane (IAM) chromato-graphy are gaining increasing interest, especially

DOI 10.1002/jps JOUR

the latter which combines simulation of cellmembranes with rapid measurements.23–25

Lipophilicity, however, is not merely a usefulparameter in drug design. Being the outcome ofmore than one component, it encodes a highcontent of structural information which can beused to understand better the behaviour of asolute in the biological environment.26 In thisaspect, the differences between the variouslipophilicity descriptors may concern intermole-cular recognition forces, developed between themolecule and the two phases, or intramolecularinteractions, which may differ in their manifesta-tion between the different biphasic systems. In thepresent review, the complex nature of lipophilicitywill serve as the background for the comparison ofalternative measures of lipophilicity. Consideringoctanol–water as the reference system for lipo-philicity, weaknesses associated with the effi-ciency of predictions are outlined and similarities/dissimilarities with reversed-phase chromato-graphic indices are discussed. Liposome partition-ing and IAM chromatography are treated in ananalogous manner and their merit to correlatebiological data is evaluated.

THE NATURE OF LIPOPHILICITY

The dual nature of lipophilicity as the outcome ofnonpolar and polar interactions is well establish-ed.26–30 Nonpolar interactions are equated withhydrophobicity and can be expressed by stericterms like molecular volume, polarizability, molarrefractivity. They contribute positively to lipophi-licity. More complex are the polar interactionswhich include ion–dipole, dipole–dipole or hy-drogen bond interactions and can be expressedby electronic constants, dipole moments or hy-drogen bond parameters. A global polarity para-meter, designated as L, has been proposed byTesta.27 Intramolecular interactions, promoted bycertain structural characteristics, like substitu-ents in aromatic or heterocyclic systems, presenceof conjugated systems in the molecules, internalhydrogen bonds, also affect lipophilicity consider-ably influencing the intermolecular forces.31–33

In addition, conformational changes inducedby polar or nonpolar environment may rendermolecules more hydrophilic or more lipophilicthan expected, depending on the nature of thegroups buried within the molecule or exposed tothe solvent.29,34 Ionisation dramatically alters alltypes of intramolecular interactions involving the

NAL OF PHARMACEUTICAL SCIENCES, VOL. 97, NO. 8, AUGUST 2008

Page 3: log p

2986 GIAGINIS AND TSANTILI-KAKOULIDOU

charged centre, limiting the validity of additivityrules established for neutral compounds.35,36

The outcome of polar/nonpolar inter- andintramolecular interactions can be summarisedin Eq. (1):

Lipophilicity ¼ hydrophobicity � polarity (1)

The above-mentioned issues refer to any biphasicsystem and are solvent (phase) dependent. Thedifferences between the systems can be demon-strated by the so-called solvatochromic analysis.37–39

Partition coefficients are factorised into steric andpolar terms which express different types of inter-molecular interactions as described in Eq. (2):

log P ¼ vV=100 þ aA þ bB þ sS þ eE þ c (2)

where V represent Mc Gowan’s characteristicmolar volume, A and B (formerly a and b) expressoverall solute’s hydrogen-bonding acidity andbasicity respectively, S (formerly p�) is a combin-ed dipolarity/polarizability term reflecting thesolute’s capacity to elicit orientation and induc-tion forces and E (formerly R) is excess molarrefractivity. v, a, b, s, e, c are constants whichcharacterise the system and are derived bymultiple linear regression analysis.40,41

The terms in Eq. (2) are significant when thegiven type of interaction elicited between thesolute and the two phases is not of equal energy.The coefficients of the terms reflect then themagnitude of this difference.

According to this type of analysis octanol–water partition coefficients differ from alkane(hexadecane)–water partition coefficients mainlyin the acidity term A, which is not significant inthe first case but has a large negative contributionin the latter. This is shown in Eqs. (3) and (4)derived for a large number of data measured in then-octanol–water system and in the alkane–watersystem respectively:39

log P ¼ 3:81V þ 0:034A � 3:460B � 1:054S

þ 0:562E þ 0:088

n ¼ 614; r ¼ 0:997 s ¼ 0:116

(3)

where n refers to the number of data derived fromthe n-octanol–water system.

log P ¼ 4:28V � 3:587A � 4:869B � 1:617S

þ 0:667E þ 0:087

n ¼ 370; r ¼ 0:998; s ¼ 0:124

(4)

JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 97, NO. 8, AUGUST 2008

where n refers to the number of data derived fromalkane–water system.

The contribution of the hydrogen bond basicityterm B in octanol–water partitioning (Eq. 3) hasbeen widely recognised. Some authors comparedsuccessfully the basicities of solutes estimated byequations of type 2 to those experimentallydetermined in CCl4.42,43

Partitioning into n-Octanol

n-Octanol–water is the widely accepted referencesystem for the determination of lipophilicity.n-Octanol offers certain advantages which ad-vocate for its suitability to simulate biologicalmembranes. Initially, these advantages wereattributed to its structure, namely the presenceof a hydrophobic chain with a polar head, as wellas to its moderate water saturation, which allowsrelatively easier and more reliable experimentaldetermination.9 The hydroxyl group in then-octanol molecule has a hydrogen bond capabilityboth as donor and acceptor. However, membranesand receptors are different as far as theirhydrogen bond properties are concerned. Thus,they may contain amphiprotic groups, and in suchcases n-octanol is the most suitable lipophilicphase, or they may contain mainly hydrogenbond donors or hydrogen bond acceptors. In thefirst case chloroform–water may provide a bettersimulation, while in the second case dibutyletheror propyleneglycoldiperlagonate have been sug-gested as lipophilic phases. Finally, it is possiblethat no hydrogen bond sites are available, so analkane–water system or dichloroethane–watersystem would be preferable.39,44–48

The degree of water saturation of organicsolvents may affect the accuracy of the partition-ing experiments, since very high or very low watersaturation values are associated with practicaldifficulties. For n-octanol, the degree of saturationis 0.25 mol % and, as already commented, itpermits more reliable measurements. More thanthat, the water present in wet octanol seems toplay an essential role in the interaction withsolutes rendering its structure more complex thananticipated for an isotropic system.49,50 Accordingto the results of NMR and near IR as well as ofX-ray diffraction analysis, the water is notuniformly dispersed in octanol, but it formsclusters of four molecules. The water moleculesare surrounded by about 16 octanol molecules,which orientate their hydroxyl groups towards the

DOI 10.1002/jps

Page 4: log p

ALTERNATIVE MEASURES OF LIPOPHILICITY 2987

water molecules constructing a network of hydro-gen bonds. On the other hand, the hydrocarbonchains create a water free region similar with thatin the interior of lipid bilayer with a dielectricconstant e¼ 8.51–53

Partitioning into an organic solvent–watersystem is generally entropy driven, althoughoctanol has an exothermic heat of transfer(negative enthalpy), due to H-bond stabilisationof the transferred solute. Thermodynamic data aredifferent regarding cyclohexane–water partition-ing for which a positive enthalpy has been found.54

Figure 1. Log D/pH profile of (A) mefenamic acid(B) alprenolol, in presence of 0.01 M (&) and 0.15 M(&) KCl. Data calculated according to Prolog D (Pallas v3.3.2.4).

Ionisation Effect in Partitioning

Partition coefficient refers to the neutral monomerspecies. If ionic species coexist with the neutralspecies, the distribution coefficient is considered asthe ratio of the sum of the concentration of allspecies in octanol to the corresponding concentra-tion sum in water. If partition of ions in octanol isneglected, log D can be converted to log P by meansof a correction term, based on the Henderson–Hasselbach equation.55 Eqs. (5) and (6) are used formonoprotic acids and bases, respectively, denotingthat distribution coefficients depend on pH assum-ing that only the neutral form partitions into thelipid phase.56

Log D ¼ log P � logð1 þ 10pH�pKaÞ (5)

Log D ¼ log P � logð1 þ 10pKa�pHÞ (6)

Theoretically, the slope of the linear part ofEqs. (5) and (6), plotting log D versus pH, is j1j,although deviations may be observed.36 In case ofextended ionisation, the nature, as well as theconcentration of the counter ions define the lowestlevel asymptote corresponding to the partitioncoefficient of the ionised species. Figure 1 showsan example of lipophilicity profiles, log D versuspH, for an acid (mefenamic acid) and a base(alprenolol) in presence of 0.01 M and 0.15 M KCl(data calculated according to Prolog D, imple-mented in Pallas v. 3.3.2.4, see also LipophilicityPrediction Section). As a rule of thumb, thedifference between partition coefficients of theneutral and ionised form is considered to be �3 logunits for bases and �4 log units for acids at 0.15 MKCl.56,57 This general rule may not be valid if theionised species produce charge delocalisation,contain hydroxyls adjacent to carboxylic groupsor as a result of steric hindrance.57,58 Specialattention deserves the lipophilic behaviour of

DOI 10.1002/jps JOUR

ampholytes. Ordinary ampholytes (pKaacid>pKabasic) exhibit an acid and base lipophilicity/pH profile described by Eq. (7).

log P ¼ log D

þ logð1 þ 10pKa1�pH þ 10pH�pKa2Þ (7)

where pKa1 is the basic pKa and pKa2 is theacidic pKa.

In the case, however, of close overlapping pKa’sor if pKaacid<pKabasic, zwitterionic species areformed around the isoelectric point. The presenceof a positive and a negative charge alters theintramolecular interactions and often contra-dictory results are reported in literature concern-ing their lipophilicity.59–61 Partial neutralisationoccurring via through-bond and through-spaceinteractions leads to distribution coefficienthigher than that corresponding to the cationicor anionic species, but lower than the log P ofthe neutral form. Charge compensation results ina bell shaped lipophilicity profile with constant

NAL OF PHARMACEUTICAL SCIENCES, VOL. 97, NO. 8, AUGUST 2008

Page 5: log p

2988 GIAGINIS AND TSANTILI-KAKOULIDOU

lipophilicity in a pH range defined by the two pKa

values. Thus, zwitterionic ampholytes are con-sidered to behave like ‘lipophilicity buffers’. Insome cases structural factors do not allow chargecompensation and a U-shaped log D/pH profilemay be obtained.59,61

Lipophilicity Prediction

Lipophilicity prediction is essential in early drugdiscovery process and this necessity was recog-nised from the beginning of the QSAR era. Theintroduction of hydrophobic substituent constant,p, in the 1960s was followed by the fragmentalsystems of Rekker and Mannhold,62 Leo63 andHansch and Fujita,8 based on the additive andconstitutive character of log P. The latter isencoded in C log P, which is considered by manyauthors as the benchmark software. Calculation isperformed by summation of fragmental contribu-tions, while certain corrections are needed toaccount for deviations from additivity. The nextstep was the development of atomic contributionsystems, which are based on atom types, takinginto consideration their neighbouring structuralenvironment. The advantage of atomic contribu-tion systems is that most of them do not requirecorrection terms.64–67 Recently, nonlinear sys-tems based on the construction of artificial neuralnetworks (ANN) have been proposed. Electroto-pological state indices are used as input data, thusincorporating electronic effects in the calculationprocedure.68,69 Atom types have also been used inthe construction of ANN for log P prediction.70

Despite the large arsenal of calculation systems,lipophilicity prediction is not an easy task. Oftenfor complex molecules different systems yielddifferent estimates, which may considerablydeviate from the true log P value.29,33 The sourcesfor prediction errors are associated with thecomplexity in the inter- and intramolecularinteractions involved in partitioning process aswell as with conformational effects, as alreadydiscussed. To this point it should be noted that themolecular lipophilicity potential (MLP) approach,developed by Testa et al. aimed to confrontthose problems. MLP is a useful tool in molecularmodelling, however it has not found generalapplication in lipophilicity screening of large datasets.71 Moreover, the systems described so farpredict the lipophilicity of the neutral species.Limited choices are available for the predictionof distribution coefficients and the results are

JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 97, NO. 8, AUGUST 2008

susceptible to larger errors, since ionisation mayalter the additivity rules as previously comment-ed. Ionisation correction terms are based onHenderson–Hasselbach equation, while the ionicstrength should also be considered. Prolog Dimplemented in Pallas software (CompuDrugInternational, Inc., Sedona, AZ) calculates log Dvalues at any pH and any K/Na]þ/[C]l� concentra-tion,72 while ADME Boxes (PharmaAlgorithms,Torontto, Canada) and ACD Labs consider astandard ionic strength of 0.15 M KCl/NaCl.

A classification of the different methods forcalculation of lipophilicity along with the relevantsoftware programs can be found elsewhere.73

Measurement of Partition/Distribution Coefficients

Different experimental protocols for log P/log Ddetermination have been suggested in literature.The classical shake-flask method for direct parti-tioning experiments is tedious and time consum-ing, not suitable for degradable compounds, lessamenable to automation, while it presents limita-tions concerning the log P or log D range that canbe reliable measured. Detailed reviews havedescribed the associated pitfalls and difficulties,as well as how to choose the best conditionsfor reliable measurements.74–76 For ionisablecompounds potentiometric dual phase pH metrictitration is possible.77–79 Partition coefficients arecalculated according to the difference in the pKa

values, in presence and absence of n-octanol (or anyother immiscible organic solvent), as described byEqs. (8a) and (8b) for acids and bases respectively:

P ¼ ð10poKa�pKa � 1ÞVH2O=Vorg

for monoprotic acids(8a)

P ¼ ð10�ðpoKa�pKaÞ � 1ÞVH2O=Vorg

for monoprotic bases(8b)

where poKa is the apparent pKa in presence of theorganic solvent.

Centrifugal chromatography offers another pos-sibility for direct log P or log D measurements.80

Attempts for the development of high throughputmeasurements scaled down to 96-well microtitreplate techniques have also been reported.81

Summarizing The Nature of Lipophilicity Sec-tion n-octanol–water partition or distributioncoefficients remain through the decades a usefultool in the design of new drugs while they providevaluable information for the understanding of

DOI 10.1002/jps

Page 6: log p

ALTERNATIVE MEASURES OF LIPOPHILICITY 2989

solutes in the biological environment. Calculationsystems permit a rapid lipophilicity screening oflarge drug libraries. However due to the interplayof different forces involved in n-octanol–waterpartitioning, calculations may be less reliable forcomplex structures and in particular for chargedspecies. Thus, experimental determination at alater drug development stage remains necessary.Although there is considerable progress to faci-litate partitioning experiments, many researchersdirect their efforts in the exploitation of user’sfriendlier chromatographic techniques, which willbe discussed in the following section.

ALTERNATIVE MEASURES OFLIPOPHILICITY-REVERSED-PHASECHROMATOGRAPHIC INDICES

Reversed-phase chromatographic techniques, inparticular high performance liquid chromato-graphy (RP-HPLC) and thin layer chromatogra-phy (RP-TLC) have proved to simulate octanol–water partitioning and are considered as popularalternatives for lipophilicity assessment. Theyoffer several practical advantages, includingspeed, reproducibility, insensitivity to impuritiesor degradation products, broader dynamic range,online detection and reduced sample handlingand sample sizes.82–87 These advantages haveattracted considerable interest and literature isrich in research articles, which investigate therelationship of chromatographic retention withoctanol–water partitioning and the commonfactors underlying the two processes.84,87–92 Theselection of either technique is associated withthe state of the art concerning their technology.Especially RP-HPLC got wide acceptance inlipophilicity assessment and has officially beenrecommended by the OECD.93 Moreover, allassumptions dealing with the complex nature oflipophilicity as the outcome of intermolecularand intramolecular interactions, involving elec-tronic, steric, or conformational effects, embracechromatographic retention, as well.

The lipophilicity indices measured by RP-HPLCand RP-TLC are derived by the retention time tr

and the migration distance relative to the solventfront, Rf, both converted to the logarithm of theretention factors log k and RM, according toEqs. (9) and (10), respectively:

log k ¼ logtr � to

to

� �(9)

DOI 10.1002/jps JOUR

tr being the retention time of an unretained solute.

RM ¼ log1

Rf� 1

� �(10)

Rf being the migration distance of the solute Zf

divided by the distance of the solvent front fromthe starting point Zs.

Rf ¼Zf

Zs(11)

Isocratic retention factors represent a relativescale of lipophilicity and are converted to log P orlog D values via Collander type Eqs. (12) and (13):

log Pðor log DÞ ¼ a log k þ b (12)

log Pðor log DÞ ¼ a RM þ b (13)

Isocratic retention factors, log k and RM, arepreferred by some authors, as they require fewerexperiments.94 However care should be taken touse an adequate training set of solutes in respect tothe solutes for which the log P or log D values haveto be determined, since often isocratic retentionfactors may lead to under- or overestimation oflipophilicity.89,95 In this aspect, extrapolated reten-tion factors, corresponding to pure water as mobilephase and termed as log kw, RMw are consideredas more representative lipophilicity indices, sincetheir values are of the same order of magnitudewith octanol–water log P and log D.82–87 In fact,extrapolated log kw or RMw values are derivedusing the linear part of the log k/w or RM/wrelationships according to Eqs. (14) and (15),respectively:

log k ¼ �S’þ log kw (14)

RM ¼ �S’þ RMw (15)

where w is the concentration of the organicmodifier in the mobile phase.

In many cases, when log kw or RMw values areused in equation of type (12) or (13), thecorresponding regression coefficients a and b tendto approach 1 and 0, respectively. Considerableresearch efforts are directed towards the stan-dardisation of chromatographic conditions, whichguarantee 1:1 correlation between extrapolatedchromatographic indices and log D values. In thisfield, outstanding progress has been achieved inthe case of RP-HPLC. Currently, standardisationof chromatographic conditions for basic andneutral drugs at physiological pH has beenestablished and analogous efforts for acidic andzwitterionic drugs are bestowed.96–98 As regards

NAL OF PHARMACEUTICAL SCIENCES, VOL. 97, NO. 8, AUGUST 2008

Page 7: log p

2990 GIAGINIS AND TSANTILI-KAKOULIDOU

the RP-TLC, the results are less spectacular, eventhough it is assumed that both reversed phasetechniques are governed by analogous retention.However, it should be noted that the stationaryphase in RP-HPLC undergoes a uniform, rela-tively well-defined transformation during equili-bration process with the mobile phase, while inthe case of RP-TLC a stationary phase with agradient composition is formed upon equilibrationcreating less reproducible conditions. Moreover,important factors in HPLC retention, like themobile pH and mobile phase additives do notaffect the retention on the TLC plates to the sameextent, thus reducing the flexibility to investigatedifferent chromatographic conditions.99–103 Forthis reason, the next paragraph will focus onstationary/mobile phase alternatives in RP-HPLCand reference to RP-TLC will be assigned whennecessary.

CHROMATOGRAPHICCONDITIONS—STATIONARY PHASES

C-18 silanised silica gel is the preferred materialfor reversed phase chromatographic columns andplates for the assessment of drug lipophilicity.However, the interference of silanophilic interac-tions in the partitioning mechanism in RP-HPLCand RP-TLC has been recognised long ago as aserious drawback.104–107 Silanophilic interactionsare attributed to the remaining free silanol sitesand include hydrogen bonding, as well as electro-static forces, especially in the case of positivelycharged compounds, producing considerableincrease in retention. They also depend on thedegree of ionisation of the silanol groups, beingless pronounced at low pH.108,109 The problem isextensively studied in the case of RP-HPLC andpartially faced by the development of columnswith reduced free or accessible silanol sitesand the use of masking agents as mobile phaseadditives. Analogous progress is not achieved forTLC layers for the reasons mentioned above. InRP-TLC, the poor wetability of the plates presentsan additional problem, more pronounced forfully silanised silica gel material. The spreadof overpressure TLC in the future may offer asolution,110,111 but up to now, application ofoverpressure TLC in the field of log P assessmenthas not been reported.

End-capping of the silanol residues by trimethy-lchlorosilane (TMCS) or hexamethyldisilazane(HMDS) is usually performed during manufacture

JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 97, NO. 8, AUGUST 2008

process, leading to a higher degree of silanisa-tion.105 Hence, base deactivated silica represents apacking material more suitable (e.g. BDS C-18) forbasic solutes. In addition, recent technology has ledto the development of polar embedded and polarend-capped stationary phases, which are consid-ered to be further deprived from silanophiliceffects.91,96,112,113 They contain polar functionalgroups such as amide, carbamate, ether orsulphonamide, incorporated at the bottom of thealkyl-bonded chains. These functional groupsprovide electrostatical shielding to the surfacesilanol sites. The LC-ABZþ and the Discovery-RP-Amide-C16 stationary phases belong to thepolar embedded type columns, which have beenused for lipophilicity determination. However,these packing materials may exhibit other polarinteractions with analytes, such as the stronginteraction between the polar embedded groupsand the phenolic analytes.113 In respect to the polarend-capped columns, a second reaction is used tobond a short carbon chain (usually C3–C4) witha polar end to the surface silanol sites.114 Afavourable advantage for both types is the factthat a higher degree of orientation for the alkylchains is achieved and thus they can be used withmobile phases containing high amounts of water oreven pure water without the problem of hydro-phobic collapse.112–114 The pH limitation of theabove-mentioned columns lies in the range 2.5–7.5.Thus, for strong bases determination of retentionfactors corresponding to the neutral form cannot beachieved.

Recently bidentate stationary phases (e.g.Zorbax-extend C-18) that include a propylenebridge, as well as surface modified silica columns(e.g. XTerra C-18), where organic functionalgroups have become a constituent of the silicabackbone, have been developed, allowing the useof mobile phases with pH up to 12.115 However,the applicability of such columns in the lipophi-licity assessment has not been systematicallyinvestigated yet. In a recent publication, 1:1correlation has been reported between log kw

and log P for 40 basic compounds measured atpH 10.5 on a Zorbax-extend C-18 column withoutaddition of any masking agent.116

An alternative choice, the polymer-basedoctadecyl-poly(vinyl alcohol) (ODP) stationaryphase, which is completely devoid of reactivesilanol groups, has also been used for lipophilicitymeasurements.117,118 ODP column presents stabil-ity to acidic and strongly basic conditions (atpH between 2 and 13).119 However, it has been

DOI 10.1002/jps

Page 8: log p

ALTERNATIVE MEASURES OF LIPOPHILICITY 2991

reported that the retention mechanism on ODPstationary phase compared to octanol–water par-titioning is controlled by a different balance offorces as deduced by solvatochromic analysis.120

In fact, there was a lower contribution of volumeand hydrogen bond basicity, while the polariz-ability term was found nonsignificant in therelevant equation. Thus, the derived data maynot be so suitable to reproduce the classical logP or log D values. It should be noted that the sameauthors reported identical solvatochromic equa-tions for log P and log kw data derived from aDiscovery RP-Amide column.120

CHROMATOGRAPHICCONDITIONS—MOBILE PHASES

The most widely used organic modifiers forlipophilicity assessment by RP-HPLC or RP-TLC are methanol, acetonitrile or THF. Methanolhas a more water-like structure and is considerednot to disturb the hydrogen-bonding network ofwater. Moreover, during stationary/mobile phaseequilibration in RP-HPLC, methanol moleculesassociate with the stationary phase forming amonolayer, which provides a hydrogen-bondingcapability in better agreement with n-octanol.117

The buffer composition of the aqueous compo-nent in the mobile phase may also play an activerole in retention especially in the case of proto-nated basic compounds, which may form ion pairswith the buffer counter ions. Morpholinepropa-nesulphonic acid (MOPS) is considered as thebuffer of choice for lipophilicity assessment by RP-HPLC.96,97,99 It exhibits a large buffering capacitycoupled to poor ion-pair formation ability dueto its zwitterionic nature and thus it does notinterfere either with solutes or with stationaryphase. The benefit of using MOPS in lipophilicityassessment by RP-TLC has also been reported.However, it should be noted that the migration ofsolutes is not affected considerably by changes inpH and for this reason many authors prefer to usewater instead of buffer in RP-TLC.99,102,103 Thisissue will be further discussed in a followingsection.

The addition of a masking agent is a criticalprerequisite in order to achieve the prevalenceof partition mechanism in RP-HPLC in thecase of neutral and basic dugs, while itseffect is not manifested on RP-TLC.99 Hydro-phobic amines, such as n-decylamine and N,N-dimethyl-octylamine, are considered to be the

DOI 10.1002/jps JOUR

most suitable masking agents to suppress silano-philic interactions, especially when combinedwith methanol as organic modifier.96,117 Theireffect on retention is less evident with acetonitrileas organic modifier. Acetonitrile, as a weakhydrogen-bonding solvent, is not capable of sol-vating the stationary phase with sufficient water,thus presumably preventing the positivelycharged amine to be dragged on the column andthus to exert its role as a masking agent.117

Recently, room temperature ionic liquids ofthe imidazolium tetrafluoroborate family, suchas 1-butyl-3-methylimidazolium (BMIM BF4),have been reported as suitable masking agentscombining the silanol masking effect of theimidazolium cation with the chaotropic characterof the BF4� anion.121–123 Nevertheless, the above-mentioned masking agents cannot be used in thecase of anionic compounds, since they would formion pairing with the negatively charged centre.

In the last few years, the addition of smallamounts of octanol in the methanol fraction of themobile phase has proved favourable for lipo-philicity assessment by HPLC.96–98 It is consid-ered to produce an octanol-like character to thechromatographic system, although its role inretention is not fully investigated. Almost 1:1correlation between log D and log kw is obtained inthe case of neutral and basic drugs with mobilephase containing small amounts of n-octanoland n-decylamine using either an ABZ or aBDS column. Based on these conditions, a rapidautomated method to convert log kw values to logP or log D by a general calibration equation hasbeen developed, known as the E log D approach.96

In the case of acidic drugs, the addition ofn-octanol has also been recognised as a criticalfactor to obtain 1:1 correlation between log P andlog kw values, measured at low pH at whichionisation is suppressed.98 At pH 7.4, a 1:1correlation was reported in the case of weak acidsupon addition of n-octanol in the mobile phase,while for fully ionised compounds the affinitywith the stationary phase, although reduced,remained strong and the relevant equationpossessed a large intercept.124 Examples of log Dversus log kw plots are illustrated in Figures 2and 3. Figure 2 concerns neutral and basic drugsin absence and presence of n-octanol using aBDS column, MOPS as buffer and n-decylamineas masking agent (data taken from Ref. 97).Although correlation coefficient is slightly lowerin presence of n-octanol the corresponding equa-tion has a nonsignificant intercept, representing

NAL OF PHARMACEUTICAL SCIENCES, VOL. 97, NO. 8, AUGUST 2008

Page 9: log p

Figure 2. Log D7.4/log kw relationships for 64 struc-turally diverse basic and neutral drugs in absence(A) and in presence (B) of n-octanol as mobile phaseadditive using a BDS column (data taken from Ref. 97).

2992 GIAGINIS AND TSANTILI-KAKOULIDOU

1:1 correlation. Figure 3 demonstrates the be-haviour of weak and strong acids on a BDS columnin presence of n-octanol, fitting two differenttrendlines.124

EFFECT OF IONISATION ON RETENTION

To adjust retention factors for ionisation, thesame corrections are used as for the distribution

Figure 3. Log D7.4/log kw7.4 relationships for 22 weakly(^) and 21 strongly (~) ionised acidic drugs in presence ofn-octanol as mobile phase additive using a BDS column(reproduced from Ref. 124 by permission).

JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 97, NO. 8, AUGUST 2008

coefficients. However, whether the effects ofionisation in the octanol–water partition systemand RP-HPLC are similar remains to be clarified.The organic modifiers are capable of affectingthe pKa of ionised solutes, as well as the acidity ofthe surface silanol groups and the pH of the mobilephase. In general, the pKa of bases decreases andthe pKa of acids increases as the organic modifierconcentration increases. Substantial structure-dependent differences in pKa shifts at a givenorganic solvent composition and pKa variations atdifferent organic solvent composition have beenreported for RP-HPLC.125–127 These effects areminimised in extrapolation procedure, providinga further argument for the use of log kw valuesinstead of isocratic log k in the case of basiccompounds.

Ionisation correction for RP-TLC RM or RMw

factors is more disputable. Dross et al.100 reportedthat ionisation of basic drugs at pH 7.4 wassuppressed in TLC and RMw values withoutionisation correction correlated better with log P.Analogous results for a series of b-blockersanalogs were attributed to silanophilic interac-tions which compensated the decrease in lipo-philicity of the ionised solutes.101 Malawska102

correlated extrapolated RMw values obtained atpH 7.4 with calculated values of the neutral form,while the same author used water as the aqueouscomponent of the mobile phase for a similar seriesof compounds.103 Investigation of the pH effectfor a set of basic drugs measured at pH 7.4 and11 revealed that, despite differences in migration,a reduced net effect of ionisation may occurred asa result of the formation of gradient conditions onthe TLC plate.99

EFFECT OF CONFORMATION IN RETENTION

The effect of conformation in lipophilicity has beencommented under the relevant section. Analogouseffects, although not necessarily to the sameextent, may be expected in retention. Moreover,topographical factors induced by the anisotropy ofreversed phase stationary phase and the fact thatsolutes approach it gradually with the movementof the mobile phase is an additional reason forconformational effects to be present. In such cases,differences in the partitioning behaviour in theoctanol–water system and the chromatographicsystem may be manifested, thus affecting thequality of the relationships between log D andchromatographic indices and rendering inter-pretations ambiguous.92

DOI 10.1002/jps

Page 10: log p

ALTERNATIVE MEASURES OF LIPOPHILICITY 2993

OTHER CHROMATOGRAPHIC DATA ASLIPOPHILICITY—RELEVANT EXPRESSIONS

The slope S of the linear Eqs. (14) or (15) isconsidered to encode significant information onthe lipophilic behaviour of the solute. By someauthors, the slope S is considered to reflect thesolute/solvent interactions during the retentionprocess and is related to the specific hydrophobicsurface area.128 The strong influence of volume inthe slope S, derived by RP-HPLC and RP-TLC,was demonstrated for a series of substitutedcoumarins using multivariate data analysis—partial least squares (PLS).92 Within a series ofsuch compounds, a good relationship betweenthe slope S and log kw or RMw is anticipated(Eqs. 16 and 17), if there are no considerabledifferences in the forces involved in solute/stationary phase interactions (mainly concerninghydrogen-bonding or the extent of silanophilicinteractions).

S ¼ a log kw þ b (16)

S ¼ a RMw þ b (17)

The organic modifier concentration wo, whichproduces an equal molar distribution between thestationary and mobile phase, leading to log k¼ 0,has been proposed as a measure to rank lipophi-licity. The wo indices correspond to the quotient asexpressed in Eqs. (18) and (19):

’o ¼ log kw=S (18)

’o ¼ RMw=S (19)

Based on the wo concept, a fast gradient RP-HPLC method has been proposed by Valko andcoworkers129–132 to determine the chromato-graphic hydrophobicity index (CHI) as a high-throughput alternative to the other lipophilicitymeasures. For this purpose, gradient retentiontimes (tg) are measured and converted to CHIvalues by means of a calibration equation, derivedby a set of standards with well-determined CHI(wo) values (Eq. 20):

CHI ¼ slope x tg þ intercept (20)

The absolute magnitude of the CHI parameterdepends on the values assigned to the set ofstandards. The method has the advantage that,once the calibration equation has been establish-ed, the retention parameter is obtained from asingle fast gradient run, thus saving time andsolvents. The CHI parameter has been reported

DOI 10.1002/jps JOUR

to correlate satisfactorily with log P. For a series ofsubstituted coumarins however principal compo-nent analysis (PCA) and partial least squaresanalysis (PLS) showed different informationcontent for wo indices, derived from RP-HPLCand RP-TLC, compared to that of the correspond-ing log kw and RMw values.92

Summarizing, reversed phase chromatographyis a friendly and rapid technique, providingindices which correlate directly with log P or log Dvalues. All assumptions however related to thecomplex nature of lipophilicity should be consi-dered also in chromatographic retention, whilesilanophilic interactions are an additional dif-ficulty to be confronted. RP-HPLC is most popularand current research is directed towards theestablishment of standard conditions, so thatextrapolated log kw values can be converted tolog P or log D by a general calibration equation.The CHI concept on the other hand is valuable forrapid screening of compound libraries, rankingthe compounds according to their lipophilicity, butdoes not provide knowledge on lipophilicity valuesper se.

PARTITIONING INTO LIPOSOMES

Partitioning into liposome–water system offers analternative attracting considerable interest, sincethey are considered to represent direct modelsfor biological membranes. Liposomes constituteanisotropic media, predominantly composed ofamhiphilic molecules, like phospholipids, whichform spherical closed structures of curved bilayersthat entrap part of the surrounding solvent intheir interior. The number of bilayer sheetsvaries from one to several hundreds and smallunilamellar (SUV), large unilamellar (LUV) andlarge multilamellar (LMV) vesicles.133–135 Phos-phatidylcholines are most frequently used toobtain standardised and easily reproducible lipo-some systems.136–138 The physical state of lipo-somes is temperature dependent, being in gelstate at low temperatures and as crystal liquids athigher temperature. Transition temperature islow for liposomes prepared from lipids of naturalsources and relatively high when synthetic lipidshave been used.139

From a thermodynamic point of view, transferinto liposomes has been found to be entropy drivenbelow the gel/liquid transition temperature.54

Above the transition temperature, studies, carriedout using highly sensitive micro-calorimetry,

NAL OF PHARMACEUTICAL SCIENCES, VOL. 97, NO. 8, AUGUST 2008

Page 11: log p

2994 GIAGINIS AND TSANTILI-KAKOULIDOU

showed that the process is predominatelyenthalpy driven, in contrast to the well establi-shed ideas of entropy-driven partitioning ofdrugs.140–142

The polar headgroups of phospholipids are ofparticular importance determining the electricalproperties of the membrane surface. Their con-formation can be affected by charged solutes thatpartition into the lipid bilayer. Conformationchanges are expected to be different in presenceof positive or negative charge. Binding of acationic solute to the phosphate anion causes arepulsion of the positively charged quaternarynitrogen towards the water phase, ‘opening’the polar region and permitting the penetrationof the solute into the hydrophobic core.141 Neutralmolecules seem to have no effect on polar head-group conformation.143

Another important factor, determining theelectrostatic interactions between a solute and amembrane, is the surface charge. The surfacecharge depends on the lipids present in the outerlayer and may be changed by pH. It is measuredby the zeta-potential, which is defined as thesurface charge at a distance of 2 A

´from the

membrane surface.144 Phosphatidylcholine lipo-somes, with a positive and negative group andthereupon a zero net charge in their headgroup,have neutral zeta-potential in a pH range 2–11.The binding of charged solutes to the membranealters the surface charge.145

Phospholipids possess a dipole field that isdetermined by the polar headgroups, the surfacewater molecules and the lipid carbonyls, espe-cially the ester group carbonyls, which linkthe fatty acids to the glycerol backbone. Thedipole field is positive inside the liposome andnegative outside, denoting that the C––O dipole isoriented towards the lipid–water interface.143

The above described issue imply that partition-ing into liposomes, a process known also asphospholipophilicity, involves electrostatic inter-actions as additional intermolecular recognitionforces, not encoded in traditional lipophilicityparameters and could be factorised as expressedby Eq. (21):

Phospholipophilicity

¼ Hydrophobicity � Polarity

þ Ionic bonds (21)

As a result, charged species seem to partitioninto membranes considerably more strongly than

JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 97, NO. 8, AUGUST 2008

they do into octanol. Thus, in the case of ionisabledrugs, a lower difference is observed betweenthe partitioning of the neutral and the ionisedform.146,147 In this aspect, phospholipophilicitymay be considered to represent a border casebetween partitioning and binding.

Determination of liposome water distributioncoefficients requires that the sample is equili-brated with a suspension of liposomes, followedby a separation procedure and quantitation ofthe sample in the lipid free phase. For theseparation of liposomes from the aqueous compo-nent ultrafiltration, ultrafiltration/centrifugationor equilibrium dialysis are the commonly usedmethods.148–150 Potentiometry, a pH-metric pro-cedure analogous to that applied for solvent/waterdistribution coefficients does not involve a separa-tion step.151,152 H-NMR spectroscopy is anotheralternative to study solute/membrane interac-tions which does not require separation ofphases.153–155 Line-width broadening upon inter-action of solutes with liposomes is a linearfunction of phospholipids concentration, with aslope (called NMR slope), which can used toquantify the degree of this interaction.156

Working with liposomes is associated withdifficulties and considerable care is required,compared to octanol. The preparation of liposomesshould be performed under inert atmosphere andreduced temperatures. They have limited stabilityand liposome/water partitioning studies are moresusceptible to experimental conditions, whileNernst conditions are not always achievable.157

These difficulties limit the use and the wide scaledevelopment of liposomes as an alternative tooctanol in partitioning studies.

IMMOBILISED ARTIFICIALMEMBRANE CHROMATOGRAPHY

The criticism towards octanol as an isotropicmedium with only a superficial similarity tobiomembranes and the difficulties associatedwith the use of liposomes as more representativemodels, have triggered the development ofIAM stationary phases for use in HPLC. IAMchromatography offers a promising alternativeto simulate liposome/water partitioning andcell membrane permeation and has unfoldednew perspectives in the application of HPLCas a tool to mimic specific interactions withphospholipids.158–160

DOI 10.1002/jps

Page 12: log p

ALTERNATIVE MEASURES OF LIPOPHILICITY 2995

IAM stationary are prepared from phos-pholipids covalently bonded to a propylamino-silica support material at monolayer densities.Remaining propylamine residues are treatedin a second step to suppress undesired basicfunction on the silica backbone. Moreover, freesilanol groups, although not easily accessible,may interfere in secondary interactions. Themost frequently used IAM column is IAMPC,which contains phosphatidylcholine. In fact, threedifferent types of IAMPC have been introduced inthe market, the single chain IAMPC-DD, thedouble chain IAMPC-MG and IAMPC-DD2, whichdiffer on the way the remaining propylamineresidues are treated. It is reported that doublechain IAM surfaces simulate better naturalphospholipids and thus the resulting chromato-graphic indices correlate better with permeabilitydata.160–162

IAM columns permit the use of aqueousmobile phases without addition of organic modi-fier, leading to directly measured log kw valuesand reducing considerably the time of analysis.The buffer of choice is phosphate buffered saline(PBS) in order to mimic physiological conditions.The pH limitations of the column restrict mea-surement in the pH range 2.5–7.4. Many authorsprefer the use of pH 7.0, which is close tophysiological pH and safer for the column.160,162

In the case of compounds with strong affinityfor the IAM surface, acetonitrile up to 30% ispreferably added and log kw values are obtainedby linear extrapolation. The use of methanol asorganic modifier is avoided, since it affects thestability of the column, leading to methanolysis ofthe phospholipids. Nevertheless, the ageing of thecolumn should be checked from time to time usingstandard compounds.163–165

The structural characteristics of IAM surfaceshave been investigated and compared to those ofliposomes. The synthetic phospholipids linked tosilica-propylamine skeleton are fully saturatedand usually contain acyl chains with 16 carbonatoms (propylamine included), while egg phos-phatidylcholine, the most widely used phospholi-pid for liposome preparation, contains longer acylchains and cis double bonds.133,166 Liposomesconstitute a fluid phase and each vesicle is formedby a phospholipids bilayer, which separatesthe external from the inner aqueous phase. Incontrast, IAM are solid-phase system, where aphospholipid monolayer is covalently linked to thesilica backbone. As a result, the hydrophobic coreis half large in IAM surface than in liposomes and

DOI 10.1002/jps JOUR

phospholipids are more ordered and deprivedfrom molecular dynamics. The most importantdifference concerns the density of the polarheadgroups, which in IAM surface is smallerthan in liposomes. However, the polar interfacialregion in IAM surface mimics well that of fluidmembranes. Thus, in both systems the cholinegroups exhibit larger motional fluctuations thanphosphate groups and the distribution of waternear the polar headgroups is the same.167

According to Ong and Pidgeon, solute partition-ing seems to be the principal retention mechanismin IAM retention, with similar thermodynamics onboth the single-chain and double-chain IAMPCsurfaces. For a set of phenol derivatives, thepartitioning into IAMPC surfaces was found tobe both enthalpy and entropy driven, while forbeta-blockers, it was entropy driven.168 Otherauthors report negative DS values for two acidicdrugs (warfarin and salicylic acid) and positive DSvalues for three basic drugs (lidocaine, propranololand diazepam), whereas negative DH valueswere calculated for all five drugs.169 For neutralcompounds the intermolecular forces resemblethose underlying partitioning in octanol/waterand retention in reversed phase HPLC. Hence,besides the hydrophobic/solvophobic interactions,polar interactions, mainly expressed as H-bondacceptor basicity, are the predominant factorsin IAM retention. According to the results ofsolvatochromic analysis however, the hydrophobicterm seems to have a smaller positive contributionin IAM retention compared to its contribution inoctanol/water partitioning.170 Moreover, it is con-sidered that the IAM surface provides a hydro-phobic environment that resembles a RP-C3 HPLCcolumn, whereas for lipophilicity assessment RP-C-18 stationary phase are used.171 For ionisedspecies, polar interactions are electrostatic innature.168 Therefore, protonated basic compoundsare stronger retained as a result of their interactionwith the phosphate anions of the stationary phase.It is reported that IAM retention of protonated b-blockers was found to be stronger, compared toisolipophilic neutral compounds.172 In a studyconcerning structurally diverse basic and neutralcompounds, the degree of protonation had to beconsidered as an additional parameter in order toobtain a good correlation between log kwIAM andlog D values at pH 7.4.173 Otherwise, a bettercorrelation was obtained with log P values, imply-ing that the decrease in the retention due toionisation was partly compensated by the electro-static interactions. These results are depicted in

NAL OF PHARMACEUTICAL SCIENCES, VOL. 97, NO. 8, AUGUST 2008

Page 13: log p

2996 GIAGINIS AND TSANTILI-KAKOULIDOU

Figure 4. In the same study, IAM retention wascompared to reversed phase chromatographicretention. Very characteristically, the strong basemetformin, fully protonated at pH 7.4, eluted withthe dead time in reversed phase HPLC, while itwas retained in IAM chromatography due to theelectrostatic interactions of its positively chargedcentre with the phosphate anions.173 In this aspect,the question may arise whether IAM indices aresuitable to predict passive diffusion of drugs, whichaccording to the well known pH-partition hypo-thesis implies reduced absorption of chargedspecies, or they should rather be used to simulatedrug–membrane interactions. As already com-mented about liposomes, IAM chromatographymay also be considered as a border case betweenpartitioning and binding. Electrostatic forces,however, have been reported to be weaker inIAM chromatography than in liposome partition-ing and the reason may be related to the smallerdensity of the polar headgroups in IAM surfaces.174

The contribution of hydrogen bond seems also to beless important for the affinity to IAM stationaryphase than to liposomes.175

Despite these differences, IAM chromatographicindices have successfully been correlated withliposomes partitioning data; however, such studies

Figure 4. Relationship of (A) log D7.4 (B) log P versuslog kwIAM measured at pH 7.4, for neutral (&) and basic(&) drugs.

JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 97, NO. 8, AUGUST 2008

include a rather limited number of compounds.169

In the case of basic drugs, silanophilic interactionshave been reported to affect the log kwIAM/pHprofile as compared to the corresponding pH/partition diagram in liposomes.169,175

Less investigated is the effect of conformationin IAM retention. In a study of substitutedcoumarins, such effects, induced by the extendedplanar conformation of some large derivatives,were found to have a similar impact in IAMretention and reversed-phase HPLC and TLC,while they did not affect the n-octanol–waterpartitioning.176 Therefore, they should ratherbe attributed to the particular features of thechromatographic procedure, according to which incontrast to solvent/water partitioning, the solutesapproach the stationary phase with the movementof the mobile phase.

The potential of IAM chromatographic indicesto predict passive transport through variousbiological barriers, as well as to estimate pharma-cokinetic properties and certain pharmacologicalactivities has recently been reviewed by Bar-bato.177 IAM chromatographic indices have beensuccessfully correlated with Caco 2 cells perme-ability and intestinal absorption in rat. In thelatter case, molecular weight should be introducedas an additional parameter to improve thecorrelation.178–180 Other pharmacokinetic para-meters (protein binding, partition to erythrocytes,apparent volume of distribution), as well aspenetration across the blood brain barrier havealso been satisfactorily correlated with IAMindices.159,181,182 Most studies, however, concernsmall data sets of congeneric compounds. Never-theless, the belief that IAM chromatographyshould be considered to be always a better choicefor modelling biological processes is disputed in arecent publication on the similarity between IAMcolumns, conventional HPLC columns, octanol–water partitioning and biopartitioning systems bymeans of solvatochromic analysis.183 To thispoint, it should be noted that the last years othertypes of biomimetic stationary phases, like humanserum albumin (HSA) or a-acid glycoproteincontaining surfaces, are available, creating newperspectives on biochromatography as an alter-native rapid technique to investigate biologicalprocesses.184

SUMMARY AND CONCLUSIONS

Figure 5 provides a rough summary of the mostimportant information discussed in the present

DOI 10.1002/jps

Page 14: log p

Figure 5. Schematic representation of the basic features of lipophilicity andphospholipophilicity, the related indices and the impact of chromatographic techniquesin lipophilicity/phospholipophilicity assessment.

ALTERNATIVE MEASURES OF LIPOPHILICITY 2997

review. Far for being complete, Figure 5 focuses onthe distinction between lipophilicity and phos-pholipophilicity, as well as on the impact andcommon characteristics of chromatographic tech-niques in the assessment of both properties. Itsummarises the difficulties in direct partitioningexperiments, (which in the case of liposomesare the major obstacle), and the advantages ofreversed phase chromatographic indices with awarning for potential silanophilic interactions.Question marks refer to the parallelism betweenliposome partitioning and IAM retention, denot-ing that further investigations are necessary.

In conclusion, despite criticisms, for more thanforty years n-octanol–water partitioning system isused as a reference system to assess lipophilicity.During these decades, large progress has beenachieved concerning prediction of log P and log D,as well as the understanding of the particularnature of log P as a prerequisite to appreciate thebehaviour of solutes in a lipophilic environment.On the other hand, based on the similarity ofreversed phase retention with n-octanol–waterpartitioning, chromatographic techniques havebeen developed and conditions refined to produce

DOI 10.1002/jps JOUR

lipophilicity indices analogous to log P or log D.The turn towards liposomes as more representa-tive models for membranes triggered the devel-opment of IAM stationary phases to overcome thedifficulties associated with liposome partitioningand to guarantee rapid measurements. Althoughthe advantages of IAM chromatographic indices tomimic membrane interactions and/or permea-bility are still under investigation, biochromato-graphy is attracting an increasing interestoffering new perspectives in the research ofbiological processes.

ACKNOWLEDGMENTS

The authors would like to thank CompuDrugInternational, Inc. and Dr. Ferenc Darvas whokindly offered Pallas software v.3.3.2.4 for log Dcalculations.

REFERENCES

1. van de Waterbeemd H, Smith DA, BeaumontK, Walker DK. 2001. Property-based design:

NAL OF PHARMACEUTICAL SCIENCES, VOL. 97, NO. 8, AUGUST 2008

Page 15: log p

2998 GIAGINIS AND TSANTILI-KAKOULIDOU

Optimization of drug absorption and pharmacoki-netics. J Med Chem 44:1313–1333.

2. Oprea T. 2002. Current trends in lead discovery:Are we looking for the appropriate properties?J Comput Aided Mol Des 16:325–334.

3. van de Waterbeemd H, Testa B. 1987. The para-metrization of lipophilicity and other structuralproperties in drug design. In: Testa B, editor.Advances in drug research. Vol. 16 New York:Academic Press. pp 85–227.

4. Hansch C, Leo A, Hoekman D. 1995. Fundamen-tals and applications in chemistry and biology inexploring QSAR. Washington, DC: American Che-mical Society.

5. Meyer H. 1899. Theorie der Alkoholnarkose. ArchExp Pathol Pharmakol 42:109–118.

6. Overton E. 1901. Studien uber die Narkose. Jena:Fischer.

7. Gaudette LE, Brodie BB. 1959. Relationshipbetween lipid solubility of drugs and their oxida-tion by liver microsomes. Biochem Pharmacol 2:89–96.

8. Hansch C, Fujita T. 1964. r-s-p analysis. A me-thod for the correlation of biological activity andchemical structure. J Am Chem Soc 86:1616–1626.

9. Leo A, Hansch C, Elkins D. 1971. Partition coef-ficients and their uses. Chem Rev 71:525–616.

10. Sangster J. 1994. Log KOW Databank. Montreal,Quebec, Canada: Sangster Research Laboratories.

11. Howard PH, Meylan W. 2000. PHYSPROP Data-base. Syracuse: Syracuse Research Corp.

12. Physicochemical Parameter Database, MedicinalChemistry Project, Pomona College, Claremont,CA, USA. www.biobyte.com/bb/prod/cqsar.html.

13. Hansch C. 1993. Quantitative structure-activity-relationships and the unnamed science. Acc ChemRes 26:147–153.

14. Lien E. 1975. Drug design. Vol. 5, In: Ariens EJ,ed, New York City: Academic Press, pp 81–132.

15. Hansch C. 1984–1985. The QSAR paradigm in thedesign of less toxic molecules. Drug Metab Rev 15:1279–1294.

16. Hansch C, Bjorkroth JP, Leo A. 1987. Hydropho-bicity and central nervous system agents: Onthe principle of minimal hydrophobicity in drugdesign. J Pharm Sci 76:663–687.

17. Lipinski CA, Lombardo F, Dominy BW, Feeney PJ.1997. Experimental and computational appro-aches to estimate solubility and permeability indrug discovery and development settings. AdvDrug Deliv Rev 23:3–25.

18. Congreve M, Carr R, Murray C, Jhoti H. 2003.A ‘rule of three’ for fragment-based lead discovery?Drug Discov Today 8:876–877.

19. Smith DA, Jones B, Walker DK. 1996. Design ofdrugs involving the concepts and theories of drugmetabolism and pharmacokinetics. Med Res Rev16:243–266.

JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 97, NO. 8, AUGUST 2008

20. Lipophilicity in Drug Research and Toxicology,University of Lausanne, 21–24 March 1995.

21. The Second LogP Symposium: Lipophilicity inDrug Disposition, University of Lausanne, 5–9March 2000.

22. LogP2004 Symposium—Physicochemical andBiological Profiling in Drug Research, ETHZurich (Swiss Federal Institute of Technology),29 February–4 March 2004.

23. Kramer S. 2001. Liposomes/water partitioning:Theory techniques and applications. In: Testa B,van de Waterbeemd H, Folkers G, Guy R, editors.Pharmacokinetic optimization in drug research,Verlag Helvetica Chimica Acta. Weinheim: Zurichand Wiley—VCH.

24. Pidgeon C, Ong S, Liu H, Qui X, Pidgeon M,Dantzig AH, Munroe J, Hornback WJ, KasherJS, Glunz L, Szczerba T. 1995. IAM chromato-graphy: An in vitro screen for predicting drugmembrane permeability. J Med Chem 38:590–594.

25. Ong S, Liu H, Pidgeon C. 1996. Immobilized arti-ficial membrane chromatography: Measurementsof membrane partition coefficient and predictingdrug membrane permeability. J Chromatogr A728:113–128.

26. Testa B, Carrupt P-A, Gaillard P, Tsai R-S. 1996.Intramolecular Interactions encoded in Lipophili-city: Their nature and significance. In: Pliska V,Testa B, van de Waterbeemd H, editors. Lipophi-licity in drug action and toxicology. Weinheim:VCH. pp 49–71.

27. El Tayar N, Testa B, Carrupt P-A. 1992. Polarintermolecular interactions encoded in partitioncoefficients: An indirect estimation of hydrogen-bond parameters of polyfunctional solutes. J PhysChem 96:1455–1459.

28. Raevsky OA, Schaper K-J, Seydel JK. 1995. H-Bond contribution to octanol-water partition coef-ficients of polar compounds. Quant Struct ActRelat 14:433–436.

29. Tsantili-Kakoulidou A, Varvaresou A, Siatra-Papastaikoudi TH, Raevsky O. 1999. A compre-hensive investigation of the partitioning andhydrogen bonding behavior of Indole containingderivatives of 1,3,4-thiadiazole and 1,2,4-triazoleby means of experimental and calculative ap-proaches. Quant Struct Act Relat 18:482–489.

30. Yang GZ, Lien EJ, Guo R. 1986. Physicalfactors contributing to hydrophobic constant.Quant Struct Act Relat 5:12–18.

31. Bradshaw J, Taylor PJ. 1989. Rationalizationsamong heterocyclic partition coefficients. Part 3:p-excessive heterocycles. Some Comments on theClogP Algorithm. Quant Struct Act Relat 8:279–287.

32. Gago F, Pastor M, Perez-Budragueno J, Lopez R,Avarez-Builla J, Elguero J. 1994. Hydrophobicityof heterocycles: Determination of the p values of

DOI 10.1002/jps

Page 16: log p

ALTERNATIVE MEASURES OF LIPOPHILICITY 2999

substituents on N-phenylpyrazoles. Quant StructAct Relat 13:165–171.

33. Vrakas D, Tsantili-Kakoulidou A, Hadjipavlou-Litina D. 2003. Exploring the consistency of logPestimation for substituted coumarins. QuantStruct Act Relat 22:622–629.

34. Gaillard P, Carrupt P-A, Testa B. 1994. The con-formation-dependent lipophilicity of morphineglucuronides as calculated from the molecularlipophilicity potential. Bioorg Med Chem Lett 4:737–742.

35. Caron G, Ermondi G. 2006. New insights into thelipophilicity of ionized species. In: Testa B, KramerSD, Wunderli-Allenspach H, Folkers G, editors.Pharmacokinetic profiling in drug research. Wiley,VCH. Zurich: pp 165–185.

36. Dellis D, Giaginis C, Tsantili-Kakoulidou A.2007. Physicochemical profile of Nimesulide. Ex-ploring the interplay of lipophilicity, solubilityand ionization. J Pharm Biomed Anal 44:57–62.

37. Kamlet MJ, Doherty RM, Abraham MH, MarcusY, Taft RW. 1988. Linear solvation energy rela-tionship. 46. An improved equation for correlationand prediction of octanol/water partition coef-ficients of organic nonelectrolytes (includingstrong hydrogen bond donor solutes). J Phys Chem92: 5244–5255.

38. Platts J, Abraham MH, Butina D, Hersey A.J Pharm Sci. 1999. 88:670–679.

39. Abraham MH, Chadha HS, Whiting GS, MitchellRC. 1994. Hydrogen bonding. An analysis ofwater-octanol and water-alkane partitioning andthe delta log P parameter of Seiler. J Pharm Sci83:1085–1100.

40. Abraham MH, Grellier PL, Prior DV, Morris JJ,Taylor PJ, Doherty RM. 1990. Hydrogen bonding.11. A quantitative evaluation of the hydrogen-bond acidity of imides as solutes. J Org Chem55:2227–2229.

41. Abraham MH, McGowan JC. 1987. The use ofcharacteristic volumes to measure cavity termsin reversed phase liquid chromatography. Chro-matographia 23:243–246.

42. Ouvrard C, Lucon M, Graton J, Berthelot M,Laurence C. 2004. Determination of the hydro-gen-bond basicity of weak and multifunctionalbases: The case of lindane (g-hexachlorocy-clohexane). J Phys Org Chem 17:56–64.

43. Berthelot M, Graton J, Ouvrard C, Laurence C.2002. Hydrogen-bond basicity of solutes in hydro-xylic solvents from octanol-water partition coef-ficients. J Phys Org Chem 15:218–228.

44. Avdeef A, Barrett DA, Shaw PN, Knaggs RD,Davis SS. 1996. Octanol-, chloroform-, andPGDP-water partitioning of morphine-6-glucuro-nide and other related opiates. J Med Chem 39:4377–4381.

DOI 10.1002/jps JOUR

45. Leahy DE, Taylor PJ, Wait AR. 1989. Model sol-vent systems for QSAR. 1. Propylene glycol dipe-largonate (PGDP). A new standard for use inpartition coefficient determination. Quant StructAct Relat 8:17–31.

46. Taylor PJ. 1990. Hydrophobic properties ofdrugs. In: Hansch C, Sammes PG, Taylor JB,editors. Comprehensive medicinal chemistry.Vol. 4 Oxford: Pergamon. pp 241–294.

47. Bouchard G, Pagliara A, Carrupt P-A, Testa B,Gobry V, Girault HH. 2002. Theoretical andexperimental exploration of the lipophilicity ofzwitterionic drugs in the 1,2-dichloroethane/watersystem. Pharm Res 19:1150–1159.

48. Murthy KS, Zografi G. 1970. Oil-water partition-ing of chlorpromazine and other phenothiazinederivatives using dodecane and n-octanol.J Pharm Sci 59:1281–1285.

49. Tsai R-S, Fan W, El Tayar N, Carrupt P-A, TestaB, Kier LB. 1993. Solute-water interactions in theorganic phase of a biphasic system. 1. Structuralinfluence of organic solutes on the ‘water-dragging’effect. J Am Chem Soc 115:9632–9639.

50. Fan W, Tsai RS, El Tayar N, Carrupt P-A, Testa B.1994. Solute-water interactions in the organicphase of a biphasic system. 2. Effects of organicphase and temperature on the ‘water-dragging’effect. J Phys Chem 98:329–333.

51. Iwahashi M, Hayashi Y, Hachiya N, MatsuzawaH, Kobayashi H. 1993. Self-association of octan-1-ol in the pure liquid state and in decane solutionsas observed by viscosity, self-diffusion, nuclearmagnetic resonance and near-infrared spectro-scopy measurements. J Chem Soc 89:707–712.

52. Franks NP, Abraham MH, Lieb WR. 1993. Mole-cular organization of liquid n-octanol: An X-raydiffraction analysis. J Pharm Sci 82:466–470.

53. Avdeef A. 2001. Physicochemical profiling (solubi-lity, permeability and charge state). Curr Top MedChem 1:277–351.

54. Davis SS, James MJ, Anderson NH. 1986. Thedistribution of substituted phenols in lipid vesi-cles. J Chem Soc 81:313–327.

55. Hasselbalch KA. 1916. Die Berechnung derWasserstoffzahl des Blutes aus der freien undgebunden Kohlensaure desselben, und die Sauer-stoffbindung des Blutes als Funktion der Wasser-stoffzahl. Die Biochem Z 78:112–144.

56. Avdeef A. 1996. Assessment of distribution—pHprofiles. In: Pliska V, Testa B, van de WaterbeemdH, editors. Methods and principles in medicinalchemistry. Vol. 4 Weinheim, Germany: VCH Pub-lishers. pp 109–139.

57. Scherrer RA. 1984. The treatment of ionizablecompounds in quantitative structure-activitystudies with special consideration of ion partition-ing. In: Magee PS, Kohn GK, Menn JJ, editors.Pesticide syntheses trough rational approaches.

NAL OF PHARMACEUTICAL SCIENCES, VOL. 97, NO. 8, AUGUST 2008

Page 17: log p

3000 GIAGINIS AND TSANTILI-KAKOULIDOU

ACS Symposium Series 225. Washington, DC:American Chemical Society. pp 225–246.

58. Scherrer RA, Crooks SL. 1989. Titrations in water-saturated octanol: A guide to partition coefficientsof ion pairs and receptor-site interactions. QuantStruct Act Relat 8:59–62.

59. Pagliara A, Carrupt P-A, Caron G, Gaillard P,Testa B. 1997. Lipophilicity profiles of ampholytes.Chem Rev 97:3385–3400.

60. Takacs-Novak K, Jozan M, Szas G. 1995. Lipophi-licity of amphoteric molecules expressed by thetrue partition coefficient. Int J Pharm 113:47–55.

61. Pagliara A, Testa B, Carrupt PA, Jolliet P, MorinC, Morin D, Urien S, Tillement JP, Rihoux JP.1998. Molecular properties and pharmacokineticbehavior of cetirizine, a zwitterionic H1-receptorantagonist. J Med Chem 41:853–863.

62. Rekker RF, Mannhold R. 1992. Calculation of druglipophilicity. Weinheim: VCH.

63. Leo A. 1993. Calculating logPoct from structures.Chem Rev 93:1281–1306.

64. Ghose AK, Crippen GM. 1986. Atomic physico-chemical parameters for three-dimensional struc-ture-directed quantitative structure-activityrelationships I. Partition coefficients as a measureof hydrophobicity. J Comput Chem 7:565–577.

65. Broto P, Moreau G, Vandycke C. 1984. Molecularstructures, perception, autocorrelation descriptorand SAR studies; System of atomic contribution forthe calculation of the octanol-water partition coef-ficient. Eur J Med Chem 19:71–78.

66. Suzuki T, Kudo Y. 1990. Automatic logP estima-tion based on combined additive modeling meth-ods. J Comp Aided Mol Des 4:155–198.

67. Meylan WM, Howard PH. 1995. Atom/fragmentcontribution method for estimating octanol-waterpartition coefficients. J Pharm Sci 84:83–92.

68. Partham M, Hall L, Kier M. 2000. Accurate pre-diction of logP using E-state indices with NeuralNetwork Analysis. Washington DC: AmericanChemical Society Meeting.

69. Tetko IV, Tancuk VY. 2002. Application of asso-ciative neural networks for prediction of lipophi-licity in ALOGPS 2.1. Program. J Chem ComputSci 42:1136–1145.

70. Papp A, Molnar L, Keseru GM, Gulyas Z, DarvasF. 2004. Prediction of logP using neural network incombination with atomic fragmental descriptors,The 15th European Symposium on QuantitativeStructure-Activity Relationships and MolecularModeling, Book of Abstracts B. 106, p. 338.

71. Testa B, Carrupt C-A, Gaillard P, Billois F, WeberP. 1996. Lipophilicity in molecular modeling.Pharm Res 13:335–343.

72. Csizmadia F, Tsantili-Kakoulidou A, Panderi I,Darvas F. 1997. Prediction of distribution co-efficient from structure: 1. Estimation method.J Pharm Sci 86:865–871.

JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 97, NO. 8, AUGUST 2008

73. Mannhold R. 2006. Calculation of lipophilicity: Aclassification of methods. In: Testa B, Kramer SD,Wunderli-Allenspach H, Folkers G, editors.Pharmacokinetic profiling in drug research.Zurich: Wiley, VCH. pp 333–352.

74. Dearden JC, Bresnen GM. 1988. The measure-ment of partition coefficients. Quant Struct ActRelat 7:133–144.

75. Hersey A, Hill AP, Hyde RM, Livingstone DJ.1989. Principles of method selection in partitionstudies. Quant Struct Act Relat 8:288–296.

76. Sangster J. 1997. Octanol-water partition coeffi-cients. New York: John Wiley & Sons, Inc.

77. Avdeef A. 1992. pH-Metric logP. 1. Difference plotsfor determining ion-pair octanol-water partitioncoefficients of multiprotic substances. QuantStruct Act Relat 11:510–517.

78. Avdeef A. 1996. Assessment of distribution—pHprofiles. In: Pliska V, Testa B, van de WaterbeemdH, editors. Methods and principles in medicinalchemistry. Vol. 4 Weinheim, Germany: VCH Pub-lishers. pp 109–139.

79. Franke U, Munk A, Wiese M. 1999. Ionizationconstants and distribution coefficients of phe-nothiazines and calcium channel antagonistsdetermined by a pH—Metric method and correla-tion with calculated partition coefficients. J PharmSci 88:89–95.

80. Tsai R-S, Lisa G, Carrupt P-A, Testa B. 1996.Centrifugal partition chromatography for lipophi-licity measurements. In: Pliska V, Testa B, vande Waterbeemd H, editors. Lipophilicity in drugaction and toxicology. Weinheim, Germany: VCHPublishers. pp 109–139.

81. Hitzel L, Watt AP, Locker KL. 2000. An increasedthroughput method for the determination of parti-tion coefficients. Pharm Res 17:1389–1395.

82. Braumann T. 1986. Determination of hydrophobicparameters by reversed-phase liquid chromato-graphy: Theory, experimental techniques, andapplication in studies on quantitative structure-activity relationships. J Chromatogr 373:191–225.

83. Kaliszan R. 1990. High-performance liquid-chro-matographic methods and procedures of hydro-phobicity determination. Quant Struct Act Relat9:83–87.

84. Dorsey JG, Khaledi MG. 1993. Hydrophobicityestimations by reversed-phase liquid chromato-graphy. Implications for biological partitioningprocesses. J Chromatogr A 656:485–499.

85. Van de Waterbeemd H, Kansy M, Wagner B,Fischer H. 1996. Lipophilicity measurement byhigh performance liquid chromatography (RP-HPLC). In: Pilska V, Testa B, van de WaterbeemdH, editors. Lipophilicity in drug action and toxi-cology. Weinheim, Germany: VCH Publishers.pp 73–87.

DOI 10.1002/jps

Page 18: log p

ALTERNATIVE MEASURES OF LIPOPHILICITY 3001

86. Biagi GL, Barbaro MC, Sapone A, Recanatini M.1994. Determination of lipophilicity by means ofreversed-phase thin layer chromatography. 1.Basic aspects and relationships between slopeand intercept of TLC equations. J Chromatogr A662:341–361.

87. Dross KP, Rekker RF, de Vries G, Mannhold R.1998. The lipophilic behavior of organic compounds:3. The search for interconnections betweenreversed-phase chromatographic data and logPoct

values. Quant Struct Act Relat 17:549–557.88. Tsantili-Kakoulidou A, Antoniadou-Vyza A. 1988.

Hydrophobicity studies on diadamantyl deriva-tives. Comparison between calculated partitioncoefficients and experimental lipophilicity indicesdetermined by reversed phase high-performanceliquid chromatography and thin layer chromato-graphy. J Chromatogr 445:317–326.

89. Abraham MH, Chandha HS, Leo A. 1994. Hydro-gen-bonding. 35. Relationship between high-performance liquid-chromatography capacity fac-tors and water-octanol partition-coefficients.J Chromatogr A 685:203–211.

90. Nasal A, Siluk D, Kaliszan R. 2003. Chromato-graphic retention parameters in medicinalchemistry and molecular pharmacology. CurrMed Chem 5:381–426.

91. Pagliara A, Khamis E, Trinh A, Carrupt PA, TsaiRS, Testa B. 1995. Structural properties governingretention mechanisms on RP-HPLC stationaryphases used for lipophilicity measurements.J Liq Chromatogr 18:1721–1745.

92. Vrakas D, Panderi I, Hadjipavlou-Litina D, Tsantili-Kakoulidou A. 2005. Investigation of the relation-ships between logP and various chromatographicindices for a series of substituted coumarins.Evaluation of their similarity/dissimilarity usingmultivariate statistics. Quant Struct Act Relat24:254–269.

93. OECD Guideline for testing of Chemicals 107,Partition Coefficient (n-octanol/water)-FlaskShaking Method, OECD, Paris 1981.

94. Yamagami C, Yokota M, Takao N. 1994. Hydro-phobicity parameters determined by reversed-phase liquid chromatography. IX: Relationshipbetween capacity factor and water-octanol parti-tion coefficient of monosubstituted pyrimidines.Chem Pharm Bull 42:907–912.

95. Lodge K. 1999. Octanol-water partition coef-ficients of cyclic C-7 hydrocarbons and selectedderivatives. J Chem Eng Data 44:1321–1324.

96. Lombardo F, Shalaeva MY, Tupper KA. Gao F2001. ElogDoct: A tool for lipophilicity determina-tion in drug discovery. 2. Basic and neutral com-pounds. J Med Chem 44:2490–2497.

97. Giaginis C, Theocharis S, Tsantili-Kakoulidou A.2006. Contribution to the standardization of thechromatographic conditions for the lipophilicity

DOI 10.1002/jps JOUR

assessment of neutral and basic drugs. Anal ChimActa 573–574:311–318.

98. Liu X, Tanaka H, Yamauchi A, Testa B, ChumanH. 2005. Determination of lipophilicity byreversed-phase high-performance liquid chroma-tography. Influence of 1-octanol in the mobilephase. J Chromatogr A 1091:51–59.

99. Giaginis C, Dellis D, Tsantili-Kakoulidou A. 2006.Effect of the mobile phase aqueous componenton RP-TLC retention and its inference on lipophi-licity determination for a series of structurallydiverse drugs. JPC-J Planar Chromatogr 19:151–156.

100. Dross K, Mannhold R, Rekker RF. 1992. Druglipophilicity in QSAR practice. II. Aspects ofRM-determinations and their interrelations withpartition coefficients. Quant Struct Act Relat 11:36–44.

101. Panderi I, Angelopoulos CH, Tsantili-KakoulidouA, Kazanis M, Vavayiannis A. 1997. Lipophilicitystudies on b-blocker benzoxazinone derivatives.II Farmaco 52:167–172.

102. Malawska B. 1998. Determination of the lipo-philicity of some N-substituted amides ofa-piperazine-g-hydroxybutyric acid. JPC J PlanarChromatogr 11:137–140.

103. Malawska B, Kulig K, Wisniewska M. 2000. Deter-mination of the lipophilicity of antiarrhythmic andantihypertensive 1-[2-hydroxy- or 1-[2-acetoxy-3-(4-aryl-1-piperazinyl)propyl]-pyrrolidin-2-onederivatives. JPC J Planar Chromatogr 13:187–190.

104. El Tayar N, Tsantili-Kakoulidou A, RoethlisbergerT, Testa B, Gal J. 1988. Different partitioningbehaviour of sulphonyl-containing compoundsin reversed-phase high-performance liquidchromatography and octanol-water systems.J Chromatogr 439:237–244.

105. Neue UD, Tran K, Iraneta PC, Alden BA. 2003.Characterization of HPLC packings. J Sep Sci 26:174–186.

106. Vervoor RJ, Debets AJ, Claessens HA, CramersCA, de Jong GJ. 2000. Optimization and charac-terization of silica-based reversed-phase liquidchromatographic systems for the analysis ofbasic pharmaceuticals. J Chromatogr A 897:1–22.

107. Cserhati T. 1991. Dependence of the silanophiliceffect on the concentration of pre-adsorbed saltsand on chemical structure of peptides in reversedphase thin-layer chromatography. J Chromatogr553:467–475.

108. Neue DU. 1997. HPLC columns. Weinheim, Ger-many: VCH Publishers.

109. Mendez A, Bosch E, Roses M, Neue UD. 2003.Comparison of the acidity of residual silanolgroups in several liquid chromatography columns.J Chromatogr A 986:33–44.

NAL OF PHARMACEUTICAL SCIENCES, VOL. 97, NO. 8, AUGUST 2008

Page 19: log p

3002 GIAGINIS AND TSANTILI-KAKOULIDOU

110. Nyiredy SZ. 2001. Planar chromatography, aRetrospective View for the Third Millenium.Budapest: Springer.

111. Sherma J, Fried B, editors. 2003. Handbook ofThin Layer Chromatography. 3rd edition, NewYork: Marcel Dekker.

112. Ascah TL, Kallury KMR, Szafranski CA, CormanSD, Liu F. 1996. Characterization and high per-formance liquid chromatographic evaluation of anew amide-functionalized reversed phase column.J Liq Chromatogr 19:3049–3073.

113. Neue UD, Cheng YF, Lu Z, Alden BA, Iraneta PC,Phoebe CH, Tran K. 2001. Properties of reversedphase packings with an embedded polar group.Chromatographia 54:169–177.

114. Layne J. 2002. Characterization and comparisonof the chromatographic performance of conven-tional, polar-embedded, and polar-endcappedreversed-phase liquid chromatography stationaryphases. J Chromatogr A 957:149–164.

115. Claessens HA, van Straten MA. 2004. Review onthe chemical and thermal stability of stationaryphases for reversed-phase liquid chromatography.J Chromatogr A 1060:23–41.

116. Stella C, Galland A, Liu X, Testa B, Rudaz S,Veuthey J-L, Carrupt P-A. 2005. Novel RPLCstationary phases for lipophilicity measurement:Solvatochromic analysis of retention mechanismsfor neutral and basic compounds. J Sep Sci 28:2350–2362.

117. Bechalany A, Tsantili-Kakoulidou A, El Tayar N,Testa B. 1991. Measurement of lipophilicity indicesby reversed-phase high-performance liquid chro-matography: Comparison of two stationary phasesand various eluents. J Chromatogr 541:221–229.

118. Donavan SF, Pescatore MC. 2002. Method formeasuring the logarithm of the octanol–waterpartition coefficient by using short octadecyl–poly(vinyl alcohol) high-performance liquid chro-matography columns. J Chromatogr A 952:47–61.

119. Supelco. 2001. catalog, Bellefonte, PA, USA 16823(http://www.sigmaaldrich.com).

120. Liu X, Tanaka H, Yamauchi A, Testa B, ChumanH. 2004. Lipophilicity measurement by reversed-phase high-performance liquid chromatography(RP-HPLC): A comparison of two stationaryphases based on retention mechanisms. HelvChim Act 87:2866–2876.

121. Ruiz-Angel MJ, Card-Broch S, Berthod A. 2006.Ionic liquids versus triethylamine as mobilephase additives in the analysis of b-blockers.J Chromatogr A 1119:202–208.

122. Marszall MP, Baczek T, Kaliszan R. 2005. Reduc-tion of silanophilic interactions in liquid chroma-tography with the use of ionic liquids. Anal ChimActa 547:172–178.

123. Kaliszan R, Marszałł MP, Jan Markuszewski M,Baczek T, Pernak J. 2004. Suppression of dele-

JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 97, NO. 8, AUGUST 2008

terious effects of free silanols in liquid chromato-graphy by imidazolium tetrafluoroborate ionicliquids. J Chromatogr A 1030:263–271.

124. Giaginis C, Theocharis S, Tsantili-Kakoulidou A.2007. Octanol/water partitioning simulation byreversed phase HPLC for structurally diverseacidic drugs: Effect of n-octanol as mobile phaseadditive. J Chromatogr A 1166:116–125.

125. Buckenmaier SMC, McCalley DV, Euerby MR.2004. Determination of ionisation constants oforganic bases in aqueous methanol solutions usingcapillary electrophoresis. J Chromatogr A 1026:251–259.

126. Espinosa A, Bonsch E, Roses M. 2002. Retention ofionizable compounds in high-performance liquidchromatography. 14. Acid-base pK values in acet-onitrile-water mobile phases. J Chromatogr A 964:55–66.

127. Espinosa A, Bonsch E, Roses M. 2002. Retention ofionizable compounds on HPLC. IX. Modeling inreversed phase liquid chromatography as a func-tion of pH and solvent composition with acetoni-trile-water mobile phase. J Chromatogr A 947:47–58.

128. Horvath C, Melander W, Molnar I. 1976. Solvo-phobic interactions in liquid chromatography withnonpolar stationary phases. J Chromatogr 125:129–156.

129. Valko K, Slegel P. 1993. New chromatographichydrophobicity index (f0) based on the slopeand the intercept of the log k( versus organicphase concentration plot. J Chromatogr 631:49–61.

130. Valko K, Bevan C, Reynolds D. 1997. Chromato-graphic hydrophobicity index by fast-gradientRP-HPLC: A high-throughput alternative to logP/log D. Anal Chem 69:2022–2029.

131. Du CM, Valko K, Bevan C, Reynolds D, AbrahamMH. 1998. Rapid gradient RP-HPLC method forlipophilicity determination: A solvation equationbased comparison with isocratic methods. AnalChem 70:4228–4234.

132. Valko K. 2004. Application of high-performanceliquid chromatography based measurementsof lipophilicity to model biological distribution.J Chromatogr A 1037:299–310.

133. Lasic DD. 1993. Liposomes: From physics to appli-cations. Amsterdam: Elsevier.

134. Cevc G, editor. 1993. Phospholipid Handbook. 1stedition, New York: Marcel Dekker, Inc. p 5.

135. Hauser H, Pascher I, Pearson RH, Sundell S. 1981.Preferred conformation and molecular packing ofphosphatidylethanolamine and phosphatidylcho-line. Biochim Biophys Acta 650:21–51.

136. Vemuri S, Rhodes CT. 1995. Preparation andcharacterization of liposomes as therapeutic deliv-ery systems: A review. Pharm Acta Helv 70:95–111.

DOI 10.1002/jps

Page 20: log p

ALTERNATIVE MEASURES OF LIPOPHILICITY 3003

137. Plember van Balen G, Marca-Martinet C, CaronG, Bouchard G, Reist M, Carrupt P-A, Fruttero R,Gasco A, Testa B. 2004. Liposome/water lipophi-licity: Methods, information content, and pharma-ceutical applications. Med Chem Rev 24:299–324.

138. Hope MJ, Bally MB, Webb G, Cullis PR. 1985.Production of large unilamellar vesicles by a rapidextrusion procedure. Characterization of size dis-tribution, trapped volume and ability to maintaina membrane potential. Biochim Biophys Acta 812:55–65.

139. Marsch D, editor. 1990. Handbook of lipid bilayers.Boston: CRC, Press.

140. Thomas PG, Seelig J. 1993. Binding of the calciumantagonist flunarizine to phosphatidylcholinebilayers: Charge effects and thermodynamics.Biochem J 291:397–402.

141. Bauerle HD, Seelig J. 1991. Interaction of chargedand uncharged calcium channel antagonists withphospholipid membranes. Binding equilibrium,binding enthalpy, and membrane location. Bio-chem 30:7203–7211.

142. Wenk MR, Fahr A, Reszka R, Seelig J. 1996.Paclitaxel partitioning into lipid bilayers.J Pharm Sci 85:228–231.

143. Seelig J, Macdonald PM, Scherer PG. 1987. Phos-pholipid headgroup as sensors of electrical chargein membranes. Biochem 26:7535–7541.

144. Kramer SD, Wunderli-Allenspach H. 1996. ThepH dependence in the partitioning behaviour of(RS)-[3H]propranolol between MDCK cell lipidvesicles and buffer. Pharm Res 13:1851–1855.

145. Kramer SD, Jakits-Dieser C, Wunderli-Allen-spach H. 1997. Free fatty acids cause pH-depen-dent changes in drug-lipid membrane interactionsaround physiological pH. Pharm Res 14:827–832.

146. Alcorn CJ, Simpson RJ, Leahy DE, Peters TJ.1993. Partition and distribution coefficients ofsolutes and drugs in Brush Border membranevesicles. Biochem Pharmacol 45:1775–1782.

147. Pauletti GM, Wunderli-Allenspach H. 1994. Parti-tion coefficients in vitro: Artificial membranes as astandardized distribution model. Eur J Pharm Sci1:273–281.

148. Miller KW, Yu S-CT. 1977. The dependence ofthe lipid bilayer membrane: Buffer partition coef-ficient of pentobarbitone on pH and lipid composi-tion. Br J Pharm 61:57–63.

149. Austin RP, Davis AM, Manners CN. 1995. Parti-tioning of ionized molecules between aqueous buf-fers and phospholipid vesicles. J Pharm Sci 54:1180–1183.

150. Ottiger C, Wunderli-Allenspach H. 1997. Partitionbehavior of acids and bases in a phosphatidylcho-line liposome buffer equilibrium dialysis system.Eur J Pharm Sci 5:223–231.

151. Balon K, Mueller BW, Riebesehl BU. 1999. Drugliposome partitioning as a tool for the prediction of

DOI 10.1002/jps JOUR

human passive intestinal absorption. Pharm Res16:882–888.

152. Avdeef A, Box KJ, Comer JEA, Hibbert C, TamKY. 1997. pH-metric logP. 10. Determination ofvesicle membrane—Water partition coefficients ofionisable drugs. Pharm Res 15:208–214.

153. Xiang T-X, Anderson BD. 1995. Phospholipid sur-face density determines the partitioning and per-meability of acetic acid in DMPC: Cholesterolbilayers. J Membrane Biol 148:157–167.

154. Xiang T-X, Anderson BD. 1995. Development of acombined NMR paramagnetic ion-induced line-broadening dynamic light scattering method forpermeability measurements across lipid bilayermembranes. J Pharm Sci 84:1308–1315.

155. Fruttero R. 2001. NMR spectroscopy for the studyof drug-phospholipid interactions. In: Testa B, vande Waterbeemd H, Folkers G, Guy RH, editors.Pharmacokinetic optimization in drug research:Biological physicochemical and computationalstrategies. Verlag Helvetica Chimica Acta. Wein-heim: Zurich and Wiley—VCH. pp 513–524.

156. Seydel JK. 1991. Nuclear magnetic resonance anddifferential scanning calorimetry as tools forstudying drug-membrane interactions. TIPS 12:368–371.

157. Kramer S. 2001. Liposomes/water partitioning:Theory techniques and applications. In: Testa B,van de Waterbeemd H, Folkers G, Guy R, editors.Pharmacokinetic optimization in drug research,Verlag Helvetica Chimica Acta. Weinheim: Zurichand Wiley—VCH.

158. Pidgeon C, Ong S, Liu H, Qui X, Pidgeon M,Dantzig AH, Munroe J, Hornback WJ, KasherJS, Glunz L, Szczerba T. 1995. IAM chromatogra-phy: An in vitro screen for predicting drug mem-brane permeability. J Med Chem 38:590–594.

159. Ong S, Liu H, Pidgeon C. 1996. Immobilized arti-ficial membrane chromatography: Measurementsof membrane partition coefficient and predictingdrug membrane permeability. J Chromatogr A728:113–128.

160. Taillardat-Bertschinger A, Carrupt PA, BarbatoF, Testa B. 2003. Immobilized artificial membraneHPLC in drug research. J Med Chem 46:655–665.

161. Markovich RJ, Stevens JM, Pidgeon C. 1989.Fourier transform infrared assessed of membranelipids immobilized to silica: Leaching and stabilityof immobilized artificial membrane-bonded phase.Anal Biochem 182:237–244.

162. Barbato F, di Martino G, Grunetto G, La RotondaMI. 2004. Prediction of drug-membrane interac-tions by IAM-HPLC: Effects of different phospho-lipid stationary phases on the partition of bases.Eur J Pharm Sci 22:261–269.

163. Morse K, Pidgeon C. 2001. Importance of mobilephase in immobilized artificial membrane chroma-tography. In: Testa B, van de Waterbeemd H,

NAL OF PHARMACEUTICAL SCIENCES, VOL. 97, NO. 8, AUGUST 2008

Page 21: log p

3004 GIAGINIS AND TSANTILI-KAKOULIDOU

Folkers G, Guy R, editors. Pharmacokineticoptimization in drug research, Verlag HelveticaChimica Acta. Weinheim: Zurich and Wiley–VCH.pp 351–381.

164. Rhee D, Markovich R, Chae WG, Qiu X, Pidgeon C.1994. Chromatographic surfaces prepared fromphosphatidylcholine ligands. Anal Chim Acta297:377–386.

165. Taillardat-Bertschinger A, Galland A, CarruptPA, Testa B. 2002. Immobilized artificial mem-brane (IAM)-HPLC: Proposed guidelines for tech-nical optimization of retention measurements.J Chromatogr A 953:39–53.

166. Ong S, Liu H, Pidgeon C. 1996. Immobilized arti-ficial membrane chromatography: Measurementsof membrane partition coefficient and predictingdrug membrane permeability. J Chromatogr A728:113–128.

167. Sheng Q, Schulten K, Pidgeon C. 1995. Moleculardynamics simulation of immobilized artificialmembranes. J Phys Chem 99:11018–11027.

168. Ong S, Pidgeon C. 1995. Thermodynamics ofsolute partitioning into immobilized artificialmembranes. Anal Chem 67:2119–2128.

169. Ottiger C, Wunderli-Allenspach H. 1999. Immobi-lized artificial membrane (IAM)-HPLC for parti-tion studies of neutral and ionized acids and basesin comparison with the liposomal partition sys-tem. Pharm Res 16:643–650.

170. Taillardat-Bertschinger A, Barbato F, QuerciaMT, Carrupt P-A, Reist M, La Rotonda MI, TestaB. 2002. Structural properties governing retentionmechanisms on immobilized artificial membrane(IAM) HPLC columns. Helv Chim Acta 85:519–532.

171. Pidgeon C, Venkataram UV. 1989. Immobilizedartificial membrane chromatography: Supportscomposed of membrane lipids. Anal Biochem176:36–47.

172. Barbato F, Martino di G, Grumetto L, La RotondaMI. 2005. Can protonated beta-blockers inter-act with biomembranes stronger than neutralisolipophilic compounds? A chromatographicstudy on three different phospholipid stationaryphases (IAM-HPLC). Eur J Pharm Sci 25:379–386.

173. Vrakas D, Giaginis C, Tsantili-Kakoulidou A.2006. Different retention behaviour of structurallydiverse basic and neutral drugs in immobilizedartificial membrane (IAM) and reversed-phaseHPLC. Comparison with octanol-water partition-ing. J Chromatogr A 1116:158–164.

JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 97, NO. 8, AUGUST 2008

174. Fruttero R, Caron G, Fornatto E, Boschi D,Ermondi G, Gasco A, Carrupt PA, Testa B.1998. Mechanisms of liposomes/water partitioningof (p-methylbenzyl)alkylamines. Pharm Res 15:1407–1413.

175. Taillardat-Bertschinger A, Marca-Martinet CA,Carrupt P-A, Reist M, Caron G, Fruttero R, TestaB. 2002. Molecular factors influencing retentionon immobilized artificial membranes (IAM) com-pared to partitioning in liposomes and n-octanol.Pharm Res 19:729–737.

176. Vrakas D, Hadjipavlou-Litina D, Tsantili-Kakoulidou A. 2005. Retention of substituted cou-marins using immobilized artificial membrane(IAM) chromatography: A comparative study withn-octanol partitioning and reversed-phase HPLCand TLC. J Pharm Biomed Anal 39:908–913.

177. Barbato F. 2006. The use of Immobilised artificialmembrane (IAM) chromatography for determina-tion of lipophilicity. Curr Comp Aided Drug Design2:341–352.

178. Osterberg T, Svensson M, Lundahl P. 2001.Chromatographic retention of drug molecules onimmobilised liposomes prepared from egg phos-pholipids molecules and from chemically purephospholipids. Eur J Pharm Sci 12:427–439.

179. Stewart BH, Chung FY, Tait B, John C, Chan OE.1998. Hydrophobicity of HIV protease inhibitorsby immobilized artificial membrane chromatogra-phy: Application and significance to drug trans-port. Pharm Res 15:1401–1406.

180. Genty M, Gonzalez G, Clere C, Desangle-Gouty V,Legendre JY. 2001. Determination of passiveabsorption through the rat intestine usingchromatographic indices and molar volume. EurJ Pharm Sci 12:223–229.

181. Ducarme A, Neuwels M, Goldstein S, MassinghamR. 1998. IAM retention and blood brain barrierpenetration. Eur J Med Chem 33:215–223.

182. Reichel A, Begley DJ. 1998. Potential of immobi-lized artificial membranes for predicting drugpenetration across the blood-brain barrier. PharmRes 15:1270–1274.

183. Lazaro E, Rafols C, Abraham MH, Roses M. 2006.Chromatographic estimation of drug dispositionproperties by means of IAM and C18 columns.J Med Chem 48:4861–4870.

184. Barbato F, di Martino G, Grumetto L, La RotondaMI. 2007. Retention of quinolones on humanserum albumin and a1-acid glycoprotein HPLCcolumns: Relationships with different scales oflipophilicity. Eur J Pharm Sci 30:211–219.

DOI 10.1002/jps