`
`J. Pharm. Pharmaco!. 1985, 37:771-775
`Received May 3, 1985
`
`© 1985 J. Pharm. Pharmacol.
`
`Physicochemical characterization of the human nail:
`solvent effects on the permeation of homologous
`alcohols*
`
`KENNETH A. WALTERSt, GORDON L. FLYNN AND JOHN R. MARVEL~
`
`College of Pharmacy, University of Michigan, Ann Arbor, Michigan, 48109 and ¢Dermatological Division, Ortho
`Pharmaceuncal Corporation, Raritan, New Jersey, USA
`
`To assess how vehicles might influence permeation through human nail, the diffusion of
`homologous alcohols (methanol to decanol) administered as neat liquids through finger nail
`plate has been studied using in-vitro diffusion ceil methods and compared with permeation
`data for the same compounds in aqueous media. Permeation rates of the homologous
`alcohols through lipid depleted nail plate have also been assessed and the influences of
`dimethylsulphoxide (DMSO) and isopropyl alcohol on permeation rates of methanol and
`hexanol have been examined. With the exception of methanol, permeability coefficients are
`uniformly about five-fold smaller when the alcohols are undiluted than when they are
`applied i’n water. Overall parallelism in the permeability profiles under these separate
`circumstances of application is an indication that tile external concentrations of the alcohols
`themselves are a determinant of their permeation velocities through the nail plate matrix.
`The even separation of the profiles suggests a facilitating role of water within the nail matrix.
`Chloroform/methanol delipidization of the nail led to increased penetration rates of water,
`methanol, ethanol and butanol. On the other hand, it caused a six-fold decrease in the
`permeation rate of decanol. Application of methanol and hexanol in DMSO somewhat
`retards their rates of permeation. Isopropyl alcohol also slows the permeation rate of
`hexanol but has little influence on that of methanol. Thus it appears that solvents which tend
`to promote diffusion through the skin horny layer have little promise as accelerants of nail
`plate permeability.
`
`Complex influences of DMSO and other organic
`solvents on the permeability of skin have been
`demonstrated (Kligman 1965; Dugard & Embery
`1969; Scheuplein & Ross 1970; Scheuplein & Blank
`1973: Astley & Levine 1976). Depending on the
`balance of factors, solvents may enhance or retard
`penetration. In-vivo and in-vitro experiments with
`DMSO and like solvents indicate they may act to
`enhance skin permeation by several mechanisms,
`including extraction of the stratum corneum lipids
`(Kligman 1965: Dugard & Embery 1969), displace-
`ment of stratum corneum bound water (Scheuplein
`& Ross 1970), the latter possibility being accompan-
`ied by keratin denaturation§, and stratum corneum
`delamination (Chandrasekaran et al 1977). Yet,
`even with all these factors operating, decreased skin
`
`* This work supported through the generosity of Ortho
`Pharmaceutical Corporation. Raritan, New Jersey. Pre-
`sented in part at the British Pharmaceutical Conference,
`Brighton, UK, 1981.
`t Correspondence and present address: Fison plc, S & T
`Laboratories, Bakewell Road. Loughborough. Leicester-
`shire.
`§ Recent work in the laboratories of the College of
`Pharmacy, University of Michigan strongly indicates that
`denaturation is a major factor influencing the permeation
`facilitating role of DMSO
`
`permeation of lipophilic compounds can still occur as
`the result of exaggeratedly reduced thermodynamic
`activity of the permeants in their solubilized states in
`such solvents (Gatmaitan et al 1979). Regardless. it
`is clear that there are actual permeation altering
`effects of organic solvents within the horny layer
`structure which in large measure depend on reorgan-
`ization of lipid and fibrous protein and displacement
`of water.
`As a closely corresponding anatomical structure
`human nail might at first be expected to exhibit
`barrier behaviour paralleling that of the stratum
`corneum. Nail plate, however, contains far less lipid
`than found in the skin’s horny layer and it also has far
`less ability to absorb water (Baden et al 1973) and
`other materials (Kligman 1965) suggesting its protein
`regime is also molecularly different from that of the
`stratum corneum. The inherent differences in nail
`are evident in its unusual structure-permeability
`sensitivities to the homologous alkanols (Waiters et
`al 1981, 1983). It can be reasoned on these bases that
`the nail might act differently in the presence of
`penetration enhancers, but this idea has yet to be
`tested.
`Earlier, the construction of, and technique of use
`
`ARGENTUM EX1042
`
`Page 1
`
`
`
`772
`
`K. A. WA.LTERS ET AL
`
`of, a special diffusion chamber for studying nail plate
`permeability in-vitro was described (Waiters et aI
`1981, 1983). We have now used this cell to charac-
`terize solvent influences on the permeation behav-
`iour of the n-alkanols. The alkanols were applied to
`the nail surface as neat liquids. These data were
`compared with past results (Waiters et al 1983)
`where the alkanols were applied as highly dilute
`aqueous solutions. Permeation of select alkanols
`through chloroform-methanol delipidized nails was
`then examined, as was the permeability rates of
`methanol and hexanol from binary aqueous solutions
`of DMSO and isopropyl alcohol.
`
`MATERIALS AND METHODS
`Tritiated water and radiolabelled alcohols were
`obtained from New England Nuclear ([3H]water,
`[3H]methanol, [14C]ethanol, [14C]butanol) and ICN
`([l~C]hexanol, [14C]octanol, [14C]decanol). Reagent
`grade pure alcohols were used throughout.
`
`Permeation procedures
`Details of the diffusion cell and permeation pro-
`cedures have been given previously (Walters et al
`1981). Briefly, trimmed human nail plate sections
`were placed between two halves of a diffusion cell. A
`known amount of radiolabelled permeant in solution
`in either its corresponding neat alcohol or a binary
`(DMSO/saline, isopropyl alcohol!saline) solvent
`mixture was placed in the donor chamber and
`samples were taken at predetermined intervals from
`the receptor chamber. Isotope activity was moni-
`tored using a Beckmann LS 9000 liquid scintillation
`counter. In all the experiments the receptor chamber
`was filled with normal saline.
`
`Extraction of nail lipids
`Nail plates were immersed in a chloroform-
`methanol (3: 1) mixture for 24h to dissolve and
`remove accessible lipids. After this extraction the
`nail plates were rinsed in four changes of saline and
`placed in the diffnsion cells as described above.
`
`Data handling
`Permeability coefficients (P) were calculated from:
`
`dC!dt
`P=V--
`A-AC
`
`(1)
`
`across the membrane. V(dC/dt) gives the diffusion
`flux in mass per unit time. The diffusion cells with
`nail plate membranes in place were scrupulously
`checked for intercompartmental leakage using sol-
`uble but impenetrable polyethyleneglycol markers.
`No leaks were evident.
`
`RESULTS AND DISCUSSION
`It is obvious from Fig. 1 that the unique ability of the
`nail plate to restrict increasingly the diffusive passage
`of low molecular weight homologous alcohols as the
`alkyl chain is lengthened, seen previously when the
`alkanols were applied as highly dilute aqueous
`solutions (Waiters et al 1983). carries over to the
`circumstance when the alcohols are applied as neat
`liquids. Except for methanol, mass transfer coeffi-
`cients are uniformly about five times smaller when
`the pure liquids are permeating nail than when the
`alcohols are in dilute aqueous solution. Since per-
`meability coefficients are concentration normalized
`mass transfer parameters, the observed differences
`signify that the actual molecular rates of penetration
`are five-fold smaller when nail is exposed to the neat
`alcohols than when the nail is exposed to the alcohols
`under hydrating conditions. The chemical potential
`difference between the respective pure liquid states
`of the alcohols and their respective high aqueous
`
`3O
`
`T
`0 -
`
`’7_, 1
`
`~0.3
`
`.Q
`
`E
`;0
`n
`
`002
`
`T
`o
`/
`/
`
`Where V is the volume of the receiver half cell, dC/dt
`is the rate of change in concentration in the
`pseudo-steady state portion of the receiver con-
`centration versus time plot, A is the diffusional area
`and AC is the concentration differential of permeant
`
`Atkyt chQin length
`
`Frd. 1. Permeability coefficients of the n-alkanols through
`nail plate from dilute aqueous~ solutions (O) and from neat
`alcohols (O) as a function of alkyl chain length. Error bars
`indicate standard deviation. * Indicates significant differ-
`ences (P < 0-05).
`
`Page 2
`
`
`
`SOlVENT EFFECTS ON NAIL PERMEATION
`
`773
`
`dilutions is known to change rapidly as the alkyl
`chain is lengthened as the alkanols exponentially
`become less and less soluble in water. On this basis
`one might expect the separation between the two
`permeability coefficient curves shown in Fig. 1 to
`systematically widen with increasing alkyI chain
`length, but this does not appear to occur. Rather, the
`roughly five-fold spread in values is maintained past
`methanol despite exponentially increasing o/w par-
`titioning. A highly ’polar route’ through the nail
`plate capable of excluding permeants on the basis of
`their relative hydrophobicities was suggested as
`controlling the permeation of the alcohols, through
`octanol, when administered in water (Waiters et al
`1983). This suggestion was made in part because
`diffusivities of the alkanols in the nail plate matrix
`calculated from lag times were not greatly different
`from one another. Incrementally decreasing mem-
`brane/water partition coefficients were thus postu-
`lated to explain the sharp and systematic permeabil-
`ity coefficient decline seen from methanol to oetanol
`(Fig. 1, upper curve). It is notable that lag times in
`the present studies, while not documented, were not
`longer than those seen before. If the separation in
`the curves in Fig. 1 were strictly brought about by
`decreased diffusivity as the result of hydration,
`roughly five-fold longer lag times should have been
`observed with the neat alcohols. Thus, the per-
`meability behaviour of the neat liquid alkanols to
`octanoI seems to have no simple thermodynamic
`(partitioning) or kinetic (diffusivity) explanation to
`tie in neatly with the previous study performed with
`dilute aqueous solutions. However, the data re-
`ported here for the nail plate are to an extent
`compatible with diverse observations which suggest
`that the lower alcohols have a diminishing capacity to
`solvate hard horny tissue as the alkyl chain is
`lengthened (Harrison & Speakman 1958; Tillman &
`Higuchi 1961; Wu 1983). In this sense nail plate is
`more like callus and hair (wool) than it is like stratum
`corneum. Upon considering chemical compositions
`of these tissues, the stratum corneum seems to be set
`apart from the others behaviourally through its high
`concentration of lipids (Baden et al 1973).
`Whether administered in dilute aqueous solution
`or as the neat alcohol, decanol’s behaviour is
`experimentally set apart from that of the lower
`alkanols (Waiters et al 1983). Though molecularly
`larger, it proves a better permeant than immediately
`lower alkanols, because of its higher mass transfer
`coefficients irrespective of the method of applica-
`tion. Dodecanol behaves similarly whether adminis-
`tered in water (previous study) or in a decanol
`
`medium (not reported). The new data do not appear
`totally compatible with the previously proposed idea
`(Walters et al 1983) that there is a non-structurally
`unique parallel lipid pathway which these long chain
`compounds preferentially, diffusionally negotiate,
`because neat octanol, for example, would have equal
`access and facility of diffusion were such a route to
`exist. It was previously reasoned that exponentially
`increasing oil/water partition coefficients finally
`became sufficient to push the higher alcohols across
`the presumed intercellular lipid route, but partition-
`ing out of the neat alcohols would not follow the
`same relationship. Thus it appears there is a regime
`of diffusion offering selective access to the C~0 and
`C12 alkyl chain length compounds. There remains
`the impression that some more specific structural
`requirement is met by decanol and dodecanol which
`affords them greater solubility or mobility in this
`phase of the nail.
`The nail plate comprises less than 1% lipid (Baden
`et al 1973) compared with the =10% or more lipid
`present in stratum corneum. In the latter tissue,
`increases in the permeability rates of the polar
`n-alkanols are exaggerated after lipid extraction with
`the chloroform/methanol (Blank et al 1967) as
`extraction of lipids results in a functionally porous
`matrix. The removal of lipid from the human nail
`plate in the present studies caused an increase in the
`penetration rates of water, methanol, ethanol and
`butanol but reduced the rates of transfer of the Cl0
`and Cl2 compounds (Table 1). The presence of only
`minor amounts of lipid in the nail plate, a significant
`portion of which must be carried over as lipid from
`former cell membranes, suggests that this structure
`should be far less sensitive than the stratum corneum
`to the effects of extraction, as is observed with the
`lower chain length homologues. On the other hand,
`the high permeability coefficient of n-decanol
`
`Table 1. Permeability coefficients of water and n-alkanols
`across normal and dclipidizcd nail.
`
`Permeability coefficients"
`(X 10~3 cln h 1
`
`Permeant
`Water
`Methanol
`Ethanol
`Butanol
`Decanol
`
`Normal nail
`16.5 +_ 5.9(6)
`5.6 + 1.2 (26)
`5.8 + 3.1 (8)
`0.6 + 0.3 (4)
`2.5 _+ 1.7 (10)
`
`Delipidized nail
`22.4 _+ 3.6(5)
`10-5 + 2.3 (20)t’
`6.9 + 0.3 (5)
`2.6 + 0.8 (5)b
`0.5 + 0.05 (5)b
`
`a Data include standard deviation and ( ) number of
`experiments.
`h Indicates significant differences (P < 0.005).
`
`Page 3
`
`
`
`774
`
`K. A. WALTERS ET AL
`
`12
`
`1"0
`
`0"8
`
`0"6
`
`ca. 0-2
`
`//
`
`//
`//
`//
`
`//
`/-/
`//
`//
`//
`//
`//
`.//
`//
`C1 261Cl
`
`Saline
`
`"1-
`
`CE
`
`[c6
`50
`25
`Ccsolvent concentration (%)
`
`75
`
`Ic61
`
`86
`
`FIG. 2. Effect of co-solvent concentration on the nail permeation of methanol (C~) and n-hexanol (C6). The data is
`normalized with respect to the permeability coefficient in saline. Error bars indicate standard deviation (n = 4). Hatched
`columns DMSO; open columns isopropyl alcohol.
`
`appears to be dependent on the presence of lipid in
`the barrier as a six-fold decrease in the permeation
`rate of decanol across the nail is evident following
`delipidization. This observation is consistent with a
`minor lipid pathway which becomes rate-controlling
`for molecules as hydrophobic as decanol. The
`selectivity of effects with decanol and dodecanol
`under all conditions indicates the lipid is organized in
`a way favourable to their inclusion.
`In Fig. 2, data are provided which indicate that
`DMSO as a vehicle acts to reduce the mass transfer
`coefficients of methanol, a very polar solute, and
`also n-hexanol, an alkanol which is hydrophobic and
`practically insoluble in water. The influence on
`hexanol is decidedly greater. Studies by Kurihara (T.
`Kurihara, personal communication) in these labora-
`tories have shown, through partial vapour pres-
`sures, that at 2% v/v, the thermodynamic activity
`of methanol is gradually, systematically increased
`as the percentage of DMSO increases in binary
`DMSO-water mixtures. The systematically declin-
`ing permeability coefficient of methanol with
`increasing DMSO concentration could therefore not
`be thermodynamic and due to declining partitioning
`tendencies. However, on the same strict thermody-
`namic grounds, the permeability coefficient of hex-
`anol should rapidly become extremely small because
`its activity drops precipitously with increasing
`DMSO. It does not. The permeability coefficient is
`suppressed not much more than five-fold even with
`86% DMSO in the solvent medium. This magnitude
`of effect is the same as when hexanol is applied in its
`
`neat state relative to a highly dilute aqueous solu-
`tion, again suggesting hydration as the facilitating
`mechanism. The permeation behaviour of hexanol
`from isopropanol-water systems is similar to its
`behaviour from DMSO-water systems. The
`methanol permeability coefficient is unaltered by the
`presence of isopropanol (Fig. 2). Thus methanol
`remains the curious exception to the general alkanol
`behaviour. The effect on hexanol of substituting
`isopropanol for water in the vehicle is some five-fold,
`there being a consistent fall in rate when the applied
`medium contained little or no water.
`The enhanced permeability of stratum corneum as
`the result of contact with strongly interacting sol-
`vents such as DMSO, is in sharp contrast to the
`effects of DMSO on nail (MacGregor 1967; Montes
`et al 1967; Allenby et al 1969; Embery & Dugard
`197l; Chandrasekaran et al 1977; Jones private
`communication). Having far less lipid and a much
`more tightly woven intracellular structure (Forslind
`& Thyresson 1975), the nail, in all likelihood, is
`much less affected by alteration caused by lipid
`extraction and there is no indication it undergoes
`delamination. The nail apparently is also incapable
`of absorbing much DMSO (Kligman 1965). The
`crystallinity of nail proteins is different from that of
`the stratum corneum. All these factors set the nail
`apart from the stratum corneum with respect to
`barrier properties in general and the DMSO actions
`in particular. The inability of DMSO to enhance
`chemical penetration of hydrophilic or tipophilic
`permeants into the keratinous matrix of ovine hoof
`
`Page 4
`
`
`
`SOLVENT EFFECTS ON NAIL PERMEATION
`
`775
`
`material (Malecki & McCausland 1982), a nail
`plate-like structure, is a related observation pointing
`to differences in hard keratin structures and stratum
`corneum. It thus seems clear that solvents which tend
`to promote diffusion through the stratum corneum
`may be of little use as facilitators of nail plate
`permeation.
`
`REFERENCES
`
`Allenby, A. C., Creasey, N. H., Edgington, J. A. G.,
`Fletcher, J. A., Schock, C. (1969) Br. J. Dermatol. 81
`(Suppl. 4): 47-55
`Astley, J. P., Levine, H. (1976) J. Pharm Sei. 65:210-215
`Baden. H. P., Goldsmith, L. A., Fleming, B. (1973)
`Biochim. Biophys. Acta. 322:269-278
`Blank. I. H., Scheuplein, R. J., Macfarlane, D. J. (1967) J.
`Invest. Dermatol. 49:582-589
`Chandrasekaran, S. K., Campbell, P. S., Michaels, A. S.
`(1977) A.I.Ch.E Journal 23:810-816
`Dugard, P. H., Embery, G. (1969) Br. J. Dermatol. 81
`(Suppl. 4): 69-74
`Embery, G., Dugard, P. H. (1971) J. Invest. Dermatol. 57:
`308-311
`
`Forslind, B., Thyresson, N. (1975) Arch. Derm. Forsch.
`254:19%204
`Gatmaitan, O. G., Flynn, G. L+, BeN, C. R., Higuchi,
`W. I., Drach, J. Ho, N. F. H. (1979) Abstracts, 27th APS
`National Meeting, Kansas City, MO., Abstract No 118,
`p. 107
`Harrison, D., Speakman, J. B. (1958) Textile Res. 28:
`1005-1007
`Kligman, A. M. (1965) J. Am. Med. Assoc. 193:796--804
`MacGregor, W. S. (1967) Ann. N.Y. Acad. Sei. 141:3-10
`Malecki, 1. C., McCausland, I. P. (1982) Res. Vet. Sci. 33:
`192-197
`Monies, L. F., Day, J. L., Ward, C. J., Kennedy, L. (1967)
`J. Invest. Dermatol. 48:184--196
`Scheuplein, R. J., Blank, I. H. (1973) J. Invest. Derrnatol.
`60:286--296
`Scheuplein, R+ J., Ross, K. (1970) J. Soc. Cosmet. Chem.
`21:853-873
`Tillman, W. J., Higuchi, T. (1961) J. Invest. DermaIol. 37:
`87-92
`Waiters, K. A., Flynn, G. L., Marvel, J. R. (1981) Ibid. 76:
`76-79
`Waiters, K. A., Flynn, G. L., Marvel, J. R. (1983) J.
`Pharm. Pharmacol. 35:28-33
`Wu M. S. (1983) J. Colloid Interface Sci. 92:273-274
`
`Page 5
`
`