- Email: [email protected]
Mechanism of Skin Penetration-Enhancing Effect by Laurocapram
Mechanism of Skin Penetration-Enhancing Effect by Laurocapram KENJI SUGIBAYASHI, SATORU NAKAYAMA, TOSHINOBUSEKI, KEN-ICHIHOSOYA, AND YASUNORIMORIMOTO~ Received A ril 2, 1990, from,the. Facuky of PharmaceuHcal Sciences, Josai University, 1-1 Keyakidai, Sakado, Saitama 350-02, iccepted for publication March 1, 1991, Japan. Abstmct 0 In order to clarify the mechanism of action of laurocapram (Azone) on the skin permeation of drugs, the following experiments were done. First, the effect of Azone on the skln components was compared with that of other penetration enhancers. Azone markedly fluidized liposomal lipids (as a model lipid system) compared with other enhancers. Ethanol extracted large amounts of the stratum comeum lipids, whereas Azone did not. These results suggest that the effect of Azone on the lipids in the stratum comeum is not the same as that of ethanol. In addition, ethanol increased the amount of free sulfhydryl (SH) group of keratin in the stratum corneum, whereas Azone did not directly affect the stratum comeum protein. Azone increased water content in the stratum comeum, as measured by skin conductance. This effect might be a reason for the action of Azone. For further understanding, the enhancing effects of Azone on the skln permeation of several model compounds (alcohols, sugars, and inorganic ions) were compared with the effects of pretreatment with distilled water, which was thought to increase water-holding capacity, and pretreatment with ethanol, which was thought to affect the lipids and protein in the skin barrier (i.e., stratum corneum). Pretreatmentwith water or ethanol enhanced skin permeation of hydrophilic compounds, whereas they decreased that of octanol, a hydrophobic compound. The tendency of h o n e to increase or decrease the skin permeation rate of most compounds was similar to that of pretreatment with water or ethanol. However, the effect of Azone on the skin permeation of inorganic ions was relatively low, whereas that of pretreatmentwith water or ethanol was high. We then hypothesizedthree permeation routes in the skln: a lipid-rich route, where hydrophobic compounds can permeate; a water-rich route, where relatively hydrophilic compounds can permeate; and a pore (shunt) route, where even inorganic ions can permeate. Azone may act on the lipid-rich route because it fluidizes the stratum corneum lipids, but also on the water-rich route because it enriches the water content. In general, increase in the pore (shunt) route would be expected when lipid has been extracted by enhancers or solvents such as ethanol. Therefore, the effect of h o n e on the enlargment in the pore (shunt) route may be negligible. The reason why Azone affects the skin permeation of a relatively broad spectrum of drugs may be that it affects at least two permeation routes in the skin barrier.
Several penetration enhancers,' prodrugs,2 superfluous vehicles,s iontophoresis,4 phonophoresis,6 and thermophoresis6 have been used as methods to increase the skin permeation of drugs. Among them, penetration enhancers are one of the most convenient methods and show relatively high effects. Ideal enhancers have no irritancy and toxicity to the skin and whole body, together with high enhancing effects. Enhancers themselves should be physicochemically stable and should not have pharmacological actions, and preferably should not have smell, color, or taste. Short- and long-chain alcohols,7.~ pyrrolidones,g urea and its derivatives,'O dimethyl sulfoxide and its relatives,*' and several surfactants12 have been screened for penetration enhancement effects. Most of them, however, did not appear on the market due to high adverse effects and low efficacy. Lack of understanding of the mechanism of their penetration-enhancing effects may be a reason 58 I Journal of Pharmaceutical Sciences Vol. 81, No. 1, January 1992
retarding their approval a s ingredients in pharmaceutical and cosmetic preparations. Laurocapram (1-dodecylazacycloheptan-2-one,Azone) is a nonirritant on the skin and can be used for a broad spectrum of drugs.13 We have reported that Azone affects mainly the skin barrier, the stratum corneum.14 However, some of the effects of Azone on the physicochemical properties of drugs, vehicles, and skin [i.e., partition coefficients between n-octano1 and water of drugs and vehicles (solvents)] are still obscure. The penetration-enhancing mechanism of Azone is also unclear. Study of the mechanism of penetration enhancers is very important, not only to discover suitable enhancers, but also to put many of them on the market. In the present study, the effects of Azone on the skin barrier components and on the skin permeation of different model compounds were compared with those of other penetration enhancers. In addition, the mechanism of the penetrationenhancing effect of Azone is discussed in relation to the experimental data and the hypothesis of penetration routes of drugs through the skin barrier.
Experlmental Section Materials and Animals-Methanol (MeOH), n-butanol (BuOH), lithium chloride (LiCl), sodium chloride (NaCl), and potassium chloride (KCl) were of JP grade. The following radiolabeled compounds were used: [l-"C]-n-octanol (W-OcOH; specific activit 5 mCi/mmol, ICN Radiochemicals, Irvine, CA), D-[5,6-3Hlglucose (l HGlu; specific activity 66.9 Ci/mmol, New England Nuclear Research, Boston, MA), and [l-3H(N)]-sucrose (3H-Suc; specific activity 0.01 Ci/mmol, New England Nuclear Research). Laurocapram (1dodecylazacycloheptan-2-one,Azone) was a kind girt from NelsonSumisho Company (Tokyo,Japan). Polysorbate 20 was obtained from Wako Pure Chemical Industries (Osaka, Japan). Trypein (Type 11, crude) and L-cysteine (free base) were supplied by Sigma Chemical Company (St.Louis, MO). Liposome compositions were as follows:egg lecithin and cholesterol (Wako Pure Chemical Industries); 5,6carboxyfluorescein (6-CF; Eastman Kodak Company, New York, NY) and 1,6-diphenylhexatriene(DPH; Aldrich Chemical Company, Milwaukee, WI);and 5,6-dithiobis-2-nitrobemicacid (DTNB) and polyoxyethylene (10) octyl phenyl ether (Triton X-100; Wako Pure Chemical Industries). Other reagents were of reagent grade and were used without further purification. Enhancers used are isopropyl myristate (IPM; Tokyo Kasei, Tokyo, Japan), dodecyl acetamide (DA; Nissan Chemical Industries, Tokyo, Japan), dodecyl pyrrolidone ( D P Niiwan Chemical Industries), dimethyl sulfoxide (DMSO; Wako Pure Chemical Industries), methyl pyrmlidone (Mp, Tokyo Kasei), Sefsol-318 (S-318; Nikko Chemical Company, Tokyo, Japan), sodium salicylate (SA-Na; Wako Pure Chemical Industries), and urea (Wako Pure Chemical Industries). When making Azone, IPM, DA, or DP solutions, polysorbate 20 was used as a n emulsifier, since these are not miscible in water. DMSO, DMAC, MP, 5318, SA-Na, and urea were dissolved directly in distilled water. The concentrations of each enhancer and polysorbate 20 in water were 3.0 and 0.196, respectively.*4 Male WBNkob hairless rats weighing -150 g were used in all animal experiments. OO22-3~9/92/0100-0058$02..50/0 0 1992, American Pharmaceutical Association
Experiments with LiposomegLiposomes were prepared by a thin-layer method.16 Lecithin (32 q o l ) and cholesterol (4 q o l ) were dissolved in 2 mL of chloroform in a 30-mL round-bottomed flask, and the solvent was evaporated to form a thin layer of lipid. After the remaining solvent was dried under reduced pressure, 5 mL of20 mM 6-CF in pH 7.4 phosphate-buffered saline (PBS) or 1mL of0.28 DPH in tetrahydrofuran was added to the flask, and the mixture aeitated to make multilamellar vehicles. The resulting suspension wis ultrasonicated with a probe tip (model G50022-4 Nihonseiki Seisakusho, Tokyo, Japan) a t 100 W for 2.5 min to make smaller liposomes. The liposomes were fractionated and concentrated by ultracentrifugation (40 000 rpm, 4 "C, 30 min). Rinsing with PBS (pH 7.4) was repeated twice and, finally, 2 mL of original liposome suspension was obtained. A 1 : l O dilution of liposome suspension in PBS (0.3 mL) containing 6-CF was mixed with each enhancer solution (2.7 mL) in a thermoregulated cuvette cell (1 x 1 x 4 cm3) for a fluorescence spectrophotometer, and the release of 6-CF from liposomes was followed at 37 T.16 Increase in fluorescence intensity (excitation and emission wavelengths of 490 and 550 nm, respectively) was monitored by a fluorescence spectrophotometer (model 650-10s; Hitachi, Tokyo, Japan). At the end of each experiment, 0.1 mL of 10% (vlv) Triton X-100 was added for complete release of 6-CF from liposomes, and the resulting fluorescence intensity was assumed to represent the initial content of 6-CF in the liposomes. A 1 : l O dilution of liposome suspension (0.3 mL) containing DPH was mixed with each enhancer solution (2.7 mL) in a thermoregulated cuvette cell (1 x 1 x 4 cm3)which contained a magnetic stirrer bar. Ten minutes after stirring, steady-statefluorescence polarization (PDof DPH intercalated in liposomal bilayers was measured on a fluorescence polarization spectrophotometer (model FS-501s; Union Giken, Tokyo, Japan). The DPH was excited a t 360 nm and its fluorescence was detected using a sharp cut-off filter (L-39; Hoya Glass Works, Tokyo, Japan).*' Fluorescence polarization of DPH itself (Pfo) was measured in tetrahydrofuran. Microviscosity of liposomal lipids was calculated from the value of (Pfo - Pf)by Perrin's equation.18 Measurement of Lipid Leaching from the Excised Skin-The excised full-thickness hairless rat skin was set in two-chamber diffusion cells.19 The effective diffusional area of the cell was 0.89 cm2, and the volume of each chamber was 2 mL. Each enhancer solution (2 mL) was added to the half cell in contact with the stratum corneum, and the same volume of distilled water was added to the other half cell in contact with the dermis. The experiment was performed at 37 "C. At appropriate times, 0.5 mL of sample was withdrawn from both half cells for analysis. The sample solution and solvent were added after sampling to keep the volumes in the half cells constant. For analysis of total sterols,m sterols in 0.5 mL ofeach sample were extracted twice with 0.5 mL of ch1oroform:methanol (2:l) mixed solvent. After evaporating the organic phase, 3 mL of glacial acetic acid was added to dissolve lipids and then 1 mL of Fecl, solution was added. The resulting solution was colorimetrically determined at 550 nm by a spectrophotometer (model 100-60; Hitachi, Tokyo, Japan). The amount of sterols was normalized to that of cholesterol. Experiments with the Stratum Corneum Sheet-Stratum corneum sheets were prepared by the following method.21 After separating the epidermis from whole excised skin, by incubating the whole skin in warm distilled water (60 "C) for 30 8 , the resulting epidermis waa placed with its epidermis side down on filter paper saturated with 1% trypsin:pH 7.4 PBS at 37 "C for 4 h. The stratum comeum was then carefully peeled from the epidermis with light vortexing in distilled water and covered with the trypsin solution for 1 h. Remaining nucleated cells were removed from the stratum comeum. Resulting stratum comeum samples were dried under reduced pressure between two filter papers and kept in a desiccator under reduced pressure until use. For.measurement of free SH mourn leached from the stratum _ corneum, the stratum comeum sh&t w-as immersed in each enhancer solution at 37 "C for 24 h. After several rinsings with distilled water, the treated stratum corneum sheet was dried under reduced pressure. Sulfuric acid solution [15% (v/v), 1 mL1 was then added to 20 mg of the treated sheet, and the mixture was heated at 105 2 1 "C for 2 h with continuous shaking. This was followed by centrifugation of the mixed eolution at 4 oc for 10min to collect the ~i~~ milliliters of buffer (1.0M, pH 8 ) and 0.5 mL of m B solution (pH 4) were added to 5 mL of the supernatant, and this mixture was
spectrophotometrically analyzed at 412 nm for SH groups.22 The amount of free SH groups was normalized to that of cysteine. The stratum corneum surface was analyzed by the attenuated total mflection (ATR) method using Fourier-transform IR spectroscopy (mR model JIR-100; Nihon Denshi, Tokyo, Japan) by contacting a &-prism Cell (incidental angle = 60") on the spectrometer. The influence of Azone was studied after immersing the stratum corneum in a 3% Amne aqueous emulsion containing 0.1% plysorbate XU. Measurement of Moisturizing Effect-In vitro keratin moistwizing and in vivo skin moisturizing by enhancers were evaluated as reported previously.23 Skin Preparation and I n Vitro Skin Permeation E x p e r i m e n t s Rats were fixed under anesthesia (sodium pentobarbital, 60 mg i.pJ and the abdominal skin was excised as reported previously." The excised full-thickness skin (-2 x 2 cm2) was set in a two-chamber diffusion cell as described above. The half cell in contact with the stratum comeum was filled with 2 mL of distilled water or 3% h o n e aqueous emulsion (containing 0.1% polysorbate 20 as a n emulsifier) containing either 0.3 M MeOH, BuOH, LiCl, NaC1, or KCl, or a trace of "C-OcOH, 3H-Glu, or 3H-Suc. The other half cell surfacing the dermis was filled with 2 mL of distilled water. Every 2 h until 10 h, 0.1 or 0.2 mL of solution was withdrawn from the receiver cell. The same volume of distilled water was added after sampling to keep the cell volume constant. Permeability was evaluated as the permeability coefficientwhich was calculated by the average flux for 2-10 hand the initial concentration of each compound in the donor cell. The same skin permeation experiments were carried out using skins pretreated with distilled water or ethanol (EtOH). The pretreatments were done using the same two-chamber diffusion cell aa follows. The excised fresh skin was set in the cells. Both chambers were filled with 2 mL of distilled water and maintained at 37 "C for 24 h in the case of the pretreatment with water. Ethanol (2 mL) was added to the stratum corneum side and distilled water to the dermis side in the case of pretreatment with EtOH. Analysis of Model CompoundsMeOH and BuOH were measured by gas chromatography (GC). Each sample was centrifuged and the supernatant (2 &) was iqjected into a GC with a flame-ionization detector (FID; model GC 6A; Shimadzu, Kyoto, Japan). The conditions were as follows: column, 3 mm x 100 cm glass column packed with Gaschropak 54 (porous polymer, 80-100 mesh; Gasukuro Kogyo, Tokyo, Japan); column temperatures of 130 and 190 "C for MeOH and BuOH, respectively; and detector and injection port temperatures of 160 and 220 "C for MeOH and BuOH, respectively. Analyses of OcOH, Glu, and SUCwere done by using their respective radiolabeled compounds. Whole samples or fractions (100 or 10 pL, respectively) were mixed with 7 mL of scintillation cocktail (composition: 4 g of PPO,0.4 g of POPOP, 1.0 L of toluene, 0.5 L of Triton X-100) and measured on a liquid scintillation counter (model LSC-700; Aloka, Tokyo, Japan). Analyses of inorganic cations (Li', Na', K') were done by a flame-ionization absorption analyzer (model AA-8200; Japan Jerrel Ash, Osaka, Japan). The conditions were as follows: wavelengths for Li', Na', and K' were 670.2,589.0, and 766.5 nm, respectively; and fuel and oxidant gases were C2H2 and air, respectively. The C1- was determined by an ion-electrode method using an Na-K-Cl analyzer (model ION-15OC; Joko Company, Tokyo, Japan).
fJJt
Results and Discussion Effect of Azone on the Components in the Stratum Corneum-The stratum corneum, the main barrier to skin permeation of drugs, is composed of dead epidermal cells derived from viable epidermal cells. The stratum comeum has a multilamellar structure in which keratinized, protein-rich intracellular layers and lipid-rich intercellular layers exist alternatelv.26 The effect of Azone may be divided into the effects on iipids and keratin in the stratum corneum and the effects on the whole comeurn. First, the effects of Azone on the lipids were considered. were used 88 a model lipid ~ i t h i n - c h o l e s ~ r oliposomes l membrane. The Of 6-CF entrapped in the liposomes was increased with increasing fluidity of liposomal lipids and '%breaking''of liposomes. Figure 1A shows the time course of 6-CF remaining in the liposomes, and Figure 1B shows a Journal of Pharmaceutical Sciences I 59 Vol. 81, No. 1, January 1992
A
B
100-
2.0
50
-
gb
m .r
E .C
a
c I .r
E
v
U E
c, m
El.0'
10:
u 0
be
c, W
m
5-
W
m
PI W
1'
0
30
0
60
90
I,o - ~
I
1o
1
Time ( min )
1o
-~
-~
Concentration ( I )
Flgurel-Effect of Azone and other enhancerson the release of 6-CF from liposomes. Each point represents the mean of at least three measurements. Key:(A)0.W1%Azone(0);0.0005%Azone(~);0.W1%IPM(O);0.W1%DMSO(O);(B)Azone(O);IPM(O);MP(A);DMSO(O);OcOH(A);DMAC (D);DP (0); DA (0).
relationship between the first-order release rate constant, derived from the slope in Figure 1A and the concentration of enhancer in the liposome suspension. No 6-CF release was observed without polysorbate and enhancers (data not shown). Azone markedly enhanced the release of 6-CF at a (10 pg/mL) compared with other concentration of enhancers. The effects of Azone-like compounds, DA and DP, were also high, and their release rate constants at lo-'% were larger than those of IPM, MP, DMSO, OcOH, and DMAC. Table I shows the effect of Azone and other enhancers on the microviscosity of the liposomal lipids. Azone markedly decreased the microviscosity at a concentration of and DP and DA also decreased the microviscosity. These effecta were higher than that of IPM. The effects of OcOH,MP, DMSO, and DMAC were negligible. It is suggested from the results shown in Figure 1and Table I that Azone a t low concentrations enhances the fluidity of lipids and that DP and DA, which are similar to Azone with respect to their chemical structures, also increase fluidity of lipids. These effectsare much higher than observed with other enhancers, such as DMSO, DMAC, OcOH, and IPM. However, it is possible that the effects of Azone on the lipid fluidity are due to its effect on lipid depletion from the liposomemembranes. In order to measure this possibility, the lipid depletion from excised, full-thickness hairless rat skin was measured. Figure 2 shows the effect of Azone and other Table I-Effect of Enhancers on the Mlcrovlscosky of the Llpld Bllayer of Llposomes
Concentration,
Microviscosity, poise.
YO
DA
DP Azone IPM OcOH MP DMAC DMSO
lo-*
1.15 1.14 1.00 0.41
1.15 1.15 1.03 0.49
1.16 1.15 1.05 0.64
1.26 1.22 1.31 0.93
1.17 1.13 1.24 1.07
1.28 1.19 1.33 1.14
1.27 1.18 1.30 1.21
1.28 1.25 1.22 1.17
Control value (without enhancer) was 1.31 poise; values are the means of three determinations. 60 I Journal of Pharmaceutical Sciences Vol. 81, No. 7, January 7992
+I-
E
a
P
50
W .r
c,
m
z
c
-
V
0
8
16 Time (
24 h
32
)
Flgum 2-Influence of Azone and other enhancers on the leachlng of sterols from skin. Each point represents the mean of at least three 0.1% polysorbate 20 (0);3%IPM (A); measurements. Key: water (0); 3% Azone (A);3% DA (0);neat EtOH (D).
enhancers on the time courses of the leaching of total sterols from the stratum corneum side of the excised skin. The rank order of sterol leaching was neat EtOH > 3%IPM > 3% DA > 3%Azone. The amount of leaching with 3%Azone containing 0.1% polysorbate 20 as a n emulsifier was similar to that of distilled water and 0.1% polysorbate 20. These results support the hypothesis that Azone directly affects the fluidity of lipids. Stratum corneum lipids may be fluidized by Azone in the same way as the liposomal lipids. In order to investigate the effect of Azone on the stratum corneum protein (keratin), the effecta of Azone and other enhancers on the leaching of free SH groups from the stratum corneum sheet were determined. Figure 3 summarizes the experimental results. The amount of leaching of free SH groups with Azone was not significantly different from those in distilled water and 0.1% polysorbate 20 (p > 0.05). On the other hand, SA-Na, S-318,and EtOH significantly enhanced
SH ( imole/g stratum corneum )
20
10
0
40
30
,
Water
0.1% Tween20 3% Azone
3% OA
t
3% IPM 3% 5-318 3% OcOH
3% 3% 3% 3% 3%
j
l
t
DMSO DMAC
MP
Urea SA-Na 100% EtOH
t
t
Flgure 3-Influence of hone and other enhancers on the amount of free SH groups in the stratum corneum. Each value represents the mean f SE for five measurements. Key: significant in comparison with distilled water (p < 0.05).
r)
the amount of free SH groups leached from the stratum corneum (p < 0.05), and DMSO significantly decreased the leaching of SH groups (p < 0.05). These results show that the effect of Azone on the stratum corneum protein is low and that h o n e does not denature the stratum corneum protein. The FTIR spectra as shown in Figure 4 give some informa-
I
I
Skin surface
B
( water treatment )
Skin surface
( Azone treatment )
W
L u
2 &
s I-
4000
n.
3000
Azone
2000
1500
1000
500
WAVENUMBER ( cm-l)
Flgure&Effect of hone on the FT-IR spectra of the stratum mrneum.
tion on physicochemical and physiological changes of the skin surface.26 When comparing the FTIR spectra of stratum comeurn without treatment of Azone with those with h o n e , the absorption spectra at 1740 cm-’, resulting from c--O stretching vibration, was found to be decreased, and the spectra a t 1640cm-’, called Amide I, was markedly increased by Azone treatment. There was no significant change in the absorption spectra a t 1550 cm-’ for Amide 11, which is dependent on C-N stretching and N-H bending vibrations. In addition, spectra between 1300 and 800 cm-’ were not changed much. The spectrum of h o n e alone between 1300
and 800 cm-’ showed a similar profile to that of the stratum comeurn treated with Azone. These FTIR data suggest that Azone was entrapped in the stratum corneum and that it might affect the stratum corneum lipid. The effect of Azone on the lipid may be larger than that on the keratin in the stratum corneum with regard to penetration enhancement. Figure 5A shows the effect of Azone on the water-holding capacity in keratin tablets, and Figure 5B shows the effect of poly(acry1ic acid) gel containing Azone or DMSO on the skin conductance using hairless rat skin. Addition of Azone to the keratin tablets increased the water-holding capacity compared with tablets without Azone. However, there was no significant difference between 3, 10, and 20% Azone tablets. The in vivo skin conductance was markedly increased by applying Azone gel, and the effect was much higher than that of plane gel and DMSO gel. These results suggest that h o n e might increase water content or the moisturizing effect on the skin surface (stratum corneum). However, it is unclear whether Azone affects the stratum corneum lipid, or keratin, or both. The effects of Azone on the components in the stratum corneum can be summarized as follows.First, Azone increases the fluidity of the stratum corneum lipid, thereby decreasing its microviscosity. (The effect of h o n e is not due to lipid depletion.) Second, an effect of Azone on the stratum corneum protein is not directly observed. Finally, one of the most important effects of Azone is to increase the water content in the stratum corneum. Effect of Azone on the Skin Permeability of Model Compounds-Table I1 shows permeability coefficients of model compounds across excised hairless rat skin with and without h o n e treatment. Data after pretreatments with water or EtOH are also shown in Table 11. In the case of alkanols (MeOH, BuOH, and OcOH), there are six groups in the experiments. Data for non-pretreatment with Azone treatment can be compared with those for water or EtOH pretreatment without Azone to distinguish the mechanisms of action. Further comparison of the data for continuous treatment (water or EtOH pretreatment with Azone treatment) with those for the pretreatment alone (without Azone) may help further understanding of the difference between the action of Azone and that of water or EtOH. The rank order of the skin permeability coefficient of alkanols in the case of non-pretreatment without Azone (control) was OcOH >> BuOH > MeOH. This order is the same as the order of their lipophilicities. The rank order among the permeability coefficients of MeOH and BuOH was Azone treatment (nonpretreatment with Azone) > EtOH pretreatment without Azone > water pretreatment without Azone. The values for non-pretreatment without Azone were lowest. On the other hand, the rank order of permeability coefficients was totally opposite for OcOH permeation. One reason why the effect of Azone on the skin permeation of OcOH was negligible may be that it could be trapped in the Azone phase and/or polysorbate 20 micelles in the donor vehicle, since OcOH is highly lipophilic, resulting in a decrease of the partition of OcOH from the vehicle to skin. On the basis of the results in which its permeability coefficient after pretreatment with water or EtOH was lower than that for non-pretreatment, Azone may hydrate the stratum corneum, and as a result, it may then further decrease the partition of OcOH into the “hydrated” stratum corneum. Azone might not markedly enhance the skin permeability of highly lipophilic compounds such as OcOH, especially at low concentrations of the lipophilic compounds. When comparing the continuous treatment (water or EtOH pretreatment with Azone treatment) with the pretreatment alone, the penneability coefficients for continuous treatments were higher (in Journal of Pharmaceutical Sciences I 61 Vol. 81, No. 1, Januaty 1992
B 150
1
10% Azone CP-aq-gel\
Azone CP-aq-gel without a d d i t i v e
2
6
4
8
- 24
0
Water Time ( s )
Time ( h )
Figure +Effect of Azone on the water-holdingcapacity in (A) keratin tablet (in vitro) and (B) rat skin (in vivo). Each point represents the mean of 3% Azone three measurements. Key: (A) 10% h o n e (A); 20% Azone (0);3% Azone (0);without additive (0);(B) 10% Azone CP-aq-gel (0); CP-aq-gel (0);CP-aq-gel (m); CP-DMSO-gel (A); control (A). Table lCPermeabiiity Coefflclents of Several Compounds
Compounde Alkanol MeOh - Azone f Azone BuOH - Azone + Azone &OH - Azone + Azone Saccharide Glu - h o n e + Azone SUC- Azone + Azone Inorganic Ion (UCI) Li' - Azone + Azone CI- - Azone + Azone Inorganic ion (NaCI) Na' - Azone + Azone CI- - Azone + Azone Inorganic ion (KCI) K' - Azone + Azone CI- - Azone + Azone
Permeability coefficient, cmlsb Non-pretreatment
Water Pretreatment
EtOH Pretreatment
0.98 f 0.27 11.80 ? 0.60 (12.0) 2.35 2 0.32 13.90 ? 1.70 (5.9) 12.70 ? 2.30 0.93 0.09 (0.07)
1.13 f 0.16 (1.2) 17.80 2 2.30 (18.2) 4.34 0.58 (1.9) 11.10 2 0.60 (4.7) 7.49 f 2.18 (0.60) 0.42 f 0.06 (0.03)
5.81 f 0.27 (5.9) 16.70 f 1.10 (17.0) 8.78 2 0.83 (3.7) 12.50 2 0.30 (5.3) 7.79 f 2.31 (0.61) 0.70 2 0.14 (0.06)
2.52 x lo-' 2.20 x 10-8 (87.3) 4.38 x 10-6 (14.0) 6.15 X
2.28 x 10-6 (90.4) 5.97 x (23.7) 4.32 X lo-' (98.6) 2.54 x 10-5 (580)
3.22 x lo-' (128) 9.19 1.66 x 10-'(365) lo-' (37.9)
1.31 x 10-7 (6.6) 8.65 x 1.10 x 10-7 5.09 x (4.6)
2.39 x 10-7 (1.8) 1.69 x 10-6(1.3) 1.57 X 10-7(1.4) 6.34 x (5.8)
2.00 x 10-'(15.3) 1.3 x 10-5 (1.5) 9.32 x 10-7 (8.5) 8.50 X 10-'(117)
1.93 x 2.95 X 1.32 x 3.61 X
10-'(7.9) 10-6(12.1) lo-' (20.4) lo-' (55.7)
5.75 x (23.6) 8.92 x lo-' (36.6) 3.47 x 10-6 (53.5) 6.93 x 10-'(107)
1.19 x 10-7 (3.2) 5.57 x 10-7 (15.1) 1.43 x 10-7 (2.5) (5.6) 3.17 X
1.72 x lo-' (46.5) 7.11 x 10-'(192) 7.86 X (13.8) (113) 6.50 X
*
*
2.44 x 3.98 x 6.48 x 4.54 x
10-7 10-7 (1.6) lo-' 10-7 (7.0)
3.70 x lo-' 2.19 x 10-7 (5.9) 5.71 X lo-' 3.04 x 10-7 (5.3)
a Key: (- Azone) without Azone; (+ Azone) with Azone. Mean 2 SE (n = 3) x ratios against the control value (non-pretreatment without Azone).
the case of MeOH, 15 times for water pretreatment and three times for EtOH pretreatment; in the case of BuOH, about two times for both cases). Among the Azone groups, there were little differences with and without pretreatments with water or EtOH. Glucose (Glu) and sucrose (Suc) were selected as watersoluble, nonionic model compounds. Experiments similar to those with alkanols were performed. The permeability coef62 I Journal of Pharmaceutical Sciences Vol. 81, No. 1, January 1992
2.65 x
(60.5)
or mean in duplicate; the values in parenthesesare enhancing
ficient of Glu in the case of non-pretreatment without Azone was similar to that of Suc. A dependency of molecular size was not to be expected. The rank order of Glu permeability was EtOH pretreatment without Azone > water pretreatment without Azone > Azone treatment (non-pretreatment) > non-treatment without Azone. Although an Azone effect was found, it was lower than those for water and EtOH pretreatments. Since these saccharides are highly water-soluble and
nonionic compounds, lipid depletion and hydration of the stratum corneum would increase their skin permeabilities. The effects of Azone and pretreatments with water and EtOH on the skin permeabilities of Li' and C1- (LiCl), Na' and C1- (NaCl),and K' and C1- (KCl),selected as model ions, were measured. The rank order of permeability coefficientsof cations for nonpretreatment without Azone was Li' = Na' > K' . The permeabilities of the corresponding counteranion, C1-, had an order of LiCl > NaCl = KCl. The permeability coefficient or C1- was not the same as the corresponding countercation, but the rank order was the same as that of the cations. The rank orders of skin permeabilities of the cations and corresponding C1- were almost the same and independent of the kinds of pretreatment and treatment. It became clear from these results that a cation and a counteranion did not permeate through the skin as a 1:l complex, but that one ion might affect the skin permeation of the counterion. The skin permeabilities of Li', Na', and K' with Azone (non-pretreatment) were -7, 2, and 6 times higher, respectively, than that of the control, whereas the effect of EtOH pretreatment without Azone on their skin permeations was -15-,23-, and 46-fold, respectively. The effects of Azone on the skin permeation of ions might be lower than its effect on the permeation of un-ionized, water-soluble compounds such MeOH and sugars. The difference may originate from differences in charge of penetrants. The effects of Azone on the skin permeations of several model compounds may be summarized as follows. Azone effectively altered the skin permeation of MeOH, BuOH, Glu, and SUC. The effects were comparable or a little higher than the effects of EtOH pretreatment. In contrast, the permeability of OcOH was decreased by Azone, and the permeabilities of inorganic ions such as Na' and K' were increased less by the addition of Azone than by EtOH pretreatment. Azone has been found to be effective for increasing the skin permeation of a wide range of substances.13 The present experiments confirm this finding. However, it became clear that there are several compounds for which the effects of Azone are not marked. Permeation Routes of Drugs and the Effect of AzonPermeation routes of drugs through the skin may be classified as follows: transcellular route, consisting of keratinized, cornified cells; intercellular route, where lipids fill up the space between the cornified cells; and transappendageal routes, such as hair follicles, sweat ducts, and sebaceous glands. All drugs will not permeate evenly through the three routes; there will be predominant routes for each compound. For example, lipid-soluble drugs would penetrate by the lipid-rich route, whereas water-soluble drugs might diffuse through water-rich domains. However, since the cytochemical or morphological meanings are not fixed for the lipid- and water-rich domains, it is difficult to relate a site of action of Azone to any of the actual permeation routes discussed above. We therefore proposed three cytuchemically meaningless permeation routes; that is, lipid-rich route, water-rich route, and pore shunt route in the skin barrier (Figure 6A). The results of the present experiments indicate that MeOH and BuOH would permeate through all three routes we have proposed; OcOH permeates mainly through the lipid-rich route, Glu and SUCthrough the water-rich and shunt routes, and inorganic ions mainly through the shunt route. By using this concept for permeation routes, we can compare the mechanism of the enhancing effect of Azone on the skin permeation of drugs with those of pretreatments of EtOH and water (discussed below). In the case of EtOH pretreatment, extraction of a lot of lipids by EtOH from the skin surface decreases the permeation route of lipid-soluble compounds and enlarges the pore route, resulting in enhanced skin permeation of water-soluble
. C
D
Figure +Idealized model for permeation routes of several compounds through skin membrane: (A) Non-treatment;(B)EtOH pretreatment; (C) water pretreatment; (D) hone treatment.
compounds, such as MeOH, BuOH, Glu, SUC,and inorganic ions (Li', Na+, K', and Cl-1, whereas the skin permeation of highly lipophilic compounds, such as OcOH, is suppressed (Figure 6B). In the case of water pretreatment, the skin is moisturized and hydrated by water. As a result, the permeation route of water-soluble compounds and pore routes are enriched and the skin permeation of sugars, inorganic ions, MeOH, and BuOH is increased, whereas that of OcOH is decreased due to a decrease in the partitioning of OcOH into the skin (Figure 6 0 . These model compounds may be replaced by therapeutically potent drugs which have similar solubilities in water and organic solvents and similar oil-water partition coefficients. On the other hand, Azone not only fluidizes the skin lipids, but also hydrates the skin. This results in enhancement of the skin permeation of water-soluble compounds and drugs. The permeation coefficient of OcOH was decreased because of its decrease in partition into the "hydrated" skin, as discussed above. The effect of Azone on the transport through pore routes may be ignored a s indicated from the results that its enhancing effect on the transport of inorganic ions was not high (Figure 6D).
Conclusions Azone greatly enhances the skin permeation of many compounds which are thought to permeate through the hydrophilic route of skin, but it has little effect on the skin permeation of other compounds which might permeate through the hydrophobic route or pore route. Azone is not a general skin penetration enhancer for all drugs and comJournal of Phannaceutical Sciences I 63 Vol. 81, No. 1, January 1992
pounds, and its effect is dependent on penetrant lipophilicity and hydrophilicity.
References and Notes 1. Barry, B. W. In Dermntolo ical Formulations: Percutaneous Absorption; Marcel Dekker: 8 e w York, 1983;pp 160-172. 2. Waranis, P.;Sloan, K. B. J.Pharm. Sci. 1987, 76,587-595. 3. Kondo, S.; Sugimoto, I. J . Pharmacobi+Dyn. 1987,10,587-594. 4. Okabe, K.; Yamaguchi, H.; Kawai, Y. J.ControlZedRelease 1986, 4, 79-85. 5. Brucks, R.;Nanavaty, M.; Jung, D.; Siege], F. Pharm. Res. 1989, 6, 679-701. 6. Sasaki H.; Kubota Y:.Tomita T: Nakamura, J: Shibasaki J.; Ohya, f.; Ashlzawa 'f.;Idiyasato, k. *akuzaigaku 19$6,46,259-h65. 7. Friend, D.; Catz, P.; Heller, J.; Reid, J.; Baker, R. J. Controlled Release 1988,7,243-250. 8. Tsuzuki, N.;Wong, 0.; Higuchi, T.Znt.J.P h m . 1988,46,19-23. 9. Southwell, D.; Barry, B. W. J . Invest. Dermatol. 1983,80,507517. 10. Wong,.O.; Tsuzuki, N.; N p m , B.; Kuehnhoff, J.; Itoh, T.; Masaki, K.; Huntington, J.; onishi, R.; Rytting, J. H.; Higuchi, T. Znt. J . Pharm. 1989,52,191-201. 11. Kurihara-Ber strom, T.; F1 nn, G. L.; Higuchi, W. I. J. Invest. Dermatol. 19&,89, 274-2d. 12. Ashton, P.;Hadgraft, J.;Walkrs, K. A. Pharm. Acta Helv. 1986,
64 I Journal of Pharmaceutical Sciences Vol. 81, No. 1, January 1992
61,228-235. 13. Stoughton, R. B.; McClure, W. 0.Drug Dev. Ind. Pharm. 1983,9, 725-744. 14. Sugibayashi, K.; Morimoto, Y.; Hosoya, K.; Higuchi, W. I. J. Pharm. Pharmacol. 1985,37,576-580. 15. Kinsky, S. C.; Haxby, J.; Kinsky, C. B.; Demel, R. A.; van Deenen, L. L. M. Biochem. Biophys. Acta 1968,152, 174-185. 16. Mizuno, M.; Yoshida, M.; Hashida, M.; Kimura, T.; Sezaki, H. Chem. Pharm. Bull. 1985,33,2932-2940. 17. O'Connor, C. T.;Wallace, R. G.; Iwamoto, K.; Taguchi, T.; Sunamoto, J.Biochim. Biophys. Acta 1985,817,95-102. 18. Shinitzky, M.; Barenholz, Y. J. Bwl.Chem. 1974,249, 26522657. 19. Yu, C. D.; Fox, J. L.; Ho, N. F. H.; Higuchi, W. I. J.Pharm. Sci. 1979.68, 1341-1346. 20. Naito, H. K. In Lzpid Research Methodology; Story,J. A., Ed.; Alan R. Liss: New York, 1984. 21. Knuteon, K.; Potts, R. 0.; Guzek, D. B.; Golden, D. M.; McKie, J. E.; Lambert,W. J.; Higuchi, W. I. J. ControlledR e h e 1985,2,67-87. 22. Ellman, G . L. Arch. Biochern. BioDhvs. 1958. 74.433-450. 23. Su 'bayashi, K.; Nemoto, M.; Moiimoto, Y. Chem. Pharm. Bull. 19& 36, 1519-1528. 24. Morimoto, Y.; Sugibayashi, K.; Hosoya, K.; Higuchi, W. I. Znt. J . Pharm. 1986,32,31-38. 25. Barker, N.;Hadgraft, J. Znt. J . Pharm. 1981,8,193-202. 26. Scheulein, R.J.; Blank, I. H. Physwl. Rev. 1971.51, 703-747.
Several penetration enhancers,' prodrugs,2 superfluous vehicles,s iontophoresis,4 phonophoresis,6 and thermophoresis6 have been used as methods to increase the skin permeation of drugs. Among them, penetration enhancers are one of the most convenient methods and show relatively high effects. Ideal enhancers have no irritancy and toxicity to the skin and whole body, together with high enhancing effects. Enhancers themselves should be physicochemically stable and should not have pharmacological actions, and preferably should not have smell, color, or taste. Short- and long-chain alcohols,7.~ pyrrolidones,g urea and its derivatives,'O dimethyl sulfoxide and its relatives,*' and several surfactants12 have been screened for penetration enhancement effects. Most of them, however, did not appear on the market due to high adverse effects and low efficacy. Lack of understanding of the mechanism of their penetration-enhancing effects may be a reason 58 I Journal of Pharmaceutical Sciences Vol. 81, No. 1, January 1992
retarding their approval a s ingredients in pharmaceutical and cosmetic preparations. Laurocapram (1-dodecylazacycloheptan-2-one,Azone) is a nonirritant on the skin and can be used for a broad spectrum of drugs.13 We have reported that Azone affects mainly the skin barrier, the stratum corneum.14 However, some of the effects of Azone on the physicochemical properties of drugs, vehicles, and skin [i.e., partition coefficients between n-octano1 and water of drugs and vehicles (solvents)] are still obscure. The penetration-enhancing mechanism of Azone is also unclear. Study of the mechanism of penetration enhancers is very important, not only to discover suitable enhancers, but also to put many of them on the market. In the present study, the effects of Azone on the skin barrier components and on the skin permeation of different model compounds were compared with those of other penetration enhancers. In addition, the mechanism of the penetrationenhancing effect of Azone is discussed in relation to the experimental data and the hypothesis of penetration routes of drugs through the skin barrier.
Experlmental Section Materials and Animals-Methanol (MeOH), n-butanol (BuOH), lithium chloride (LiCl), sodium chloride (NaCl), and potassium chloride (KCl) were of JP grade. The following radiolabeled compounds were used: [l-"C]-n-octanol (W-OcOH; specific activit 5 mCi/mmol, ICN Radiochemicals, Irvine, CA), D-[5,6-3Hlglucose (l HGlu; specific activity 66.9 Ci/mmol, New England Nuclear Research, Boston, MA), and [l-3H(N)]-sucrose (3H-Suc; specific activity 0.01 Ci/mmol, New England Nuclear Research). Laurocapram (1dodecylazacycloheptan-2-one,Azone) was a kind girt from NelsonSumisho Company (Tokyo,Japan). Polysorbate 20 was obtained from Wako Pure Chemical Industries (Osaka, Japan). Trypein (Type 11, crude) and L-cysteine (free base) were supplied by Sigma Chemical Company (St.Louis, MO). Liposome compositions were as follows:egg lecithin and cholesterol (Wako Pure Chemical Industries); 5,6carboxyfluorescein (6-CF; Eastman Kodak Company, New York, NY) and 1,6-diphenylhexatriene(DPH; Aldrich Chemical Company, Milwaukee, WI);and 5,6-dithiobis-2-nitrobemicacid (DTNB) and polyoxyethylene (10) octyl phenyl ether (Triton X-100; Wako Pure Chemical Industries). Other reagents were of reagent grade and were used without further purification. Enhancers used are isopropyl myristate (IPM; Tokyo Kasei, Tokyo, Japan), dodecyl acetamide (DA; Nissan Chemical Industries, Tokyo, Japan), dodecyl pyrrolidone ( D P Niiwan Chemical Industries), dimethyl sulfoxide (DMSO; Wako Pure Chemical Industries), methyl pyrmlidone (Mp, Tokyo Kasei), Sefsol-318 (S-318; Nikko Chemical Company, Tokyo, Japan), sodium salicylate (SA-Na; Wako Pure Chemical Industries), and urea (Wako Pure Chemical Industries). When making Azone, IPM, DA, or DP solutions, polysorbate 20 was used as a n emulsifier, since these are not miscible in water. DMSO, DMAC, MP, 5318, SA-Na, and urea were dissolved directly in distilled water. The concentrations of each enhancer and polysorbate 20 in water were 3.0 and 0.196, respectively.*4 Male WBNkob hairless rats weighing -150 g were used in all animal experiments. OO22-3~9/92/0100-0058$02..50/0 0 1992, American Pharmaceutical Association
Experiments with LiposomegLiposomes were prepared by a thin-layer method.16 Lecithin (32 q o l ) and cholesterol (4 q o l ) were dissolved in 2 mL of chloroform in a 30-mL round-bottomed flask, and the solvent was evaporated to form a thin layer of lipid. After the remaining solvent was dried under reduced pressure, 5 mL of20 mM 6-CF in pH 7.4 phosphate-buffered saline (PBS) or 1mL of0.28 DPH in tetrahydrofuran was added to the flask, and the mixture aeitated to make multilamellar vehicles. The resulting suspension wis ultrasonicated with a probe tip (model G50022-4 Nihonseiki Seisakusho, Tokyo, Japan) a t 100 W for 2.5 min to make smaller liposomes. The liposomes were fractionated and concentrated by ultracentrifugation (40 000 rpm, 4 "C, 30 min). Rinsing with PBS (pH 7.4) was repeated twice and, finally, 2 mL of original liposome suspension was obtained. A 1 : l O dilution of liposome suspension in PBS (0.3 mL) containing 6-CF was mixed with each enhancer solution (2.7 mL) in a thermoregulated cuvette cell (1 x 1 x 4 cm3) for a fluorescence spectrophotometer, and the release of 6-CF from liposomes was followed at 37 T.16 Increase in fluorescence intensity (excitation and emission wavelengths of 490 and 550 nm, respectively) was monitored by a fluorescence spectrophotometer (model 650-10s; Hitachi, Tokyo, Japan). At the end of each experiment, 0.1 mL of 10% (vlv) Triton X-100 was added for complete release of 6-CF from liposomes, and the resulting fluorescence intensity was assumed to represent the initial content of 6-CF in the liposomes. A 1 : l O dilution of liposome suspension (0.3 mL) containing DPH was mixed with each enhancer solution (2.7 mL) in a thermoregulated cuvette cell (1 x 1 x 4 cm3)which contained a magnetic stirrer bar. Ten minutes after stirring, steady-statefluorescence polarization (PDof DPH intercalated in liposomal bilayers was measured on a fluorescence polarization spectrophotometer (model FS-501s; Union Giken, Tokyo, Japan). The DPH was excited a t 360 nm and its fluorescence was detected using a sharp cut-off filter (L-39; Hoya Glass Works, Tokyo, Japan).*' Fluorescence polarization of DPH itself (Pfo) was measured in tetrahydrofuran. Microviscosity of liposomal lipids was calculated from the value of (Pfo - Pf)by Perrin's equation.18 Measurement of Lipid Leaching from the Excised Skin-The excised full-thickness hairless rat skin was set in two-chamber diffusion cells.19 The effective diffusional area of the cell was 0.89 cm2, and the volume of each chamber was 2 mL. Each enhancer solution (2 mL) was added to the half cell in contact with the stratum corneum, and the same volume of distilled water was added to the other half cell in contact with the dermis. The experiment was performed at 37 "C. At appropriate times, 0.5 mL of sample was withdrawn from both half cells for analysis. The sample solution and solvent were added after sampling to keep the volumes in the half cells constant. For analysis of total sterols,m sterols in 0.5 mL ofeach sample were extracted twice with 0.5 mL of ch1oroform:methanol (2:l) mixed solvent. After evaporating the organic phase, 3 mL of glacial acetic acid was added to dissolve lipids and then 1 mL of Fecl, solution was added. The resulting solution was colorimetrically determined at 550 nm by a spectrophotometer (model 100-60; Hitachi, Tokyo, Japan). The amount of sterols was normalized to that of cholesterol. Experiments with the Stratum Corneum Sheet-Stratum corneum sheets were prepared by the following method.21 After separating the epidermis from whole excised skin, by incubating the whole skin in warm distilled water (60 "C) for 30 8 , the resulting epidermis waa placed with its epidermis side down on filter paper saturated with 1% trypsin:pH 7.4 PBS at 37 "C for 4 h. The stratum comeum was then carefully peeled from the epidermis with light vortexing in distilled water and covered with the trypsin solution for 1 h. Remaining nucleated cells were removed from the stratum comeum. Resulting stratum comeum samples were dried under reduced pressure between two filter papers and kept in a desiccator under reduced pressure until use. For.measurement of free SH mourn leached from the stratum _ corneum, the stratum comeum sh&t w-as immersed in each enhancer solution at 37 "C for 24 h. After several rinsings with distilled water, the treated stratum corneum sheet was dried under reduced pressure. Sulfuric acid solution [15% (v/v), 1 mL1 was then added to 20 mg of the treated sheet, and the mixture was heated at 105 2 1 "C for 2 h with continuous shaking. This was followed by centrifugation of the mixed eolution at 4 oc for 10min to collect the ~i~~ milliliters of buffer (1.0M, pH 8 ) and 0.5 mL of m B solution (pH 4) were added to 5 mL of the supernatant, and this mixture was
spectrophotometrically analyzed at 412 nm for SH groups.22 The amount of free SH groups was normalized to that of cysteine. The stratum corneum surface was analyzed by the attenuated total mflection (ATR) method using Fourier-transform IR spectroscopy (mR model JIR-100; Nihon Denshi, Tokyo, Japan) by contacting a &-prism Cell (incidental angle = 60") on the spectrometer. The influence of Azone was studied after immersing the stratum corneum in a 3% Amne aqueous emulsion containing 0.1% plysorbate XU. Measurement of Moisturizing Effect-In vitro keratin moistwizing and in vivo skin moisturizing by enhancers were evaluated as reported previously.23 Skin Preparation and I n Vitro Skin Permeation E x p e r i m e n t s Rats were fixed under anesthesia (sodium pentobarbital, 60 mg i.pJ and the abdominal skin was excised as reported previously." The excised full-thickness skin (-2 x 2 cm2) was set in a two-chamber diffusion cell as described above. The half cell in contact with the stratum comeum was filled with 2 mL of distilled water or 3% h o n e aqueous emulsion (containing 0.1% polysorbate 20 as a n emulsifier) containing either 0.3 M MeOH, BuOH, LiCl, NaC1, or KCl, or a trace of "C-OcOH, 3H-Glu, or 3H-Suc. The other half cell surfacing the dermis was filled with 2 mL of distilled water. Every 2 h until 10 h, 0.1 or 0.2 mL of solution was withdrawn from the receiver cell. The same volume of distilled water was added after sampling to keep the cell volume constant. Permeability was evaluated as the permeability coefficientwhich was calculated by the average flux for 2-10 hand the initial concentration of each compound in the donor cell. The same skin permeation experiments were carried out using skins pretreated with distilled water or ethanol (EtOH). The pretreatments were done using the same two-chamber diffusion cell aa follows. The excised fresh skin was set in the cells. Both chambers were filled with 2 mL of distilled water and maintained at 37 "C for 24 h in the case of the pretreatment with water. Ethanol (2 mL) was added to the stratum corneum side and distilled water to the dermis side in the case of pretreatment with EtOH. Analysis of Model CompoundsMeOH and BuOH were measured by gas chromatography (GC). Each sample was centrifuged and the supernatant (2 &) was iqjected into a GC with a flame-ionization detector (FID; model GC 6A; Shimadzu, Kyoto, Japan). The conditions were as follows: column, 3 mm x 100 cm glass column packed with Gaschropak 54 (porous polymer, 80-100 mesh; Gasukuro Kogyo, Tokyo, Japan); column temperatures of 130 and 190 "C for MeOH and BuOH, respectively; and detector and injection port temperatures of 160 and 220 "C for MeOH and BuOH, respectively. Analyses of OcOH, Glu, and SUCwere done by using their respective radiolabeled compounds. Whole samples or fractions (100 or 10 pL, respectively) were mixed with 7 mL of scintillation cocktail (composition: 4 g of PPO,0.4 g of POPOP, 1.0 L of toluene, 0.5 L of Triton X-100) and measured on a liquid scintillation counter (model LSC-700; Aloka, Tokyo, Japan). Analyses of inorganic cations (Li', Na', K') were done by a flame-ionization absorption analyzer (model AA-8200; Japan Jerrel Ash, Osaka, Japan). The conditions were as follows: wavelengths for Li', Na', and K' were 670.2,589.0, and 766.5 nm, respectively; and fuel and oxidant gases were C2H2 and air, respectively. The C1- was determined by an ion-electrode method using an Na-K-Cl analyzer (model ION-15OC; Joko Company, Tokyo, Japan).
fJJt
Results and Discussion Effect of Azone on the Components in the Stratum Corneum-The stratum corneum, the main barrier to skin permeation of drugs, is composed of dead epidermal cells derived from viable epidermal cells. The stratum comeum has a multilamellar structure in which keratinized, protein-rich intracellular layers and lipid-rich intercellular layers exist alternatelv.26 The effect of Azone may be divided into the effects on iipids and keratin in the stratum corneum and the effects on the whole comeurn. First, the effects of Azone on the lipids were considered. were used 88 a model lipid ~ i t h i n - c h o l e s ~ r oliposomes l membrane. The Of 6-CF entrapped in the liposomes was increased with increasing fluidity of liposomal lipids and '%breaking''of liposomes. Figure 1A shows the time course of 6-CF remaining in the liposomes, and Figure 1B shows a Journal of Pharmaceutical Sciences I 59 Vol. 81, No. 1, January 1992
A
B
100-
2.0
50
-
gb
m .r
E .C
a
c I .r
E
v
U E
c, m
El.0'
10:
u 0
be
c, W
m
5-
W
m
PI W
1'
0
30
0
60
90
I,o - ~
I
1o
1
Time ( min )
1o
-~
-~
Concentration ( I )
Flgurel-Effect of Azone and other enhancerson the release of 6-CF from liposomes. Each point represents the mean of at least three measurements. Key:(A)0.W1%Azone(0);0.0005%Azone(~);0.W1%IPM(O);0.W1%DMSO(O);(B)Azone(O);IPM(O);MP(A);DMSO(O);OcOH(A);DMAC (D);DP (0); DA (0).
relationship between the first-order release rate constant, derived from the slope in Figure 1A and the concentration of enhancer in the liposome suspension. No 6-CF release was observed without polysorbate and enhancers (data not shown). Azone markedly enhanced the release of 6-CF at a (10 pg/mL) compared with other concentration of enhancers. The effects of Azone-like compounds, DA and DP, were also high, and their release rate constants at lo-'% were larger than those of IPM, MP, DMSO, OcOH, and DMAC. Table I shows the effect of Azone and other enhancers on the microviscosity of the liposomal lipids. Azone markedly decreased the microviscosity at a concentration of and DP and DA also decreased the microviscosity. These effecta were higher than that of IPM. The effects of OcOH,MP, DMSO, and DMAC were negligible. It is suggested from the results shown in Figure 1and Table I that Azone a t low concentrations enhances the fluidity of lipids and that DP and DA, which are similar to Azone with respect to their chemical structures, also increase fluidity of lipids. These effectsare much higher than observed with other enhancers, such as DMSO, DMAC, OcOH, and IPM. However, it is possible that the effects of Azone on the lipid fluidity are due to its effect on lipid depletion from the liposomemembranes. In order to measure this possibility, the lipid depletion from excised, full-thickness hairless rat skin was measured. Figure 2 shows the effect of Azone and other Table I-Effect of Enhancers on the Mlcrovlscosky of the Llpld Bllayer of Llposomes
Concentration,
Microviscosity, poise.
YO
DA
DP Azone IPM OcOH MP DMAC DMSO
lo-*
1.15 1.14 1.00 0.41
1.15 1.15 1.03 0.49
1.16 1.15 1.05 0.64
1.26 1.22 1.31 0.93
1.17 1.13 1.24 1.07
1.28 1.19 1.33 1.14
1.27 1.18 1.30 1.21
1.28 1.25 1.22 1.17
Control value (without enhancer) was 1.31 poise; values are the means of three determinations. 60 I Journal of Pharmaceutical Sciences Vol. 81, No. 7, January 7992
+I-
E
a
P
50
W .r
c,
m
z
c
-
V
0
8
16 Time (
24 h
32
)
Flgum 2-Influence of Azone and other enhancers on the leachlng of sterols from skin. Each point represents the mean of at least three 0.1% polysorbate 20 (0);3%IPM (A); measurements. Key: water (0); 3% Azone (A);3% DA (0);neat EtOH (D).
enhancers on the time courses of the leaching of total sterols from the stratum corneum side of the excised skin. The rank order of sterol leaching was neat EtOH > 3%IPM > 3% DA > 3%Azone. The amount of leaching with 3%Azone containing 0.1% polysorbate 20 as a n emulsifier was similar to that of distilled water and 0.1% polysorbate 20. These results support the hypothesis that Azone directly affects the fluidity of lipids. Stratum corneum lipids may be fluidized by Azone in the same way as the liposomal lipids. In order to investigate the effect of Azone on the stratum corneum protein (keratin), the effecta of Azone and other enhancers on the leaching of free SH groups from the stratum corneum sheet were determined. Figure 3 summarizes the experimental results. The amount of leaching of free SH groups with Azone was not significantly different from those in distilled water and 0.1% polysorbate 20 (p > 0.05). On the other hand, SA-Na, S-318,and EtOH significantly enhanced
SH ( imole/g stratum corneum )
20
10
0
40
30
,
Water
0.1% Tween20 3% Azone
3% OA
t
3% IPM 3% 5-318 3% OcOH
3% 3% 3% 3% 3%
j
l
t
DMSO DMAC
MP
Urea SA-Na 100% EtOH
t
t
Flgure 3-Influence of hone and other enhancers on the amount of free SH groups in the stratum corneum. Each value represents the mean f SE for five measurements. Key: significant in comparison with distilled water (p < 0.05).
r)
the amount of free SH groups leached from the stratum corneum (p < 0.05), and DMSO significantly decreased the leaching of SH groups (p < 0.05). These results show that the effect of Azone on the stratum corneum protein is low and that h o n e does not denature the stratum corneum protein. The FTIR spectra as shown in Figure 4 give some informa-
I
I
Skin surface
B
( water treatment )
Skin surface
( Azone treatment )
W
L u
2 &
s I-
4000
n.
3000
Azone
2000
1500
1000
500
WAVENUMBER ( cm-l)
Flgure&Effect of hone on the FT-IR spectra of the stratum mrneum.
tion on physicochemical and physiological changes of the skin surface.26 When comparing the FTIR spectra of stratum comeurn without treatment of Azone with those with h o n e , the absorption spectra at 1740 cm-’, resulting from c--O stretching vibration, was found to be decreased, and the spectra a t 1640cm-’, called Amide I, was markedly increased by Azone treatment. There was no significant change in the absorption spectra a t 1550 cm-’ for Amide 11, which is dependent on C-N stretching and N-H bending vibrations. In addition, spectra between 1300 and 800 cm-’ were not changed much. The spectrum of h o n e alone between 1300
and 800 cm-’ showed a similar profile to that of the stratum comeurn treated with Azone. These FTIR data suggest that Azone was entrapped in the stratum corneum and that it might affect the stratum corneum lipid. The effect of Azone on the lipid may be larger than that on the keratin in the stratum corneum with regard to penetration enhancement. Figure 5A shows the effect of Azone on the water-holding capacity in keratin tablets, and Figure 5B shows the effect of poly(acry1ic acid) gel containing Azone or DMSO on the skin conductance using hairless rat skin. Addition of Azone to the keratin tablets increased the water-holding capacity compared with tablets without Azone. However, there was no significant difference between 3, 10, and 20% Azone tablets. The in vivo skin conductance was markedly increased by applying Azone gel, and the effect was much higher than that of plane gel and DMSO gel. These results suggest that h o n e might increase water content or the moisturizing effect on the skin surface (stratum corneum). However, it is unclear whether Azone affects the stratum corneum lipid, or keratin, or both. The effects of Azone on the components in the stratum corneum can be summarized as follows.First, Azone increases the fluidity of the stratum corneum lipid, thereby decreasing its microviscosity. (The effect of h o n e is not due to lipid depletion.) Second, an effect of Azone on the stratum corneum protein is not directly observed. Finally, one of the most important effects of Azone is to increase the water content in the stratum corneum. Effect of Azone on the Skin Permeability of Model Compounds-Table I1 shows permeability coefficients of model compounds across excised hairless rat skin with and without h o n e treatment. Data after pretreatments with water or EtOH are also shown in Table 11. In the case of alkanols (MeOH, BuOH, and OcOH), there are six groups in the experiments. Data for non-pretreatment with Azone treatment can be compared with those for water or EtOH pretreatment without Azone to distinguish the mechanisms of action. Further comparison of the data for continuous treatment (water or EtOH pretreatment with Azone treatment) with those for the pretreatment alone (without Azone) may help further understanding of the difference between the action of Azone and that of water or EtOH. The rank order of the skin permeability coefficient of alkanols in the case of non-pretreatment without Azone (control) was OcOH >> BuOH > MeOH. This order is the same as the order of their lipophilicities. The rank order among the permeability coefficients of MeOH and BuOH was Azone treatment (nonpretreatment with Azone) > EtOH pretreatment without Azone > water pretreatment without Azone. The values for non-pretreatment without Azone were lowest. On the other hand, the rank order of permeability coefficients was totally opposite for OcOH permeation. One reason why the effect of Azone on the skin permeation of OcOH was negligible may be that it could be trapped in the Azone phase and/or polysorbate 20 micelles in the donor vehicle, since OcOH is highly lipophilic, resulting in a decrease of the partition of OcOH from the vehicle to skin. On the basis of the results in which its permeability coefficient after pretreatment with water or EtOH was lower than that for non-pretreatment, Azone may hydrate the stratum corneum, and as a result, it may then further decrease the partition of OcOH into the “hydrated” stratum corneum. Azone might not markedly enhance the skin permeability of highly lipophilic compounds such as OcOH, especially at low concentrations of the lipophilic compounds. When comparing the continuous treatment (water or EtOH pretreatment with Azone treatment) with the pretreatment alone, the penneability coefficients for continuous treatments were higher (in Journal of Pharmaceutical Sciences I 61 Vol. 81, No. 1, Januaty 1992
B 150
1
10% Azone CP-aq-gel\
Azone CP-aq-gel without a d d i t i v e
2
6
4
8
- 24
0
Water Time ( s )
Time ( h )
Figure +Effect of Azone on the water-holdingcapacity in (A) keratin tablet (in vitro) and (B) rat skin (in vivo). Each point represents the mean of 3% Azone three measurements. Key: (A) 10% h o n e (A); 20% Azone (0);3% Azone (0);without additive (0);(B) 10% Azone CP-aq-gel (0); CP-aq-gel (0);CP-aq-gel (m); CP-DMSO-gel (A); control (A). Table lCPermeabiiity Coefflclents of Several Compounds
Compounde Alkanol MeOh - Azone f Azone BuOH - Azone + Azone &OH - Azone + Azone Saccharide Glu - h o n e + Azone SUC- Azone + Azone Inorganic Ion (UCI) Li' - Azone + Azone CI- - Azone + Azone Inorganic ion (NaCI) Na' - Azone + Azone CI- - Azone + Azone Inorganic ion (KCI) K' - Azone + Azone CI- - Azone + Azone
Permeability coefficient, cmlsb Non-pretreatment
Water Pretreatment
EtOH Pretreatment
0.98 f 0.27 11.80 ? 0.60 (12.0) 2.35 2 0.32 13.90 ? 1.70 (5.9) 12.70 ? 2.30 0.93 0.09 (0.07)
1.13 f 0.16 (1.2) 17.80 2 2.30 (18.2) 4.34 0.58 (1.9) 11.10 2 0.60 (4.7) 7.49 f 2.18 (0.60) 0.42 f 0.06 (0.03)
5.81 f 0.27 (5.9) 16.70 f 1.10 (17.0) 8.78 2 0.83 (3.7) 12.50 2 0.30 (5.3) 7.79 f 2.31 (0.61) 0.70 2 0.14 (0.06)
2.52 x lo-' 2.20 x 10-8 (87.3) 4.38 x 10-6 (14.0) 6.15 X
2.28 x 10-6 (90.4) 5.97 x (23.7) 4.32 X lo-' (98.6) 2.54 x 10-5 (580)
3.22 x lo-' (128) 9.19 1.66 x 10-'(365) lo-' (37.9)
1.31 x 10-7 (6.6) 8.65 x 1.10 x 10-7 5.09 x (4.6)
2.39 x 10-7 (1.8) 1.69 x 10-6(1.3) 1.57 X 10-7(1.4) 6.34 x (5.8)
2.00 x 10-'(15.3) 1.3 x 10-5 (1.5) 9.32 x 10-7 (8.5) 8.50 X 10-'(117)
1.93 x 2.95 X 1.32 x 3.61 X
10-'(7.9) 10-6(12.1) lo-' (20.4) lo-' (55.7)
5.75 x (23.6) 8.92 x lo-' (36.6) 3.47 x 10-6 (53.5) 6.93 x 10-'(107)
1.19 x 10-7 (3.2) 5.57 x 10-7 (15.1) 1.43 x 10-7 (2.5) (5.6) 3.17 X
1.72 x lo-' (46.5) 7.11 x 10-'(192) 7.86 X (13.8) (113) 6.50 X
*
*
2.44 x 3.98 x 6.48 x 4.54 x
10-7 10-7 (1.6) lo-' 10-7 (7.0)
3.70 x lo-' 2.19 x 10-7 (5.9) 5.71 X lo-' 3.04 x 10-7 (5.3)
a Key: (- Azone) without Azone; (+ Azone) with Azone. Mean 2 SE (n = 3) x ratios against the control value (non-pretreatment without Azone).
the case of MeOH, 15 times for water pretreatment and three times for EtOH pretreatment; in the case of BuOH, about two times for both cases). Among the Azone groups, there were little differences with and without pretreatments with water or EtOH. Glucose (Glu) and sucrose (Suc) were selected as watersoluble, nonionic model compounds. Experiments similar to those with alkanols were performed. The permeability coef62 I Journal of Pharmaceutical Sciences Vol. 81, No. 1, January 1992
2.65 x
(60.5)
or mean in duplicate; the values in parenthesesare enhancing
ficient of Glu in the case of non-pretreatment without Azone was similar to that of Suc. A dependency of molecular size was not to be expected. The rank order of Glu permeability was EtOH pretreatment without Azone > water pretreatment without Azone > Azone treatment (non-pretreatment) > non-treatment without Azone. Although an Azone effect was found, it was lower than those for water and EtOH pretreatments. Since these saccharides are highly water-soluble and
nonionic compounds, lipid depletion and hydration of the stratum corneum would increase their skin permeabilities. The effects of Azone and pretreatments with water and EtOH on the skin permeabilities of Li' and C1- (LiCl), Na' and C1- (NaCl),and K' and C1- (KCl),selected as model ions, were measured. The rank order of permeability coefficientsof cations for nonpretreatment without Azone was Li' = Na' > K' . The permeabilities of the corresponding counteranion, C1-, had an order of LiCl > NaCl = KCl. The permeability coefficient or C1- was not the same as the corresponding countercation, but the rank order was the same as that of the cations. The rank orders of skin permeabilities of the cations and corresponding C1- were almost the same and independent of the kinds of pretreatment and treatment. It became clear from these results that a cation and a counteranion did not permeate through the skin as a 1:l complex, but that one ion might affect the skin permeation of the counterion. The skin permeabilities of Li', Na', and K' with Azone (non-pretreatment) were -7, 2, and 6 times higher, respectively, than that of the control, whereas the effect of EtOH pretreatment without Azone on their skin permeations was -15-,23-, and 46-fold, respectively. The effects of Azone on the skin permeation of ions might be lower than its effect on the permeation of un-ionized, water-soluble compounds such MeOH and sugars. The difference may originate from differences in charge of penetrants. The effects of Azone on the skin permeations of several model compounds may be summarized as follows. Azone effectively altered the skin permeation of MeOH, BuOH, Glu, and SUC. The effects were comparable or a little higher than the effects of EtOH pretreatment. In contrast, the permeability of OcOH was decreased by Azone, and the permeabilities of inorganic ions such as Na' and K' were increased less by the addition of Azone than by EtOH pretreatment. Azone has been found to be effective for increasing the skin permeation of a wide range of substances.13 The present experiments confirm this finding. However, it became clear that there are several compounds for which the effects of Azone are not marked. Permeation Routes of Drugs and the Effect of AzonPermeation routes of drugs through the skin may be classified as follows: transcellular route, consisting of keratinized, cornified cells; intercellular route, where lipids fill up the space between the cornified cells; and transappendageal routes, such as hair follicles, sweat ducts, and sebaceous glands. All drugs will not permeate evenly through the three routes; there will be predominant routes for each compound. For example, lipid-soluble drugs would penetrate by the lipid-rich route, whereas water-soluble drugs might diffuse through water-rich domains. However, since the cytochemical or morphological meanings are not fixed for the lipid- and water-rich domains, it is difficult to relate a site of action of Azone to any of the actual permeation routes discussed above. We therefore proposed three cytuchemically meaningless permeation routes; that is, lipid-rich route, water-rich route, and pore shunt route in the skin barrier (Figure 6A). The results of the present experiments indicate that MeOH and BuOH would permeate through all three routes we have proposed; OcOH permeates mainly through the lipid-rich route, Glu and SUCthrough the water-rich and shunt routes, and inorganic ions mainly through the shunt route. By using this concept for permeation routes, we can compare the mechanism of the enhancing effect of Azone on the skin permeation of drugs with those of pretreatments of EtOH and water (discussed below). In the case of EtOH pretreatment, extraction of a lot of lipids by EtOH from the skin surface decreases the permeation route of lipid-soluble compounds and enlarges the pore route, resulting in enhanced skin permeation of water-soluble
. C
D
Figure +Idealized model for permeation routes of several compounds through skin membrane: (A) Non-treatment;(B)EtOH pretreatment; (C) water pretreatment; (D) hone treatment.
compounds, such as MeOH, BuOH, Glu, SUC,and inorganic ions (Li', Na+, K', and Cl-1, whereas the skin permeation of highly lipophilic compounds, such as OcOH, is suppressed (Figure 6B). In the case of water pretreatment, the skin is moisturized and hydrated by water. As a result, the permeation route of water-soluble compounds and pore routes are enriched and the skin permeation of sugars, inorganic ions, MeOH, and BuOH is increased, whereas that of OcOH is decreased due to a decrease in the partitioning of OcOH into the skin (Figure 6 0 . These model compounds may be replaced by therapeutically potent drugs which have similar solubilities in water and organic solvents and similar oil-water partition coefficients. On the other hand, Azone not only fluidizes the skin lipids, but also hydrates the skin. This results in enhancement of the skin permeation of water-soluble compounds and drugs. The permeation coefficient of OcOH was decreased because of its decrease in partition into the "hydrated" skin, as discussed above. The effect of Azone on the transport through pore routes may be ignored a s indicated from the results that its enhancing effect on the transport of inorganic ions was not high (Figure 6D).
Conclusions Azone greatly enhances the skin permeation of many compounds which are thought to permeate through the hydrophilic route of skin, but it has little effect on the skin permeation of other compounds which might permeate through the hydrophobic route or pore route. Azone is not a general skin penetration enhancer for all drugs and comJournal of Phannaceutical Sciences I 63 Vol. 81, No. 1, January 1992
pounds, and its effect is dependent on penetrant lipophilicity and hydrophilicity.
References and Notes 1. Barry, B. W. In Dermntolo ical Formulations: Percutaneous Absorption; Marcel Dekker: 8 e w York, 1983;pp 160-172. 2. Waranis, P.;Sloan, K. B. J.Pharm. Sci. 1987, 76,587-595. 3. Kondo, S.; Sugimoto, I. J . Pharmacobi+Dyn. 1987,10,587-594. 4. Okabe, K.; Yamaguchi, H.; Kawai, Y. J.ControlZedRelease 1986, 4, 79-85. 5. Brucks, R.;Nanavaty, M.; Jung, D.; Siege], F. Pharm. Res. 1989, 6, 679-701. 6. Sasaki H.; Kubota Y:.Tomita T: Nakamura, J: Shibasaki J.; Ohya, f.; Ashlzawa 'f.;Idiyasato, k. *akuzaigaku 19$6,46,259-h65. 7. Friend, D.; Catz, P.; Heller, J.; Reid, J.; Baker, R. J. Controlled Release 1988,7,243-250. 8. Tsuzuki, N.;Wong, 0.; Higuchi, T.Znt.J.P h m . 1988,46,19-23. 9. Southwell, D.; Barry, B. W. J . Invest. Dermatol. 1983,80,507517. 10. Wong,.O.; Tsuzuki, N.; N p m , B.; Kuehnhoff, J.; Itoh, T.; Masaki, K.; Huntington, J.; onishi, R.; Rytting, J. H.; Higuchi, T. Znt. J . Pharm. 1989,52,191-201. 11. Kurihara-Ber strom, T.; F1 nn, G. L.; Higuchi, W. I. J. Invest. Dermatol. 19&,89, 274-2d. 12. Ashton, P.;Hadgraft, J.;Walkrs, K. A. Pharm. Acta Helv. 1986,
64 I Journal of Pharmaceutical Sciences Vol. 81, No. 1, January 1992
61,228-235. 13. Stoughton, R. B.; McClure, W. 0.Drug Dev. Ind. Pharm. 1983,9, 725-744. 14. Sugibayashi, K.; Morimoto, Y.; Hosoya, K.; Higuchi, W. I. J. Pharm. Pharmacol. 1985,37,576-580. 15. Kinsky, S. C.; Haxby, J.; Kinsky, C. B.; Demel, R. A.; van Deenen, L. L. M. Biochem. Biophys. Acta 1968,152, 174-185. 16. Mizuno, M.; Yoshida, M.; Hashida, M.; Kimura, T.; Sezaki, H. Chem. Pharm. Bull. 1985,33,2932-2940. 17. O'Connor, C. T.;Wallace, R. G.; Iwamoto, K.; Taguchi, T.; Sunamoto, J.Biochim. Biophys. Acta 1985,817,95-102. 18. Shinitzky, M.; Barenholz, Y. J. Bwl.Chem. 1974,249, 26522657. 19. Yu, C. D.; Fox, J. L.; Ho, N. F. H.; Higuchi, W. I. J.Pharm. Sci. 1979.68, 1341-1346. 20. Naito, H. K. In Lzpid Research Methodology; Story,J. A., Ed.; Alan R. Liss: New York, 1984. 21. Knuteon, K.; Potts, R. 0.; Guzek, D. B.; Golden, D. M.; McKie, J. E.; Lambert,W. J.; Higuchi, W. I. J. ControlledR e h e 1985,2,67-87. 22. Ellman, G . L. Arch. Biochern. BioDhvs. 1958. 74.433-450. 23. Su 'bayashi, K.; Nemoto, M.; Moiimoto, Y. Chem. Pharm. Bull. 19& 36, 1519-1528. 24. Morimoto, Y.; Sugibayashi, K.; Hosoya, K.; Higuchi, W. I. Znt. J . Pharm. 1986,32,31-38. 25. Barker, N.;Hadgraft, J. Znt. J . Pharm. 1981,8,193-202. 26. Scheulein, R.J.; Blank, I. H. Physwl. Rev. 1971.51, 703-747.