JOURNAL OF ‘
`
`NUMBER 1
`
`JANUARY 1997
`
`ISSN 0022-3573
`
`—
`
`?.H‘Afi!9lABY L!-SEA-E%’i‘
`
`CFAD V. Anacor, |PR2015-01776
`ANACOR EX. 2192 - 1/8
`Royal Pharmaceutical Society of Great Britain
`
`CFAD v. Anacor, IPR2015-01776
`ANACOR EX. 2192 - 1/8
`
`

`
`Journal of Pharmacy and Pharmacology
`
`Published by The Royal Pharmaceutical Society of Great Britain
`1 Lambeth High Street, London SE1 7JN Telephone 0171-735 9141 Telegrams Pharrnakon London SE1
`E-Mail JPP@dial.pipex.com
`
`FAX 0171-820 3917
`
`EDITOR
`
`Dr J. CHAMBERLAIN
`
`ASSISTANT EDITOR
`
`Dr A. L. SUGDEN
`
`EDITORIAL ASSISTANT
`
`G. M. MCMAHON
`
`EDITORIAL BOARD
`
`Board Members
`
`Professor B. W. BARRY, University of Bradford
`Professor E. BEUBLER, University of Graz, Austria
`Professor N. G. BOWERY, University of Birmingham
`Professor W. C. BOWMAN, University of Strathclyde
`Dr K. J. BROADLEY, Welsh School of Pharmacy, Cardiff
`Dr D. A. COWAN, King’s College, London
`Professor S. P. DENYER, University of Brighton
`Professor F. J. EVANS, School of Pharmacy, London
`Professor A. T. FLORENCE, School of Pharmacy, London
`Professor J. L. FORD, Liverpool John Moores University, Liverpool
`Professor D. GANDERTON OBE (Chairman), British Pharmacopoeia, London
`Professor P. G. JENNER, King’s College, London
`Professor T. M. JONES, Association of British Pharmaceutical Industry, London
`Professor 1. W. KELLAWAY, Welsh School of Pharmacy, Cardiff
`Dr W. E. LINDUP, University of Liverpool
`Professor R. J. NAYLOR, University of Bradford
`Professor K. D. RAINSFORD, Sheffield Hallam University, Sheffield
`Professor C. RAPER, Victorian College of Pharmacy, Parkville, Australia
`Professor B. TESTA, University of Lausanne, Switzerland
`Dr E. TOMLINSON, GeneMedicine Inc., Texas, USA
`Professor G. T. TUCKER, Hallamshire Hospital, Sheffield
`Dr B. WIDDOP, Poisons Unit, New Cross Hospital, London
`
`Secretary to the Board
`I. FERGUSON OBE
`
`COPYRIGHT © 1997 Journal ofPharmacy and Pharmacology.
`
`Authorization to photocopy items for internal or personal use, or the internal or personal use of specific clients, is granted by the Joumal of Pharmacy
`and Pharmacology for libraries and other users registered with the Copyright Clearance Center (CCC) Transactional Reporting Service, provided that
`the base fee of $2.00 per copy, plus 0.10 per page is paid directly to CCC, 21 Congress St, Salem, MA 01970, USA.
`CODE: 0022-3573/90 $2.00+0.lO PP.
`
`Annual subscription for 1997: UK and EC £360 (inc. postage); America and Japan $635; rest of world £395 (air speeded).
`Claims for missing copies cannot be considered unless received within three months of publication.
`
`CFAD V. Anacor, |PR20’|5-01776
`ANACOR EX. 2192 - 2/8
`
`CFAD v. Anacor, IPR2015-01776
`ANACOR EX. 2192 - 2/8
`
`

`
`Contents
`
`VOLUME 49 0 NUMBER 1 0 JANUARY 97
`
`Medicinal Chemistry
`A KOUROUNAKIS N BODOR J SIMPKINS
`Synthesis and evaluation of brain-targeted chemical
`delivery systems for the neurotrophomodulator 4-
`methylcatechol
`F J GARCIA—MARCH R GARClA—DOMENECH
`J GALVEZ G M ANTON-FOS J v DE JULIAN-ORTIZ
`R GINER—PONS M C RECIO—IGLESlAS
`Pharmacological studies of l—(p-chloro-
`phenyl)propano1 and 2-(l-hydroxy-3-
`butenyl)phenol: two new nonnarcotic analgesics
`designed by molecular connectivity
`Pharmaceutics
`B B LUNDBERG
`A submicron lipid emulsion coated with
`arnphipathic polyethylene glycol for parenteral
`administration of paclitaxel (Taxol)
`
`Biopharmaceutics
`T YAG1 K YAMAUCHI S KUWANO
`The synergistic purgative action of aloe-emodin
`anthrone and rhein anthrone in mice: synergism in
`large intestinal propulsion and water secretion
`H YUASA C KUNO J WATANABE
`Comparative assessment of D-xylose absorption
`between small intestine and large intestine
`D MERTIN B C LIPPOLD
`
`In-vitro permeability of the human nail and of a
`keratin membrane from bovine hooves: influence of
`the partition coefficient octanol/water and the water
`solubility of drugs on their permeability and
`maximum flux
`
`Drug Metabolism and Pharmacokinetics
`S TAKEDA T ISONO Y WAKUI Y MIZUHARA
`S AMAGAYA M MARUNO M HATTORI
`In-vivo assessment of extrahepatic metabolism of
`paeoniflorin in rats: relevance to intestinal floral
`metabolism
`
`D POISSON A DUPUIS M BRETEAU S BOUQUET
`M POTERRE W COUET
`Effect of tamoxifen on the phannacokinetics of
`theophylline in rats
`
`Toxicology
`A M TOTTERMAN N G M SCHIPPER D 0 THOMPSON
`J—P MANNERMAA
`Intestinal safety of water-soluble Li-cyclodextrins
`in paediatric oral solutions of spironolactone:
`effects on human intestinal epithelial Caco-2-
`cells
`
`s MARTIN-ARAGON J M BENEDI A M VILLAR
`Modifications on antioxidant capacity and lipid
`peroxidation in mice under fraxetin treatment
`M D COLEMAN J K SMITH A D PERRIS N S BUCK
`J K SEYDEL
`Studies on the inhibitory effects of analogues of
`dapsone on neutrophil-function in-vitro
`
`Receptor Pharmacology
`M J RAMIREZ E GARCiA-GARAYOA G ROMERO
`A MONGE J ROCA J DEL Rio B LASHERAS
`VB20B7, a novel 5-HT-ergic agent with
`gastrokinetic activity. 1. Interaction with 5-HT;
`and 5-HT4 receptors
`
`E GARCIA-GARAYOA A MONGE J ROCA J DEL Rio
`B LASHERAS
`VBZOB7, a novel 5-HT-ergic agent with
`gastrokinetic activity. II. Evaluation of the
`gastroprokinetic activity in rats and dogs
`Cardiovascular Pharmacology
`M Q PAIVA M J SANTOS A ALBINO-TEIXEIRA
`Endothelium-dependent vascular responses in
`l ,3 -dipropyl—8-sulphophenylxanthine (DPSPX)
`hypertensive rats
`JRSHEUWCKOWCHUNGHCPENGTFHUANG
`Interaction of thrombin-activated platelets with
`extracellular matrices (fibronectin and
`vitronectin): comparison of the activity of
`arg-gly-asp-containing venom peptides and
`monoclonal antibodies against glycoprotein IIb/
`Illa complex
`T-B LIU H-C LIN Y—T HUANG C—M SUN C-Y HONG
`Portal hypotensive efiects of tetrandine and
`veraparnil in portal hypertensive rats
`Neuropharmacology
`Y ZHENG B RUSSELL D SCHMIERER R LAVERTY
`The effects of aminorex and related compounds
`on brain monoamines and metabolites in CBA
`mice
`
`l02~1()4
`
`105-107
`
`108-112
`
`B V MACFARLANE A WRIGHT H A E BENSON
`Reversible blockade of retrograde axonal
`transport in the rat sciatic nerve by vincristine
`Natural Products
`L M WANG T YAMAMOTO X X WANG LYANG
`Y KOIKE K SHIBA S MINESHITA
`Effects of Oren—gedoku-to and Unsei-in,
`Chinese traditional medicines, on
`interleukin-8 and superoxide dismutase in
`rats
`
`SREEJAYAN M N A RAO
`Nitric oxide scavenging by curcuminoids
`Biochemical Pharmacology
`1 TAMAI H TAKANAGA H MAEDA H YABUUCHI
`Y SA] Y SUZUKI A TSUJI
`Intestinal brush-border membrane transport
`of monocarboxylic acids mediated by proton-
`coupled transport and anion antiport
`mechanisms
`
`113-118
`
`R IGARASHI Y TSUTSUMI H FUJII S TSUNODA
`A OCHIAI M TAKENAGAY MORIZAWA T MAYUMI
`Y MIZUSI-IIMA
`Lecithinization of IL-6 enhances its
`thrombopoietic activity in mice
`
`CFAD V. Anacor, |PR2015-01776
`ANACOR EX. 2192 - 3/8
`
`CFAD v. Anacor, IPR2015-01776
`ANACOR EX. 2192 - 3/8
`
`

`
`This material may be protected by Copyright law (Title 17 U.S. Code)
`
`J. Pharrn. Pharmacol. 1997, 49: 30-34
`Received July 9, 1996
`Accepted August 1, 1996
`
`© 1997 J. Pharm. Pharmacol.
`
`In—vitro Permeability of the Human Nail and of a Keratin
`Membrane from Bovine Hooves: Influence of the Partition
`Coefficient Octanol/Water and the Water Solubility of Drugs on
`their Permeability and Maximum Flux
`
`1>
`
`DIRK MERTIN AND BERNHARD C. LIPPOLD
`
`Department of Pharmaceutical Technology, Heinrich-Heine-University, Universitiitsstr. 1,
`D-40225 Diisseldmj‘, Germany
`
`Abstract
`
`Penetration of homologous nicotinic acid esters through the human nail and a keratin membrane from bovine
`hooves was investigated by modified Franz diffusion cells in-vitro to study the transport mechanism.
`The partition coefficient octanol/water PCOC,/W of the esters was over the range 7 to >51000. The
`permeability coefficient P of the nail plate as well as the hoof membrane did not increase with increasing
`partition coefficient or lipophilicity of the penetrating substance. This indicates that both barriers behave like
`hydrophilic gel membranes rather than lipophilic partition membranes as in the case of the stratum comeum.
`Penetration studies with the model compounds paracetamol and phenacetin showed that the maximum flux was
`first a function of the drug solubility in water or in the swollen keratin matrix. Dissociation hindered the
`diffusion of benzoic acid and pyridine through the hoof membrane. Since keratin, a protein with an isoelectric
`point of about 5, is also charged, this reduction can be attributed to an exclusion of the dissociating substance
`due to the Donnan equilibrium. Nevertheless, the simultaneous enhancement of the water solubility makes a
`distinct increase of the maximum flux possible.
`In order to screen drugs for potential topical application to the nail plate, attention has to be paid mainly to
`the water solubility of the compound. The bovine hoof membrane may serve as an appropriate model for the
`nail.
`
`Fungal diseases of the nail plate account for up to 30-45% of
`all topical mycoses and the prevalence among the population
`of the industrial nations is about 5% (Evans 1990). Serious
`infections are treated with systemic antimycotics over several
`months, which exposes the organism to a considerable amount
`of drug. Since the introduction of nail lacquers, topical treat-
`ment has become more and more successful for light and
`intermediate mycoses (Qadripur et al 1981; Polak & Zaug
`1990; Meisel 1992; Nolting & Seebacher 1993; Haria &
`Bryson 1995).
`Until now there has been no commonly accepted in-vitro
`model for testing new,
`topically applicable drugs. Because
`human nails are not available in a sufficient number, a model
`based on membranes from the bovine hoof was developed, the
`structure and permeability of which equates with those of the
`human nail. Up to now the mechanism of nail penetration is
`largely uncertain. From investigations of the penetration of
`homologous alcohols through the nail plate, Walters et al
`(1983, 1985b) concluded that it is a hydrophilic gel membrane.
`They postulated an additional lipophilic route for the diffusion
`of lipophilic substances. Therefore, the present work examines
`the dependency of the permeability of both the nail plate and
`the hoof membrane on the partition coefficient between 1-
`octanol and water (PCOC,/w). Homologous nicotinic acid esters
`were used as model compounds with PCOC,/W values over the
`range 7 to > 51000 (Table 1).
`
`Correspondence: B. C. Lippold, Department of Pharmaceutical
`Technology, Heinrich-Heine-University, Universitatsstr. 1, D-40225
`Dusseldorf, Germany.
`
`However, the maximum amount of drug, reaching the side
`of action per time after application to an area A,
`is more
`interesting than the permeability. While in the case of a lipo-
`philic partition membrane the maximum flux depends on the
`drug solubility in the vehicle and the partition coefficient
`(Hagedom-Leweke & Lippold 1995), it increases with grow-
`ing water solubility in the case of hydrophilic gel membranes.
`The dependency of the maximum flux through the nail plate
`and the hoof membrane was investigated with the model
`compounds paracetamol and phenacetin whose solubilities in
`water differ from each other by a factor of 17 (Csw = 16 800
`and 950 mg L"], PCOCUW = 5-7 and 70-8, respectively).
`The solubility of acidic or basic substances in water can be
`increased greatly by dissociation. Enhancement of the max-
`imum flux through the nail plate is possible only when the
`decrease of the permeability coefficient due to dissociation is
`not too strong. Ions cannot penetrate lipid membranes due to
`low partitioning into the membrane. However, they are able to
`overcome hydrophilic gel membranes (Zaikov et al 1988). In
`this respect the penetration of the model compounds benzoic
`acid (pKa 4-19) and pyridine (pK,, 5-19) through the hoof
`membrane was investigated at pH 2-0 and 7-4. In these milieus
`they are nearly completely dissociated or undissociated. The
`neutral benzyl alcohol served as a standard in order to consider
`the possible influence of the pH value on the keratin swelling.
`
`Materials and Methods
`
`Chemicals
`
`Phosphate buffer solution with sodium chloride pH 7-4 (DAB
`1996), hydrochloric acid buffer pH 20 or ammonium chloride
`
`CFAD V. Anacor, |PR2015-01776
`ANACOR EX. 2192 - 4/8
`
`CFAD v. Anacor, IPR2015-01776
`ANACOR EX. 2192 - 4/8
`
`

`
`IN—VITRO PERMEABILITY OF NAIL AND KERATIN
`
`Table 1. Physical and chemical parameters of the nicotinic acid esters (Le & Lippold 1995).
`Substance
`MW
`
`PCWW
`< my: ‘>
`CsOct
`CSW _ 1
`ms
`(mg mL )
` _j¥_
`
`6-9
`7624
`1106
`137-1
`Methyl nicotinate
`22
`1034
`47
`151-1
`Ethyl nicotinate
`292.
`715
`2-45
`179-2
`Butyl nicotinate
`3233
`548
`0-17
`207-3
`Hexyl nicotinate
`51182
`527
`0-01
`235-3
`Octyl nicotinate
`
`
`MW=molecular weight, C,w=solubility in water, C,oc,=so1ubility in 1-octanol, PCOC,/W:
`partition coefficient between 1-octanol and water.
`
`buffer pH 10-0 (DAB 1996) were used as vehicles. All solutions
`were standardized to an ionic strength of 0-158. Methyl, ethyl,
`butyl and hexyl nicotinate were obtained from Aldrich-Chemie
`(Steinheim), octyl nicotinate from the Department of Pharma-
`ceutical Chemistry of the University of Diisseldorf, paracetamol
`from Boehringer lngelheim KG (Ingelheim), phenacetin from
`Bayer AG (Leverkusen), benzoic acid from Caesar & Lorentz
`(Hilden), benzyl alcohol from Janssen Chimica (Geel, Belgium)
`and pyridine from Kraft GmbH (Duisburg). HPLC grade
`methanol and acetonitrile are products of Riedel-de Haen
`(Seelze, Germany). Water was used freshly distilled.
`
`Diflizsion cells
`Two different diffusion cells were used for the investigations
`of the nail plates and hoof membranes (Mertin 1995). Both
`cells were modifications of the Franz diffusion chamber (Franz
`1975). The vertically oriented donor and acceptor compart-
`ment were joined together by clamps and contained 100 or
`50 mL, respectively. As curved nails must be fixed under
`pressure to seal the cell, the two—part donor compartment was
`built from stainless steel. The nail plate was fixed between
`these two parts screwing them together.
`
`Preparation of the nail plates and hoof membranes
`Healthy nails of the big toe were taken from dead men and
`women, aged 19-61 years, 24 h post mortem at
`the latest.
`Adhering tissue of the nail bed or nail fold was removed by the
`method of Kligrnan and Christophers (Kligman & Christo-
`phers 1963). With the ventral side downwards the nails were
`placed on a cotton pad which was soaked with O-5% trypsin in
`phosphate buffer pH 7-4 and tempered at 37°C over a period of
`10 h. Afterwards the tissue was removed with tweezers, the
`nails rinsed with distilled water and put into water for 24 h to
`remove remaining trypsin. Because keratin is hardly decom-
`posed by proteinases (Falbe & Regitz 1989),
`this method
`enabled receipt of intact nail plates.
`Hooves were taken from freshly slaughtered cattle, rid of
`adhering connective and cartilaginous tissue and put into water
`for 24 h. Afterwards about 100 pm thick membranes were
`taken from the distal part of the ball horn with a microtome
`(Reichert & Jung, Nufiloch). The hoof pieces were strong
`enough not need embedding, even in the swollen state. The
`membranes were punched with a diameter of 25 mm and dried
`at room temperature. Careful inspection of the hoof mem-
`branes by electron microscopy did not show any pores through
`the membranes but only cavities. The thickness of the swollen
`nails and membranes were measured with a micrometer cal-
`liper (Tesa Micromaster, Renens, Switzerland), in the case of
`the nails with the help of a metal ball (5-5 mm diameter).
`
`Analytical conditions
`In the case of hoof membranes the determination of the
`
`acceptor concentration was carried out spectrophotometrically
`(Zeiss DMR 10, Oberkochen). The hourly-collected samples
`were measured at
`the long-wave maximum and thereafter
`returned. In the case of the nail plates the samples had to be
`analysed by HPLC due to the smaller fluxes. The concentration
`was assayed by a modular HPLC unit (LC-6A, Shimadzu,
`Duisburg) equipped with an automatic sample injection mod-
`ule (SIL-6B), system controller (SCL-6B), UV-VIS-spectro-
`photometer
`(SPD-6AV)
`and
`an
`integrator
`(CR 4AX
`Chromatopac). The analyses were carried out at ambient
`temperature with
`a
`125 x 4 mm column packed with
`LiChrospher 100 RP-18, 5 am particle diameter (E. Merck,
`Darmstadt). The mobile phase, consisting of methanol/water
`or acetonitrile/water mixtures was pumped at flow rates ran-
`ging from 1 to 2 mL min_1. Injection volumes varied between
`5 and 150 ;rL. The concentration of the samples was calculated
`from peak areas by the external standard method.
`
`Solubilities
`
`After a rough estimate in a preliminary experiment, the amount
`of substance corresponding to twice the solubility was mixed
`with 20-50 mL solvent in a 100 mL glass bottle by rotating at
`32°C. After 24 h the samples were filtered first by paper
`(Schleicher & Schuell, Dassel) and after that by cellulose
`acetate membrane filter (0-45 am, 25 mm 9) with a filter
`device of stainless steel (Sartorius, Gtittingen). All materials
`were equilibrated at 32°C. Spectrophotornetrical assay of the
`concentration followed immediately to prevent precipitation.
`
`Penetration studies
`
`After filling the acceptor compartment, the swollen nail plate
`or hoof membrane was inserted in the diffusion cell. The cells
`were equilibrated at 32°C in a water bath (thermostat Julabo
`Paratherm IM,
`Juchheim Labortechnik, Seelbach),
`then
`100 mL donor liquid was added and samples were taken from
`the acceptor periodically. The measuring interval was deter-
`mined by the penetration rate. If it was high enough (hoof
`membrane) the samples were taken hourly. In the case of the
`nail plate samples were collected daily or weekly because of
`the low permeability and therefore long test period. Since all
`investigations with nail plates were analysed by HPLC, the
`samples (1-7 mL) could not be returned but had to be replaced
`by buffer solution. The acceptor dilution was corrected arith-
`metically (computer-aided). The media were preserved with
`0-02% sodium azide in the case of a long test period. The
`acceptor was mixed by a magnetic stirrer (Ikamag EOA 9 with
`
`CFAD V. Anacor, |PR2015-01776
`ANACOR EX. 2192 - 5/8
`
`CFAD v. Anacor, IPR2015-01776
`ANACOR EX. 2192 - 5/8
`
`

`
`32
`
`DIRK MERTIN AND BERNHARD C. LIPPOLD
`-6
`
`Controller ES 5, Janke & Kunkel, Staufen i. Br., Germany)
`during the whole course.
`The donor concentration of methyl, ethyl and butyl nicoti-
`nate was about 1000 mg L"1, while hexyl and octyl nicotinate
`were used as saturated solutions at 32°C. Octyl nicotinate was
`investigated only with the hoof membrane due to its low
`solubility and low flux. Paracetamol and phenacetin were used
`as saturated solutions at 32°C. In the case of the investigations
`of charge influence the donor concentrations were between
`1000 (pyridine) and 5000 mg L” (benzoic acid).
`
`Results and Discussion
`
`Influence of the partition caefiicient 0ctanol/water on the
`permeability coefficient
`The evaluation of the data is based on Fick’s law:
`
`5:? = D33” - (CBD — CBA)
`
`(1)
`
`fr-o———+————;————§
`HN
`
`2
`
`3
`L09 PCOct/W
`
`FIG. 1. Relationship between permeability coefficient P and partition
`coefficient octanol/water PCOWW of the nicotinic acid esters across
`human nail plate (I) and bovine hoof membrane (C) at 32°C (11 = 3--8,
`means :1: s.d.). MN = methyl
`nicotinate,
`EN = ethyl
`nicotinate,
`BN = butyl nicotinate, HN = hexyl nicotinate, ON = octyl nicotinate.
`P in cm2 s‘1
`
`case of the hoof membrane: compared with the nail plate the
`keratin matrix in the hoof membrane is probably wider-
`meshed, so that the increasing molecular weight reduces the
`diffusion coefficient only insignificantly. The enlargement of a
`molecule only in one direction, as is the case with homologs,
`merely leads to a small decrease of D, because the cross-
`section remains constant assuming orientation in the direction
`of diffusion (Flynn et al 1974).
`Although the perrneabilities presented here are in the same
`order of magnitude as the results of Walters et al (1983,
`1985b), there is an essential difference in that the decrease of
`the perrneabilities from the C2-
`to the C3-alcohol
`(from
`1-6 x 10-6 to 7-5 X 1O‘8 cm s_1) is much stronger but the
`diffusion coefficients are nearly the same. The authors justified
`the assumption of a decreasing partition coefficient nail/water
`with an endergonic transfer of a methyl group from water into
`the nail compartment. This explanation is only conclusive
`when the affinity of a substance to the nail plate decreases
`faster than its affinity to water with increasing lipophilicity.
`Harrison & Speakman (1958) also reported a decreasing dif-
`fusion of n-alcanoles in keratin with increasing chain length.
`As pure alcohols were used for these investigations, the per-
`meabilities decreased probably because of the reduced swel-
`ling of keratin or increasing molecular volume. Our results
`with homologous nicotinic acid esters correspond with the
`investigations of Walters et al (1983, 1985b) as far as the nail
`plate is characterized as a hydrophilic gel membrane rather
`than a lipophilic partition membrane. In contrast to the stratum
`comeum, the lipid content of the nail plate is much lower (0-1-
`1% (Walters & Flynn 1983); stratum comeum: 15% of dry
`weight (Flynn 1985)). The high water content of the swollen
`nail plate (about 27% (Mertin 1995)) indicates the presence of
`a hydrophilic gel membrane, too. The hoof membrane behaves
`similarly: there is no dependency of the PCQC,/W or the lipo-
`philicity on the permeability of the penetrating drug.
`
`Maximum fluxes and water salubilities
`The equation for calculating the maximum flux (Jmax) follows
`directly from the first Fick’s law (eqn 1). Jmax is the amount of
`substance that penetrates through a barrier of the area A and
`the thickness h]; per unit of time from a saturated solution
`
`CFAD V. Anacor, |PR2015-01776
`ANACOR EX. 2192 - 6/8
`
`in which dM/dt is the penetrating amount per time, DB the
`effective diffusion coefficient in the barrier, hg the thickness of
`the barrier, and CBD and CBA the concentration in the barrier
`on the donor and the acceptor side, respectively. As the con-
`centration in the barrier is normally unknown, CED is replaced
`by PCB/V-CV, in which PCB/V is the partition coefficient bar-
`rier/vehicle and CV is the vehicle concentration of the pene-
`trating compound in the donor. The term DB-PCB/V is called
`permeability coefficient P. On the condition that not more that
`10% of the drug amount in the donor penetrates, P can be
`calculated from the slope of a plot M vs t:
`D -90
`-A-C
`P-A-
`B
`
`B/V
`ha
`
`V_t=
`
`CV_t
`
`hB
`
`M:
`
`then it grows with
`If P is dependent on the lipophilicity,
`increasing PCB/V. This term is normally unknown and its
`determination is difficult. So PCB/V is approached by the
`partition coefficient octanol/water (PCOCI/W)
`in the case of
`biological membranes. Considering the equation of Collander
`(1947), who quantified the relationship between PCOC,/W and
`PCB/V, log P can be calculated from:
`
`log P = log DB + b + a - log PCOCI/W
`
`(3)
`
`If the solubility properties of octanol and the lipid barrier are
`not the same, the slope a differs more or less from the ideal
`value of unity but is distinctly greater than 0. But if the barrier
`behaves like a hydrophilic gel membrane,
`then the penne-
`ability is independent of the PCOC‘/w and the slope becomes 0.
`The permeability coefficients of the hoof membrane (PH)
`exceeded that of the nail plate (PN) 10- to 30-fold (Fig. 1).
`Unlike the hoof membrane, where the lag-time was only a few
`minutes,
`steady-state penetration through the nail plate
`occurred after 10 (methyl nicotinate) up to 80 (hexyl nicoti-
`nate) h. The plot of P vs PCOC,/w in a logarithmic scale
`according to equation 3 does not show a positive slope. Rather
`the permeability of the hoof membrane is independent of the
`lipophilicity of the penetrating substances (P=0-05). How-
`ever, the decrease of the permeability in case of the nail plate is
`highly significant (P = 0-01), but also contradicts the model of
`a partition membrane. It can be explained by the decreasing
`diffusion coefficient due to the increasing molecular volume. It
`can only be assumed why this factor has no influence in the
`
`CFAD v. Anacor, IPR2015-01776
`ANACOR EX. 2192 - 6/8
`
`

`
`IN-VITRO PERMEABILITY OF NAIL AND KERATIN
`
`33
`
`
`
`LogJmax(1000um)
`
`32
`
`3-4
`
`3-6
`Log Csw
`
`3-8
`
`4-0
`
`4-2
`
`4-4
`
`FIG. 2. Relationship between the logarithm of the maximum flux Jmax
`(1000 um) and the logarithm of the water solubility of the model
`compounds paracetamol (PAR) and phenacetin (PHE) across human
`nail plate (I) and bovine hoof membrane (C) at 32°C (n=4,
`means:l:s.d.). Jmx (1000 pm) in mg cm”2 s
`1; Csw in mg L71.
`
`(sink-conditions):
`
`dM,,,,, _ P
`J‘““‘ dt-A —hB CW
`
`Taking the logarithms results in:
`P
`B
`log Jmax = log? + log CSV
`
`(5)
`
`Neglecting effects of the molecular Volume and assuming a
`hydrophilic gel membrane, the term log P/hB in equation 5 is
`constant and the maximum flux is only dependent on the
`solubility of the drug in water. Then we expect, for a plot of
`logarithmic maximum fluxes vs logarithmic water solubilities,
`a linear correlation with the ideal slope of unity. This was
`confirmed by the results of the nail plate (slope 1-082) as well
`as the hoof membrane (slope 1-003, Fig. 2). For better com-
`parability the maximum fluxes were standardized to a barrier
`thickness of 1000 pm. The slight positive deviation of the
`slope in the case of the nail plate was explained by the stronger
`influence of the molecular size on the permeability. The denser
`network of filamentous keratin compared with that of the hoof
`membrane exhibited more resistance to the more voluminous
`
`phenacetin than to paracetamol. This was expressed not only
`by the lower maximum flux (Fig. 2), but also by the lower
`penneability
`coefficient
`calculated
`from these
`results:
`1-78 :|:0-32 x IOT8 cm2 s71
`for
`paracetamol
`and
`1-40:l:O-47 x 1078 cm2 s71 for phenacetin respectively.
`It has to be emphasized that the maximum flux of a sub-
`stance firstly depends on its water solubility and, considering
`
`that, this investigation with only two model compounds must
`be interpreted carefully. This agrees with the model of a
`hydrophilic gel membrane where the solubility in the barrier,
`which determines the maximum flux, corresponds to the
`solubility in the swelling medium that is water. In contrast, the
`maximum flux through a lipophilic partition membrane is a
`function of the partition coefficient and the solubility in the
`vehicle. Approaching the solubility in the barrier by the solu-
`bility in octanol, complex equations result. With their help
`maximum fluxes can be predicted by knowledge of the drug
`solubility in the vehicle and in octanol (Hagedom-Leweke
`& Lippold 1995). The consequence is that one has to pay
`attention primarily to the highly different water solubilities and
`only secondly to the more similar permeability coefficients in
`order to screen drugs for potential topical application to the
`nail plate.
`
`Penetration of electrolytes
`Dissociation of benzoic acid and pyridine led to a reduction of
`their penetration rate through the hoof membrane. Table 2
`shows that the permeability coefficients of the ionic forms
`were significantly lower
`than those of the neutral
`form.
`Because the penetration of benzyl alcohol decreases sig-
`nificantly by a half (t-test, P = 0-05) after the transition from an
`acidic to a neutral or basic milieu, the neutral compound serves
`as a standard. To this end the quotients of the permeability
`coefficients Pbenzoic acid or Ppyridine and Pbenzyl alcohol are built
`and compared with each other (Table 3).
`Dissociation reduces the diffusion rate of benzoic acid to a
`
`fifth and that of pyridine to a quarter. Assuming the isoelectric
`point of keratin is about 5 (Marshall 1983),
`it is positively
`charged at pH 2-0 and negatively charged at pH 7-4. However,
`nothing is known about the charge density. The decrease of
`penetration caused by dissociation is due to the Donnan effect
`(Meares 1968; Higa et al 1990, 1991) or the electrostatic
`repulsion between the membrane and the diffusing molecule
`(Kobayashi et al 1994). At pH 7-4 benzoate is a co-ion to the
`negatively charged keratin and hence it is displaced from the
`membrane; the lower concentration gradient then leads to a
`reduction of the permeability. In the case of an acidic milieu,
`the pyridinium—cation is displaced from the now positively
`charged keratin leading to a decreased penetration compared
`with the neutral form. Nevertheless, diffusion of ionic sub-
`stances through the hoof membrane confirm the assumption of
`a hydrophilic gel membrane (Zaikov et al 1988).
`Some authors attribute the lowering of the electrolyte dif-
`fusion through charged polymers to sorption effects (Medley
`1957; Meares 1968; Zaikov et al 1988). Since only the free
`
`Table 2. Permeability coefficients of benzyl alcohol, benzoic acid and pyridine
`through the bovine hoof membrane.
`
`Benzyl alcohol
`
`Benzoic acid
`
`Pyridine
`
`pH
`
`207-4
`10-0
`2-0
`7-4
`2-07-4
`
`Temperature = 32°C (n = 4, means i: s.d.).
`
`Permeability coefficient (1073 cm2 s“ 1)
`
`78-24:1: 16-45
`41-72:t7-07
`48-78 :l:8-40
`78-62i 16-43
`8-29 21:1-36
`19-l4:l:7-05
`44~80i 10-35
`
`CFAD V. Anacor, |PR2015-01776
`ANACOR EX. 2192 - 7/8
`
`CFAD v. Anacor, IPR2015-01776
`ANACOR EX. 2192 - 7/8
`
`

`
`34
`
`DIRK MERTIN AND BERNHARD c. LIPPOLD
`
`Table 3. Permeability coefficients of benzoic acid (PBAC) and pyr-
`idine <PpyR) through the bovine hoof membrane at pH 2-0 and 7-4,
`standardized to benzyl alcohol (PBAL).
`
`Benzoic acid
`
`Pyridine
`
`PH
`
`2-0
`7 -4
`2-0
`7-4
`
`Temperature = 32°C (n = 4).
`
`PBAC/PBAL
`
`PPYR/PBAL
`
`1-00
`0-20
`
`fraction is able to penetrate, sorption to functional groups
`could lead to a decrease of the concentration gradient. How-
`ever, investigations with polydimethylsiloxane films contain-
`ing a silica filler show that sorption only causes a prolongation
`of
`the
`lag-time but not
`a
`reduced
`steady-state
`flux
`(Flynn & Roseman 1971). Due to the dynamic exchange of the
`free substance between membrane and donor solution,
`the
`resulting concentration of the free drug in the membrane will
`not alter after the adjustment of the sorption equilibrium at the
`steady state and is independent of the degree of saturation of
`the functional groups, on the condition that there is no decrease
`of the donor concentration by the sorption. Therefore, an
`interpretation of the hindered penetration of charged molecules
`is preferred on the basis of the Donnan potential.
`The reduced permeability of the neutral compound benzyl
`alcohol at pH 7-4 as compared with pH 2-0 is probably caused
`by a decrease of the keratin swelling due to the charge inversion
`of the keratin at the transition from an acidic to a neutral or basic
`
`environment. The weak, not significant increase of the pene-
`tration of the benzyl alcohol changing the pH from 7-4 to 10-0
`lends further support to this hypothesis, because the number of
`negatively charged groups in keratin (IEP~5) at pH 10-0 is
`nearly the same as in a milieu of pH 7-4.
`Applying these results to the human nail, they contradict the
`investigations of Walters et al (l985a), who could not find a
`dependency of the nail permeability of miconazole on the
`degree of dissociation. Maybe this contradiction is explained
`by the higher ionic strength of the solutions; in this case the
`influence of the Donnan equilibrium is lowered. Since the
`amino acid composition of keratin is dependent on the species
`(Fraser & MacRae 1973), the different results could also be
`put down to varying charge densities of hoof and nail keratin.
`The reduction of the hoof permeability to a quarter caused by
`complete dissociation of the substance is much lower than the
`possible enhancement of the solubility of hardly soluble acids
`and bases. So the solubility is increased a hundred fold at 99%
`degree of ionization (assuming sufficient solubility of the salt).
`Since the maximum flux is primarily a function of the solu-
`bility in water (see above), it is possible to heighten the per-
`meability of both the hoof membrane and the nail plate by
`dissociation of the penetrating drug if the milieu in the barrier
`causes no precipitation of the substance.
`
`References
`
`Collander, R. (1947) On “lipoid solubility”. Acta Physiol. Scand. 13:
`363-381
`Evans. E. G. V. (1990) Nail derrnatophytosis: the nature and scale of
`the problem. J. Dermatol. Treat. 1 (Suppl. 2): 47-48
`Falbe, J., Regitz, M. (1989) Rtimpp Chemie-Lexikon. G. Thieme
`Verlag, Stuttgart, pp 2195-2196
`
`(1985) Mechanism of percutaneous absorption from
`Flynn, G. L.
`physicochemical evidence.
`In: Bronaugh, R. L., Maibach, H.
`I.
`(eds) Percutaneous Absorption: Mechanisms —Methodology-
`Drug Delivery (Dermatology, Vol. 6). Marcel Dekker Inc., New
`York, pp 17-41
`Flynn, G. L., Roseman, T. J. (1971) Membrane diffusion II: influence
`of physical adsorption on the molecular flux through heterogeneous
`dimethylpolysiloxane barriers. J. Pharm. Sci. 60: 1788-1796
`Flynn, G. L., Yalkowsky, S. H.