Quantitative assessment of tissue retention, lipophilicity, ionic valence and convective transport of permeant as factors affecting iontophoretic enhancement

J. DRUG DEL. SCI. TECH., 16 (1) 91-98 2006

Quantitative assessment of tissue retention, lipophilicity, ionic valence and convective transport of permeant as factors affecting iontophoretic enhancement M. Altenbach, N. Schnyder, C. Zimmermann, G. Imanidis* Institute of Pharmaceutical Technology, University of Basel, Klingelbergstrasse 50, 4056 Basel, Switzerland *Correspondence: [email protected] Iontophoresis of the dipeptide tyrosine-phenylalanine (TyrPhe), the protected amino acid tyrosine-β-naphthylamide (Tyr-β-NA) and the glucose derivative benzyl-2-acetamido-2-deoxy-α-D-glucopyranoside (BAd-α-Glc) used as model compounds was investigated in human epidermis in vitro at pH 3 and pH 4.5 under constant voltage application, the purpose being to delineate the contribution of tissue retention, lipophilicity, ionic valence and convective transport of these compounds to iontophoretic enhancement in a quantitative fashion and ultimately gain an improved mechanistic understanding of iontophoresis. Retention of BAd-α-Glc in epidermal tissue during permeation was considerable and was reduced upon iontophoresis in the presence of extraneous tyrosine (Tyr) and phenylalanine (Phe) added to the donor solution, this evoking an increase of the apparent iontophoretic permeation. This suggests firstly, the occurrence of an interaction between BAd-α-Glc, Tyr and Phe impacting iontophoretic enhancement and secondly, the potential of using amino acids as adjuvants to modulate iontophoresis of selected compounds. This influence of tissue retention on iontophoretic permeation in the presence of Tyr and Phe was not observed for TyrPhe or Tyr-β-NA or when BAd-α-Glc, TyrPhe and Tyr-β-NA were used concomitantly rather than individually in the absence of Tyr and Phe. The effect of lipophilicity, ionic valence in the aqueous permeation pathway and convective transport by electroosmosis of the three permeants on iontophoresis was assessed simultaneously for all permeants by analyzing the experimental data using a theoretically derived model for iontophoretic enhancement that encompassed these factors. This model was based on an extension of the modified Nernst-Planck equation, whereas the difference of ionic valence between the aqueous domain of tissue and the bulk solution was evaluated on the basis of a pH shift due to the electrical double layer at the lipid/aqueous interface in the epidermis using the Poisson-Boltzmann equation. Using deduced parameter values characterizing the involved factors, a very good agreement between model calculated and experimental enhancement data was obtained. At pH 4.5, a very weak convective transport due to electroosmosis from cathode to anode was evident, indicating an isoelectric point of the epidermis slightly above 4.5. At pH 3, electroosmotic transport was approximately 10-fold stronger than at pH 4.5 which reflected a high positive surface charge density of the epidermis at pH 3 and a low one at pH 4.5. The pH in the aqueous permeation pathway, estimated to differ from that of the bulk in accordance with these charge densities, produced an ionic valence in this pathway dependent on pKa that accounted consistently for the flux due to the direct effect of the electric field on the ionic permeants. The ratio of lipid to aqueous pathway passive permeability coefficient was << 1, indicating a marginal role of the lipophilicity of these compounds for iontophoretic enhancement. Hence, the applied evaluation afforded a quantitative assessment of the effect of factors relevant for iontophoresis and provided a congruent understanding of the process. Key words: Iontophoresis – Transdermal permeation – Physicochemical model – Amino acids – Peptide – Tissue retention – Lipophilicity – Electroosmosis – Iontophoretic enhancement.

may permeate the epidermal membrane by an aqueous and a lipid pathway [20]. The relative importance of each pathway depends on the lipophilicity and the ionic state of the compound. The aqueous pathway alone is responsible for the conductance of electric current and for electroosmosis. Therefore, iontophoretic permeation of a compound and permeation enhancement depend on its lipophilicity and ionization. The latter also affects the magnitude of migration that takes place as a result of the direct effect of the electric field on an ion. Furthermore, ionization of weak electrolytes depends on the pH of the environment which for a charged membrane may be affected by the distribution of charges in the electrical double layer at interfaces within the membrane. In order to establish a formalized relationship between iontophoretic permeation enhancement and the above listed factors and make possible the evaluation of the effect of each of these factors in a quantitative fashion, a theoretical model was introduced [14, 21]. This model was based on the modified Nernst-Planck equation that included electroosmotic flow and was extended to take into account the ratio of lipid to aqueous pathway passive permeability of permeant and the ionization of the permeant in the aqueous pathway of the epidermis. The latter was considered to differ from that in bulk because of the electrical double layer at the aqueous/non-aqueous domain interface of the epidermis eliciting a pH shift in the aqueous domain that was estimated

Iontophoresis is an extensively investigated technology for enhancing and modulating topical and systemic transdermal delivery of drugs by an electric field. Using the electric field as an external physical stimulus whose strength and temporal pattern of application can easily be regulated, it was possible to control the onset and the rate of the delivery of drugs to the body as well as of the extraction of substances from the blood stream for the purpose of monitoring [1-8]. Mechanistically, next to the direct effect of the electric field driving ionic permeants through the epidermis, additional phenomena such as convective solvent flow due to electroosmosis and membrane alterations may influence the iontophoretic flux of charged and uncharged permeants. The direction of the electroosmotic flow may coincide with that of the current flow or be opposite to it, depending on the fixed charge carried by the membrane. The epidermis is a permselective membrane with an isoelectric point (pI) reportedly somewhat above 4 [9-11]. At physiological pH, epidermis is negatively charged and electroosmotic flow contributes to the current flow of an anodal delivery while cathodal delivery is retarded [12, 13]. At a pH close to the pI, a neutral membrane comes about so that no electroosmotic flow occurs. At a pH lower than the pI, the epidermis carries a positive charge and the electroosmotic flow is against the current flow of an anodal iontophoretic delivery [9, 18, 19]. Compounds 91

J. DRUG DEL. SCI. TECH., 16 (1) 91-98 2006

Quantitative assessment of tissue retention, lipophilicity, ionic valence and convective transport of permeant as factors affecting iontophoretic enhancement M. Altenbach, N. Schnyder, C. Zimmermann, G. Imanidis

using the Poisson-Boltzmann equation. The relationship between the iontophoretic enhancement under constant voltage application and all factors conceivably affecting enhancement that was yielded by the model was used earlier to analyze experimental iontophoresis data of a peptide and an amphoteric weak electrolyte drug in order to verify the validity of the underlying assumptions and delineate the contribution of the different factors acting simultaneously in the process [14, 21]. Those analyses provided quantitatively consistent results about the effect of the involved factors and a good agreement between calculation and experiment. In a previous manuscript [15], anodal iontophoresis of the glucose derivative benzyl-2-acetamido-2-deoxy-α-D-glucopyranoside (BAdα-Glc) used as an electroosmosis marker at pH 3 and pH 4.5 caused a slight flux retardation of this compound indicating electroosmotic flow occurring in the cathode-to-anode direction in accordance with the accepted view of positive fixed epidermis charges prevailing in this pH range. Significant iontophoretic enhancement of BAd-α-Glc at the same pH values was found in the presence of the amino acids tyrosine (Tyr) and phenylalanine (Phe) produced by cutaneous metabolism of the model dipeptide tyrosine-phenylalanine (TyrPhe) that was used concomitantly in the studies. This correlated with a decrease upon iontophoresis of the amount of BAd-α-Glc retained in the epidermis. Therefore, an interaction between BAd-α-Glc and the metabolic products of TyrPhe impacting tissue retention and apparent permeation of BAd-α-Glc upon iontophoresis was suggested [15]. In the present work, the effect of a priori addition of Tyr and Phe to the drug solution on epidermis retention and iontophoretic permeation of three model permeants, i.e., the glucose derivative benzyl-2-acetamido-2-deoxy-α-D-glucopyranoside (BAd-α-Glc), the dipeptide tyrosine-phenylalanine (TyrPhe) and the protected amino acid tyrosine-β-naphthylamide (Tyr-β-NA) is investigated, the purpose being to test the universality of the described effect [15] and understand its origin. TyrPhe was studied at 4°C in order to abolish its enzymatic degradation in the tissue [15] and thus establish controlled experimental conditions. Furthermore, the physicochemical model of iontophoretic enhancement developed previously [14, 21] is used for simultaneously analyzing experimental data of all three model permeants in order to quantitatively delineate the effect of permeant lipophilicity, ionic valence in the aqueous membrane domain and convective transport by electroosmosis to the measured enhancement. The goal of applying the model to the iontophoresis of more than one compounds simultaneously is to obtain results about the role of the involved phenomena and the influence of permeant properties with rather general validity. The model permeants were chosen such as to span a wide range of lipophilicity (70-fold) and ionic valence but had all approximately the same molecular size at this stage of the work to ensure comparable diffusivity and convective transport in the tortuous aqueous membrane domain. Permeation experiments were performed with heat separated human epidermis at pH 3 and pH 4.5. At these pH values the permeants carried according to their pKa in part positive (or zero) charge allowing the use of anodal iontophoresis. Also, this pH range contained the isoelectric point of human epidermis, providing the possibility to apply the theoretical model when opposite directions of the electroosmotic flow took hold. This represents the typical situation encountered in the iontophoresis of peptides and drugs that are organic bases. Finally, constant voltage iontophoresis was used because under these conditions an analytical solution of the Nernst-Planck equation can be obtained and iontophoretic transport of concomitant ionic spieces including metabolic products and buffer salts can be treated independently of each other. The Nernst-Planck equation in combination with the Poisson-Boltzmann equation provided the theoretical framework for this analysis. It is recognized that constant voltage conditions are appropriate only for a fundamental study of the process whereas constant current application is the method of choice for controlling delivery [24].

I. MATERIALS AND METHODS 1. Materials

Benzyl-2-acetamido-2-deoxy-α-D-glucopyranoside (BAd-α-Glc), a glucose derivative with a molecular weight of 311.3, was purchased from Toronto Research Chemicals, North York, Canada. BAd-α-Glc is unionized at pH 3 and pH 4.5 and was used at a concentration of 13 mM in the donor solution. Tyrosine-β-naphthylamide (Tyr-β-NA), an amino acid protected at the C-terminus with a molecular weight of 306.4, and tyrosine-phenylalanine (TyrPhe), a dipeptide with a molecular weight of 328.4, were purchased from Bachem AG (Bubendorf, Switzerland). A pKa of 3.5 for TyrPhe and 4.78 for Tyr-β-NA was determined by potentiometric titration in the range that was of interest considering the acidic pH values of the permeation experiments. No other pKa was found < 7. The concentration of TyrPhe and of Tyrβ-NA in the donor solution was 3 mM at pH 3 and 2 mM at pH 4.5. The amino acids tyrosine (Tyr) with a molecular weight of 181.2 (pKa 2.2, 9.1 and 10.1) and phenylalanine (Phe) with a molecular weight of 165.2 (pKa 2.2 and 9.2) were purchased from Fluka BioChemika (Buchs, Switzerland) and Sigma Chemical Co. (St. Louis, MO, USA), respectively. The concentration of each amino acid that was added to the donor solution was 3 mM at pH 3 and 2 mM at pH 4.5. Universal buffer was used at pH 3 and pH 4.5. It was composed of citric and phosphoric acid (each 6.67 mM) and boric acid (11.5 mM), which were dissolved in double distilled water and titrated to the desired pH with sodium hydroxide. The osmolarity in the donor and the receiver solutions was set to 300 mOsmol with sodium chloride. All chemicals were of analytical reagent grade. Human cadaver skin was obtained from the Department of Pathology, University Hospital, Basel. Heat-separated epidermis with an average thickness of 43 µm (SD 5 µm, n = 16) was used in the permeation experiments. Heat separation, mounting in the diffusion cells, a gravitational leaking test and an electrical resistance test of the barrier properties of the epidermis were carried out as described in [14]. Permeation experiments were carried out in custom made twochamber, side-by-side glass diffusion cells with a diffusion area of approximately 2 cm2 that were connected to a current source with two working Ag/AgCl electrodes and two reference Ag/AgCl electrodes, the latter reaching the membrane by buffer-filled glass capillaries. The anode was always placed in the donor and the cathode in the receiver compartment. Universal buffer was used in both chambers of the cells. The current source (built in the Department of Physics, University of Basel) supplied direct current and was operated at constant voltage for iontophoresis. The flow of electric current was measured and recorded continually on a disk with a Digital Chart Recorder (DCR 520, W + W Instruments AG, Basel, Switzerland). Further details of the experimental set up can be found elsewhere [14, 21].

2. Protocol of permeation experiments

In a first stage, the baseline passive permeability was measured for 44 h to allow accurate determination of the flux of these slowly permeating compounds. This was followed by the iontophoretic stage during which a constant voltage of 250 mV was applied for 3 h across the epidermal membrane. This was shown previously to be appropriate for determining the steady state iontophoretic permeability [14, 21]. Subsequently, the post-iontophoretic passive permeability was measured. Samples of 0.2 ml were withdrawn from the receiver compartment every 90 min in the passive stages and every 20 min in the iontophoretic stage and replaced with fresh buffer. Samples were analyzed with no previous treatment by HPLC-MS. At the beginning of each stage, 0.05-ml samples were taken from the donor compartment and diluted 400-fold before analysis by HPLC-MS. At the end of the passive stages, a potential difference of 250 mV was applied for 5 min to record the electrical resistance and check the integrity of the epidermal membrane barrier. At the end of each experiment, the 92

Quantitative assessment of tissue retention, lipophilicity, ionic valence and convective transport of permeant as factors affecting iontophoretic enhancement M. Altenbach, N. Schnyder, C. Zimmermann, G. Imanidis

absence of significant pH and osmolarity changes in both chambers was verified. Experiments were normally carried out at 37°C except when suppression of cutaneous metabolism was intended in which case 4°C were used [15, 16]. From the cumulative amount of the substances permeating the membrane as a function of time, permeability coefficients were calculated for each experimental stage using the equation: P = (dQ/dt)(1/ACD)

compound [15]. Retention in the epidermis decreased markedly upon iontophoresis in the presence of exogenous Tyr and Phe eliciting an increase of the apparent iontophoretic permeability and hence a marked iontophoretic enhancement. BAd-α-Glc, an unionized compound, was serving in those experiments as a reporter of electroosmotic flow and the amino acids Tyr and Phe were the product of epidermal metabolism of the model dipeptide TyrPhe that was used concurrently in the iontophoresis. By comparison, in the absence of Tyr or Phe, iontophoresis at the same pH values caused a marginal flux retardation of BAd-αGlc while tissue retention remained large. Flux retardation indicated electroosmotic flow occurring in the cathode-to-anode direction in accordance with the accepted view of positive fixed epidermis charges prevailing in this pH range. Therefore, an interaction was suggested to take place between BAd-α-Glc and Tyr and Phe impacting tissue retention and apparent permeation of BAd-α-Glc upon iontophoresis [15]. The a priori addition of Tyr and Phe to the donor solution produced the same effect on iontophoresis of BAd-α-Glc as the metabolism of TyrPhe (Table I). For BAd-α-Glc used alone, the enhancement factor was < 1 at pH 4.5 while remarkably at pH 3 negative iontophoretic fluxes were found, this being in principle in agreement with the previous results. The negative fluxes are an expression of the cathode-to-anode electroosmotic flow dominating over the diffusive flux due to the concentration difference between donor and receiver compartment. The addition of Tyr and Phe (compositions BAd-α-Glc + Tyr + Phe pH 3 and pH 4.5, Table I) yielded enhancement factors of BAd-α-Glc > 1 at both pH values. The > 1 enhancement factors correlated with a strong reduction of the amount of BAd-α-Glc recovered from the tissue at the end of the experiments in the presence of the amino acids compared to BAd-α-Glc used alone (Table II) that was statistically significant (22.43 versus 3.3 µg/specimen, p < 0.01 and 24.1 versus 11.2 µg/specimen, p < 0.1, t-test double sided). This further confirms that these amino acids interacted with the BAd-α-Glc retention in the epidermis during iontophoresis and caused as a result a profound increase of its apparent iontophoretic permeation. The fact that retention was measured at the end of the experiments again does not seem to compromise this conclusion [15]. The amino acids had no significant influence on the passive permeability of BAd-α-Glc. This is concluded by inspecting the permeability coefficients which were adjusted to an electrical membrane resistance of 60 kΩ·cm2 in order to be corrected for the inherent permeation barrier variation between membrane specimens [15] (Table I). Finally, the average amino acid fluxes into the receiver solution (Table I) and their amount recovered from the tissue (Table II) when the amino acids were added to the donor compartment seemed to be greater than the respective values of the blank runs reflecting endogenous amino acids. This result validates the measurements. It is further interesting to notice that, at the employed concentrations, the amount of Tyr and Phe recovered from the epidermis when these were added a priori to the donor solution was comparable to the amount recovered from the epidermis when the amino acids were produced by metabolism of the dipeptide TyrPhe in the tissue [15]. Tyr and Phe exerted separately the same effect in qualitative terms as their mixture used at the same individual concentration at pH 4.5 (compositions BAd-α-Glc + Tyr pH 4.5 and BAd-α-Glc +Phe pH 4.5, Tables I and II). The apparent enhancement factor of BAd-α-Glc in the presence of each of the amino acids was > 1 and the amount of BAd-α-Glc retained by the tissue was reduced compared to when BAd-α-Glc was used alone (24.1 versus 13.53 µg/specimen, p < 0.2 and 24.1 versus 13.38 µg/specimen, p < 0.2, t-test double sided). The magnitude of these effects, however, was clearly smaller for the individual amino acids compared to their mixture. The amino acids themselves showed separately the same behavior with respect to epidermis permeation and tissue accumulation as when used in a mixture. These data demonstrate that Tyr and Phe share the property of interacting

Eq. 1

where dQ/dt is the slope of the curve of the amount permeating in the receiver compartment versus time, A is the surface area of diffusion (≈ 2 cm2) and CD is the average donor concentration of the respective stage. For the amino acids Tyr and Phe that were of interest solely as adjuvants, the appearance in the receiver compartment is reported as time-averaged flux over the entire duration of the experiment. The ratio of permeability coefficients of the iontophoretic and the baseline passive stages gave the enhancement factor, E, of iontophoresis: E = Piont/Ppassive

J. DRUG DEL. SCI. TECH., 16 (1) 91-98 2006

Eq. 2

3. Skin extraction

After completion of the permeation experiments, the epidermal membrane that had a surface area of approximately 4.5 cm2 was dismounted from the diffusion cell and pulverized at liquid nitrogen temperature. A Freezer mill, type 6750 (Spex CertiPrep, Inc., New Jersey, USA) was used with the following settings: P Cool T3 (freezing time 10 min), Run T1 (milling time 2-5 min), Cycles (1 cycle). After milling, the pulverized tissue was taken up in HPLC mobile phase, briefly ultrasonicated and centrifuged at 10,000 rpm for 10 min. The supernatant was analyzed by HPLC-MS for the permeants and possible degradation products.

4. Partition coefficient

Partition coefficients in n-octanol/aqueous buffer pH 3 and pH 4.5 were determined at room temperature using the shaking flask method [17]. The equilibrium concentration of the compounds was determined in the aqueous and the organic phase by UV spectrometry.

5. Assay

Tyr, Phe, BAd-α-Glc, Tyr-β-NA and TyrPhe were assayed by HPLC-MS (Hewlett Packard and Agilent 1100 system). A Spherisorb ODS2 chromatography column (CC 125 x 2 mm, 5 µm, Macherey and Nagel, Switzerland) was used with the following mobile phases: for Tyr, Phe, TyrPhe and BAd-α-Glc, 5% acetonitrile and 95% aqueous ammonium acetate 10 mM solution set to pH 3.2 with acetic acid; for Tyr-β-NA, 20% acetonitrile and 80% aqueous ammonium acetate 10 mM solution set to pH 3.2 with acetic acid. The settings of the mass spectrometer were API-ES ionization source, SIM mode, positive polarity, drying gas flow 10 l/min, nebulizer pressure 30 psig, drying gas temperature 350°C, fragmentor variable 40-80 Volt, capillary voltage 4000 Volt, peak width 0.2 min, cycle time 1.2 s/cycle. Detection was performed at m/z 182, 204 and 226 for Tyr, m/z 166 for Phe, m/z 312 for BAd-α-Glc, m/z 307 for Tyr-β-NA and m/z 329 for TyrPhe. Samples of the partition coefficient experiments were analyzed by UV spectrometry (Perkin Elmer, Lamda 5) at 275 nm for TyrPhe, 279 nm for Tyr-β-NA and 216 nm for BAd-α-Glc. Quantification was always performed against a set of external standard solutions.

II. RESULTS AND DISCUSSION 1. Tissue retention and its modulation

Iontophoretic enhancement of the glucose derivative BAd-α-Glc across epidermis under anodal iontophoresis at pH 3 and pH 4.5 was found in an earlier study to correlate with tissue retention of the 93

Quantitative assessment of tissue retention, lipophilicity, ionic valence and convective transport of permeant as factors affecting iontophoretic enhancement M. Altenbach, N. Schnyder, C. Zimmermann, G. Imanidis

J. DRUG DEL. SCI. TECH., 16 (1) 91-98 2006

Table I - Permeability coefficient, P, enhancement factor, E, of BAd-α-Glc, and Flux, J, of Tyr and Phe at pH 3.0 and pH 4.5 for different donor compositionsa. pH 3.0 3.0 3.0 3.0 3.0 3.0 3.0 4.5 4.5 4.5 4.5 4.5 4.5 4.5

Donor composition BAd-α-Glc BAd-α-Glc + Tyr + Phe BAd-α-Glc + Tyr-b-NA BAd-α-Glc + TyrPhe at 4°C BAd-α-Glc + Tyr-b-NA + Tyr + Phe BAd-α-Glc + TyrPhe + Tyr + Phe at 4°C Blank BAd-α-Glc BAd-α-Glc + Tyr + Phe BAd-α-Glc + Tyr BAd-α-Glc + Phe BAd-α-Glc + Tyr-b-NA BAd-α-Glc + TyrPhe at 4°C Blank

P x 109 (cm/s)

E

Passive I

Iontophoresis

Passive If

1.55 (0.45) 20.03 (14.15) 5.56 (5.35) 2.16 (0.63) 3.28 (1.77) 5.08 (4.56) NAe 2.92 (1.46) 5.88 (3.40) 2.50 (0.91) 1.19 (0.39) 0.20 (0.089) 8.32 (3.81) NAe

d

1.71 (0.89) 2.78 (1.25) 2.18 (1.98) 0.12 (0.022) 1.98 (1.49) 2.41 (0.46) NAe 1.08 (0.30) 1.81 (0.47) 2.50 (1.11) 1.30 (0.38) 1.59 (1.56) 1.61 (0.95) NAe

102.4 (86.85) d d

11.28 (4.43) 16.17 (13.44) NAe 1.39 (0.56) 18.49 (10.25) 3.12 (0.92) 2.17 (0.49) 0.13 (0.023) 6.85 (5.04) NAe

d

4.76 (2.01) d d

4.01 (0.76) 3.92 (1.84) NAe 0.78 (0.18) 3.08 (0.40) 1.75 (0.49) 2.35 (0.66) 0.89 (0.35) 0.86 (0.31) NAe

J x 106 (µg/s/cm2)b Tyr

Phe

NMc 33.55 (19.87) NMc 1.21 (0.21) 10.19 (4.28) 7.84 (2.69) 1.02 (0.23) NMc 2.96 (0.95) 7.59 (0.33) NMc NMc 1.94 (0.53) 1.91 (0.21)

NMc 30.00 (13.05) NMc 0.92 (0.13) 5.76 (2.96) 9.24 (1.15) 0.74 (0.012) NMc 1.78 (0.11) NMc 2.72 (0.74) NMc 0.72 (0.11) 0.84 (0.14)

Mean values and standard error of the mean in parenthesis (n = 3-8). bAverage flux over 69 h of permeation. cNM, not measured. dPermeability coefficients and enhancement factors include positive and negative values. eNA, not applicable, fPermeability coefficient of passive I stage interpolated to a membrane electrical resistance of 60 kΩ·cm2 using the correlation established in [15]. a

Table II - Amount of BAd-α-Glc, Tyr and Phe extracted from the epidermis at the end of permeation experiments for different donor compositionsc. pH 3.0 3.0 3.0 3.0 3.0 3.0 3.0 4.5 4.5 4.5 4.5 4.5 4.5 4.5 a

Donor composition BAd-α-Glc BAd-α-Glc + Tyr + Phe BAd-α-Glc + Tyr-b-NA BAd-α-Glc + TyrPhe at 4°C BAd-α-Glc + Tyr-b-NA + Tyr + Phe BAd-α-Glc + TyrPhe + Tyr + Phe at 4°C Blank BAd-α-Glc BAd-α-Glc + Tyr + Phe BAd-α-Glc + Tyr BAd-α-Glc + Phe BAd-α-Glc + Tyr-b-NA BAd-α-Glc + TyrPhe at 4°C Blank

Amount (µg/skin sample) BAd-α-Glc

Tyr

Phe

22.43 (4.18) 3.30 (0.79) 23.78 (4.82) 17.90 (3.98) 8.15 (2.79) 24.63 (4.55) NAb 24.10 (5.18) 11.20 (3.02) 13.53 (4.12) 13.38 (3.99) 23.03 (6.48) 22.63 (5.20) NAb

NM 3.08 (0.49) NMa 1.10 (0.32) 3.95 (0.85) 3.94 (1.65) 1.02 (0.50) NMa 4.35 (1.56) 3.85 (0.73) 1.53 (0.52) NMa 1.93 (0.65) 1.45 (0.50)

NMa 2.53 (0.78) NMa 0.94 (0.61) 3.15 (0.40) 2.82 (1.34) 0.90 (0.40) NMa 4.13 (0.92) 1.40 (0.38) 3.75 (0.78) NMa 1.53 (0.38) 1.06 (0.43)

a

NM, not measured. bNA, not applicable. cMean values and standard error of the mean in parenthesis (n = 3-10).

with BAd-α-Glc retention by the epidermis upon iontophoresis and consequently increasing the apparent iontophoretic permeability. No difference in their potency in this respect was obvious within experimental error. The conclusions reached from the experiments with the individual amino acids carried out at pH 4.5 are considered to most likely apply also to pH 3 where the effects were, if anything, stronger as judged by the studies with the Tyr and Phe mixture. The tyrosine derivative Tyr-β-NA and the dipeptide TyrPhe did not exert the same effect as Tyr and Phe in terms of iontophoretic enhancement and reduction of tissue retention of BAd-α-Glc at both pH values (compositions BAd-α-Glc + Tyr-β-NA and BAd-α-Glc + TyrPhe 4°C at pH 3 and pH 4.5, Tables I and II). Tyr-β-NA is enzymatically stable and TyrPhe was used at 4°C in order to prevent metabolism [15]. Moreover, Tyr-β-NA and TyrPhe did not interfere with the action of Tyr and Phe on the iontophoretic behavior of BAd-α-Glc (compositions BAd-α-Glc + Tyr-β-NA + Tyr + Phe pH 3 and BAd-α-Glc + TyrPhe + Tyr + Phe pH 3, Tables I and II). This property, therefore, seems so far to be unique to these amino acids. It is not possible based on the present data to ascertain the chemical characteristics that are required for this property. Since Tyr and Phe have very similar lipophilic and ionized groups as Tyr-β-NA and TyrPhe, it seems that only the smaller molecular size of the amino acids distinguishes them from the other compounds.

Epidermis permeation and iontophoresis of Tyr-β-NA and TyrPhe (at 4°C) in the presence of Tyr and Phe added to the donor solution are shown in Tables III and IV, respectively. The amino acids had no impact on the electrical resistance-corrected passive permeability coefficients or the iontophoretic enhancement of either compound. Also, the amino acids had no impact on the epidermis retention of the compounds (Tables V and VI). This effect was examined at pH 3 because this is where it was expected to be the strongest based on the BAd-α-Glc results. This result was independent of the presence of BAdα-Glc (shown for Tyr-β-NA). Further, BAd-α-Glc did not influence Tyr-β-NA permeability, enhancement or tissue retention at either pH (Table III). These results are consistent with the BAd-α-Glc results inasmuch as the effect, or lack thereof, of the amino acids on tissue retention correlates with their effect on the measured iontophoretic enhancement of the concomitantly present compounds. Notably, in the same experiments which showed no influence of the addition of Tyr and Phe on tissue retention and iontophoretic enhancement of Tyr-β-NA and TyrPhe, these amino acids did have a major impact on the behavior of BAd-α-Glc. This demonstrates that this effect of Tyr and Phe is not universal but depends on the permeant in question. The three permeants studied here differ in that one of them contains an hexose structure and is uncharged while the other two are peptides that are partly positively charged at the used pH values. 94

Quantitative assessment of tissue retention, lipophilicity, ionic valence and convective transport of permeant as factors affecting iontophoretic enhancement M. Altenbach, N. Schnyder, C. Zimmermann, G. Imanidis

J. DRUG DEL. SCI. TECH., 16 (1) 91-98 2006

Table III - Permeability coefficient, P, and enhancement factor, E, of Tyr-β-NA at pH 3.0 and pH 4.5 for different donor compositionsa. pH 3.0 3.0 3.0 3.0 4.5 4.5

Donor composition

P x 109 (cm/s)

Tyr-β-NA Tyr-β-NA + BAd-α-Glc Tyr-β-NA + Tyr + Phe Tyr-β-NA + BAd-α-Glc + Tyr + Phe Tyr-β-NA Tyr-β-NA + BAd-α-Glc

E

Passive I

Iontophoresis

Passive I

0.36 (0.058) 1.32 (0.70) 0.98 (0.87) 0.61 (0.29) 0.96 (0.53) 0.88 (0.035)

1.20 (0.23) 7.29 (4.33) 2.65 (1.88) 4.58 (2.34) 4.76 (2.44) 4.73 (0.89)

0.41 (0.052) 0.45 (0.095) 0.60 (0.37) 0.47 (0.32) 1.02 (0.63) 1.53 (0.057)

b

3.38 (0.73) 4.46 (0.91) 4.72 (2.27) 4.56 (1.48) 5.65 (0.62) 5.35 (0.84)

Mean values and standard error of the mean in parenthesis (n = 3-8). bPermeability coefficient of passive I stage interpolated to a membrane electrical resistance of 60 kΩ·cm2 using a log-log linear correlation between passive permeability and resistance with slope -1.03 and r = 0.66 for pH 3.0 and pH 4.5. a

Table IV - Permeability coefficient, P, and enhancement factor, E, of Tyr-Phe at pH 3.0 and pH 4.5 for different donor compositionsa. pH 3.0 3.0 4.5

Donor composition

P x 109 (cm/s)

TyrPhe + BAd-α-Glc at 4°C TyrPhe + BAd-α-Glc + Tyr + Phe at 4°C TyrPhe + BAd-α-Glc at 4°C

E

Passive I

Iontophoresis

Passive I

14.86 (13.37) 26.21 (25.74) 46.75 (43.12)

22.57 (20.3) 51.13 (50.44) 74.25 (68.36)

10.01 (8.84) 11.83 (10.84) 11.59 (9.58)

b

1.57 (0.23) 1.62 (0.17) 1.06 (0.43)

Mean values and standard error of the mean in parenthesis (n = 3-4). bPermeability coefficient of passive I stage interpolated to a membrane electrical resistance of 60 kΩ·cm2 using a log-log linear correlation between passive permeability and resistance with slope -1.04 and r = 0.67 for pH 3.0 and pH 4.5. a

these results draw oneʼs attention to the possible role of tissue retention of permeants and drugs in particular for iontophoretic transport and its modulation by physiological substances such as amino acids.

Table V - Amount of Tyr-β-NA extracted from the epidermis at the end of permeation experiments for different donor compositionsa.

a

pH

Donor composition

Amount (µg/skin sample)

3.0 3.0 3.0 3.0 4.5 4.5

Tyr-β-NA Tyr-β-NA + BAd-α-Glc Tyr-β-NA +Tyr + Phe Tyr-β-NA + BAd-α-Glc +Tyr + Phe Tyr-β-NA Tyr-β-NA + BAd-α-Glc

12.38 (2.41) 12.90 (2.32) 10.43 (3.82) 12.65 (4.45) 17.45 (3.85) 16.85 (2.60)

2. Permeant lipophilicity, ionic valence and convective transport

Iontophoretic enhancement factors of the three model permeants Tyr-β-NA, TyrPhe and BAd-α-Glc for bulk pH 3 and 4.5 are collected in Table VII (last column). BAd-α-Glc values only in the absence of extraneous Tyr and Phe are considered. In the same table the weighted average net ionic valence of the permeants in the bulk, zb, is given. This was calculated based on the bulk pH and the pKa of the compounds considering that the pKa defined an equilibrium between positively charged and zwitterionic species. Also the experimental octanol/aqueous buffer partition coefficient, Ko/w, is given that was used as a measure of the lipophilicity of the compounds. Tyr-β-NA exhibited the strongest enhancement, which was greater at pHb 4.5 than at pHb 3 while for TyrPhe the opposite was true. For BAd-α-Glc a flux retardation was caused by iontophoresis at these pH values. In order to delineate in a quantitative fashion the contribution of factors that are relevant for these results and gain a understanding about the physicochemical phenomena and the permeant properties that play a role in iontophoretic enhancement, the model developed previously [14, 21] that was based on the modified Nernst-Planck equation was used to evaluate the data. The model was applied simultaneously to the results obtained with all three permeants at both pHb values for the purpose of reaching conclusions with rather general validity. In this model, the ionic valence of the permeants in the aqueous epidermis domain, the lipophilicity of the permeants and

Mean values and standard error of the mean in parenthesis (n=3-8).

Table VI - Amount of Tyr-Phe extracted from the epidermis at the end of permeation experiments for different donor compositionsa. pH

Donor composition

Amount (µg/skin sample)

3.0 3.0 4.5

TyrPhe + BAd-α-Glc at 4°C TyrPhe + BAd-α-Glc + Tyr + Phe at 4°C TyrPhe + BAd-α-Glc at 4°C

9.40 (1.08) 9.03 (2.47) 2.80 (0.18)

Mean values and standard error of the mean in parenthesis (n=4-11).

a

Interestingly, BAd-α-Glc, for which the effect was observed, exhibited by far the largest tissue retention of the three compounds when no amino acids were added. The limited sample size does not allow, of course, to reach a conclusion as to the structural characteristics of the permeants that may be relevant for this type of interaction with Tyr and Phe. Clearly, a larger study is required for this purpose. Nevertheless, Table VII - Results of model based data evaluation. pHb

Temp. (°C)

Compound

zb

zad

Kad/b

Ko/w

(Pld/Pad)passive x 104

Pe

Ecalculated

Eexperim.

3.0

37 37/4 4 37 37 4 4

Tyr-β-NA BAd-α-Glc TyrPhe Tyr-β-NA BAd-α-Glc TyrPhe BAd-α-Glc

0.98 0.00 0.76 0.66 0.00 0.091 0.00

0.96 0.00 0.56 0.63 0.00 0.082 0.00

0.41 1 0.54 0.9 1 0.99 1

4.81 0.13 0.80 7.28 0.11 0.61 0.11

175.6 2.4 22.2 146.0 2.0 11.3 2.0

-4.83

4.15

4.15

a

a

a

-4.83 -0.36 -0.36 -0.58 -0.58

1.6 5.48 0.83 1.15 0.74

1.60 5.50 0.78 1.06 0.86

4.5

a

Experimental enhancement factors partly negative; for separate evaluation see text. 95

J. DRUG DEL. SCI. TECH., 16 (1) 91-98 2006

Quantitative assessment of tissue retention, lipophilicity, ionic valence and convective transport of permeant as factors affecting iontophoretic enhancement M. Altenbach, N. Schnyder, C. Zimmermann, G. Imanidis

their convective transport due to electroosmotic flow are taken into account. The iontophoretic enhancement factor under constant voltage application is given by the following expression that was derived theoretically from the model:

E=

−zad B + Pe passive 1+ (Pld / Pad )

 −zad B + Pe  1− exp passive   1+ ( Pld / Pad ) 

simultaneously to all experimental enhancement factors (except for BAd-α-Glc at pHb 3, see below) and the deduced parameter values reflected the quantitative effect of the corresponding factors on iontophoretic enhancement. In this analysis the following assumptions were made: i) Pe was dependent on pH and temperature but not on the permeant since the three permeants had approximately the same molecular weight and therefore the same effective diffusional and convective mobility; ii) pH differences between aqueous domain of the epidermis and the bulk of either 0.05 or 0.4 units were considered; iii) the relative value of (Pld/Pad)passive using BAd-α-Glc pHb 4.5 as an absolute point of reference was estimated based on the ratio of the experimental n-octanol/aqueous buffer partition coefficient, Ko/w, to the calculated aqueous domain/bulk partition coefficient, Kad/b. An iterative procedure was applied consisting in assigning values to zad and Kad/b corresponding to the high or the low surface charge density and seeking the optimal reference (Pld/Pad)passive and Pe values. Calculations were performed using a least squares optimization software (MINSQ, MicroMath, Inc., Salt Lake City, UT, USA). The best agreement between the calculated and the experimental enhancement factors was obtained for the values of zad, Kad/b, (Pld/Pad)passive and Pe reported in Table VII. The agreement between calculation and experiment is considered very good, supporting the validity of this model approach. The values of and Kad/b in Table VII corresponded to the high surface charge density at pHb 3 and the low one at pHb 4.5. These produced small to moderate differences between the ionic valence in the aqueous tissue domain, zad, and that in the bulk, zb. Pe values were slightly negative at pHb 4.5 indicating a small cathodeto-anode convective transport due to electroosmosis, which implies an isoelectric point of the tissue of somewhat above 4.5. At pHb 3, a 10-fold stronger convective transport in the cathode-to-anode direction was found compared to pHb 4.5. This is consistent with the high surface charge density at pHb 3 and the low surface charge density at pHb 4.5 found in connection with zad and Kad/b. The difference of epidermal charge density between the two pH values is consistent with the notion that in the vicinity of the isoelectric point the tissue is virtually net uncharged while at pH values distant from the isoelectric point the fixed tissue charge increases. This estimation of the contribution of electroosmotic flow to the enhancement agrees with previous reports on the role of this factor in iontophoresis [9, 18, 19] and provides a basis for its quantitative assessment. No clear dependence of Pe on temperature was evident. (Pld/Pad)passive, although it differed considerably between permeants, reached a maximal value of 0.018, indicating that the aqueous pathway vastly dominated transepidermal permeation. Thus, this factor was not found to significantly affect iontophoretic enhancement of the studied compounds. This might be because even the most lipophilic of them was not lipophilic enough to exhibit a pronounced lipid pathway permeation. The presented analysis has yielded remarkably congruent results in that the estimated effect of the involved factors was consistent throughout and physically reasonable and the calculated iontophoretic enhancement closely matched the experimental data of all simultaneously analyzed compounds. Convective transport in the cathodeto-anode direction severely impeded iontophoretic enhancement of Tyr-β-NA at pHb 3 where its magnitude in terms of enhancement was about half of that generated by the electric field acting directly on the ionized permeant species. The same convective transport at pHb 3 all but abolished enhancement of TyrPhe which had a smaller ionic valence than Tyr-β-NA in the aqueous epidermis domain. Incidentally, the ionic valence of TyrPhe at this pH was markedly reduced in the epidermis compared to the bulk. Enhancement of Tyr-β-NA at pHb 4.5 was greater than at pHb 3 despite the diminished ionic valence of the compound because of the strong reduction of the cathode-to-anode convective transport. For TyrPhe, on the other hand, the decrease of the ionic valence in the aqueous epidermis domain at pHb 4.5 compared to pHb 3 outweighed the reduction of the cathode-to-anode convective

Eq. 3

where zad is the weighted average net ionic valence of the permeant in the aqueous tissue domain, B is a dimensionless variable given by B = F∆Ψ/RT with F the Faraday constant, R the gas constant, T the absolute temperature and ∆Ψ the electrical potential difference applied to the membrane, Pe is the Peclet number, a dimensionless variable given by Pe = νh/Dad with ν the linear convective flow velocity of the solvent that is effective for solute transport, h the thickness of the membrane and Dad the effective diffusion coefficient in aqueous domains, and (Pld/Pad)passive is the ratio of the permeability coefficients for lipid (Pld) and aqueous membrane domains (Pad) under passive conditions. ∆Ψ for the anodal iontophoresis experiments was -250 mV. With this potential difference, membrane alteration was assumed to be negligible, as verified in earlier studies [14]. The ionic valence in the aqueous tissue domain was estimated based on the pH prevailing in this domain. This pH differed from that of the bulk because of the electrical double layer at the aqueous/non-aqueous (lipid) domain interface within the epidermis and was estimated on the basis of the Poisson-Boltzmann equation for surface charge densities of the tissue of 0.0035 and 0.035 C/m2 [21-23]. These charge densities are representative of the values reported in the literature and define a range which likely includes the actual values. The calculated pH difference between epidermis aqueous domain and bulk is 0.05 and 0.4 pH units for the low and high charge density, respectively [21]. The Peclet number, Pe, is an effective variable reflecting the effect of the convective solvent flow on iontophoretic enhancement and containing the hindrance factor for transport by convection which depends on the molecular size of the permeant in relation to the dimension of the pathway. Since a theoretical calculation of the convective flow velocity on the basis of the electrical double layer and of the hindrance factors for transport by convection and diffusion in the tortuous aqueous pathway of the epidermis is not readily possible, Pe is treated in this work as an adjustable parameter. (Pld/Pad)passive finally, takes into account the fact that iontophoretic enhancement shall be diminished in case of compounds permeating the epidermis to some extent by the lipid pathway which does not contribute to the conductance of electric current or to convective solvent flow. Pld is considered here to be proportional to the lipophilicity of the permeant as expressed by its octanol/aqueous buffer partition coefficient, Ko/w, and Pad is considered to be proportional to the partition coefficient between the aqueous domain and the bulk, Kad/b, which for ionized compounds is generally different than unity because of the fixed charge carried by the tissue. Kad/b is expressed by Equation 4 [21]:

K ad / b =

10( pK a − pH ad ) + 1 ( pH b − pH ad )z2 (10 ) 10( pK a − pH b ) + 1

Eq. 4

where pKa is ionization constant of the permeant, pHb is the pH of the bulk solution, pHad refers to the pH in aqueous membrane domain and z2 is the smallest of the two ionic valences of the permeant. For the analysis of the experimental results using Equation 3, values of the parameters included in this equation were sought that yielded the measured enhancement factors. This analysis was applied 96

Quantitative assessment of tissue retention, lipophilicity, ionic valence and convective transport of permeant as factors affecting iontophoretic enhancement M. Altenbach, N. Schnyder, C. Zimmermann, G. Imanidis

transport so that iontophoretic enhancement was diminished. The measured enhancement factor of BAd-α-Glc being < 1 was consistent with the convective transport of all compounds. Hence, a complete quantitative interpretation of the experimental data is afforded by the relevant factors involved in the model. This and the fact that the magnitude of the effect of these factors was consistent within the framework of the process lend support to the validity of the used model yielding Equation 3 and of the underlying mechanistic understanding of iontophoresis. They further demonstrate that this approach, besides affording a quantitative evaluation and providing information about the contribution of individual factors to iontophoretic enhancement, may be used as a basis for predicting its outcome. The ionic valence within the aqueous domain of the epidermis rather than in the bulk appears to be relevant since this domain constitutes the effective pathway that is responsible for electric current conduction and hence iontophoretic transport. The electrical double layer at the interface between aqueous and non-aqueous domains of the epidermis which gives rise to the electroosmotic flow and thus convective transport also affects the ionic valence of permeants by creating a pH micro-environment in the aqueous domain of the epidermis. Finally, permeant lipophilicity although in theory may interfere with iontophoretic enhancement due to the lipid pathway permeation not being subject to enhancement by electrical current or electroosmotic flow, does not seem to play a significant role for permeants with rather small lipophilicity of the order used in the present study. The negative iontophoretic fluxes and hence negative enhancement factors obtained for BAd-α-Glc at pH 3 (Table VII) could be analyzed quantitatively only by relaxing the boundary condition of sink prevailing in the receiver compartment that was assumed for the derivation of Equation 3. The resulting enhancement factor for zero ionic valence applying to BAd-α-Glc for an exclusively aqueous pathway of permeation is given by Equation 5:

 CR  PeCD −  exp(Pe)   E= (1- exp(-Pe))(CD − CR )

J. DRUG DEL. SCI. TECH., 16 (1) 91-98 2006

factors. This is probably because iontophoretic enhancement factors represent relative measurements in which each iontophoretic flux reading has its own control of passive flux from the same epidermis specimen. Passive epidermis permeability of Tyr-β-NA, BAd-α-Glc and TyrPhe was not expected to correlate with the lipophilicity of the compounds expressed by their octanol/aqueous buffer partition coefficient in light of the finding of the iontophoretic enhancement analysis that the aqueous pathway dominated passive permeation. Hence, passive permeability differences should be attributable, if anything, to the different accessibility of the aqueous domain for the three permeants, which is subject to charge-charge interactions, whereas in the present situation, size considerations should not play a role because the used compounds had all approximately the same molecular weight. Yet the moderate differences of Kad/b between the different compounds and pH values (Table VII) in conjunction with the variability of the passive permeability values did not allow establishing a correlation for these three permeants. Further, no correlation between passive permeability and tissue retention of the permeants was found. * The physicochemical model taking into account lipophilicity, ionic valence in the aqueous epidermis domain and convective transport of the permeant due to electroosmosis that is employed to simultaneously analyze the iontophoretic enhancement data of Tyr-β-NA, TyrPhe and BAd-α-Glc offers the possibility to determine consistent and physically relevant parameter values that characterize the effect of the above factors on the enhancement. These, when combined in the model equation, allow enhancement factors to be computed, that closely match the experimental results of all permeants. The deduced effects of these factors constituting physicochemical phenomena and permeant properties apply to all three compounds and are in agreement with earlier results which suggests their possible usefulness towards a universal prediction of iontophoretic enhancement under constant voltage application. Iontophoretic enhancement of selected permeants may be influenced by the addition of the amino acids Tyr and Phe at levels above the endogenous ones by a mechanism involving the retention of the permeant in epidermal tissue, which suggests the potential of using these physiological substances as adjuvants to modulate iontophoresis.

Eq. 5

where CD and CR are concentration in the donor and the receiver compartment, respectively. Equation 5 was used to estimate the magnitude of electroosmotic flow that would lead to a net flux of permeants from the receiver to the donor compartment and thus against the prevailing concentration gradient. Using for CD the actual donor concentration, no reasonable estimates of Pe could be obtained. This is probably because the molecules leaving the receiver solution never reach the donor solution but accumulate in the membrane, this notion being justified by the considerable amount of BAd-α-Glc recovered from the epidermis compared to the amount found in the receiver compartment. Therefore, an estimated permeant concentration in the outer membrane layer facing the receiver solution of 20 µg/ml was used for CD for the sake of probing Equation 5. This concentration is equal to 0.5% of the concentration in the donor solution. Using this concentration infers, therefore, a movement of the solute from the receiver solution into the outer 0.5% of the thickness of the membrane. With this value of CD and the measured receiver concentrations, Pe values ranging between -40 and -0.2 and averaging approximately -15 were calculated. The high variability of the data, however, and the required assumptions precluded a meaningful quantitative interpretation of the negative flux data. Consequently, the value of these data lies primarily in the inference that the electroosmotic flow can completely abolish flux caused by a considerable concentration gradient. Passive permeability coefficients of the three model compounds adjusted to an electrical membrane resistance of 60 kΩ·cm2 to correct for differences in the transport barrier properties between tissue specimens still showed a greater variability than the respective enhancement

REFERENCES 1. 2. 3.

4. 5.

6.

7.

97

COUVREUR P., DUCHENE D., GREEN P., JUNGINGER H.E. Ed. - Transdermal Administration, A Case Study, Iontophoresis. - Editions de Santé, Paris, 1997. GUY R.H. - Iontophoresis – recent developments. - J Pharm Pharmacol, 50, 371-374, 1998. ROBERTS M.S., LAI P.M., CROSS S.E., YOSHIDA N.H. - Solute structure as a determinant of iontophoretic transport. - In: Mechanisms of Transdermal Drug Delivery, R.O. Potts and R.H. Guy Eds., Marcel Dekker, Inc., New York-Basel-Hong Kong, 1997, pp. 291-349. LOPEZ R.F.V., BENTLEY M.V.L.B., DELGADO-CHARRO M.B., GUY R.H. - Iontophoretic delivery of 5-aminolevulinic acid: effect of pH. - Pharm Res, 18, 311-315, 2001. MARRO D., KALIA Y.N., DELGADO-CHARRO M.B., GUY R.H. - Optimizing iontophoretic drug delivery: Identification and distribution of the charge-carrying species. - Pharm Res, 18, 1709-1713, 2001. DE GRAAFF A.M., LI G.L., VAN AELST A.C., BOUWSTRA J.A. - Combined chemical and electrical enhancement modulates stratum corneum structure. - J Control Release, 90, 49-58, 2003. YAN G., PECK K.D., ZHU H., HIGUCHI W.I., LI S.K. - Effects of electrophoresis and electroosmosis during alternating current

J. DRUG DEL. SCI. TECH., 16 (1) 91-98 2006

8. 9. 10.

11.

12. 13. 14. 15.

16.

17. 18.

Quantitative assessment of tissue retention, lipophilicity, ionic valence and convective transport of permeant as factors affecting iontophoretic enhancement M. Altenbach, N. Schnyder, C. Zimmermann, G. Imanidis

iontophoresis across human epidermal membrane. - J Pharm Sci, 94, 547-558, 2005. DELGADO-CHARRO M.B., GUY R.H. - Transdermal reverse iontophoresis of valproate: a non-invasive method for therapeutic drug monitoring. - Pharm Res, 20, 1508-1513, 2003. MARRO D., GUY R.H., DELGADO-CHARRO M.B. - Characterization of the iontophoretic permselectivity properties of human and pig skin. - J Contol Release, 70, 213-217, 2001. LUZARDO-ALVEREZ A., RODRIGUEZ-FERNANDEZ M., BLANCO-MENEDEZ J., GUY R.H., DELGADO-CHARRO M.B. - Iontophoretic permselectivity of mammalian skin: characterization of hairless mouse and porcine membrane models. - Pharm Res, 15, 984-987, 1998. HOOGESTRAATE A.J., SRINIVASAN V., SIMS S.M., HIGUCHI W.I. - Iontophoretic enhancement of peptides: behavior of leuprolide versus model permeants. - J Control Release, 31, 41-47, 1994. BRUNETTE R.R., ONGPIPATTANAKUL B. - Characterisation of the permselective properties of excised human skin during iontophoresis. - J Pharm Sci, 76, 765-773, 1987. DE NUZZIO J.D., BERNER B. - Electrochemical and iontophoretic studies of human skin. - J Control Release, 11, 105-111, 1990. KOCHHAR C., IMANIDIS G. - In vitro transdermal iontophoretic delivery of leuprolide - Mechanisms under constant voltage application. - J Pharm Sci, 92, 84-96, 2003. ALTENBACH M.P., SCHNYDER N., ZIMMERMANN C., IMANIDIS G. - Cutaneous metabolism of a dipeptide influences the iontophoretic flux of a concomitant uncharged permeant. - Int J Pharm, 307, 308-317, 2006. HIGO N., HINZ R.S., LAU D.T.W., BENET L.Z., GUY R.H. - Cutaneous metabolism of nitroglycerin in vitro. II. Effects of skin condition and penetration enhancement. - Pharm Res, 9, 303-306, 1992. LEO A., HANSCH C., ELKINS D. - Partition coefficients and their uses. - Chem Rev, 71, 525-616, 1971. NICOLI S., CAPPELLAZZI M., COLOMBO P., SANTI P. - Char-

19. 20.

21. 22.

23. 24.

acterization of the permselective properties of rabbit skin during transdermal iontophoresis. - J Pharm Sci, 92, 1482-1488, 2003. BATH B.D., WHITE H.S., SCOTT E.R. - Visualization and analysis of electroosmosis flow in hairless mouse skin. - Pharm Res, 17, 471-475, 2000. PECK K.D., HIGUCHI W.I. - Characterization of the passive transdermal diffusion route of polar/ionic permeants. - In: Mechanisms of Transdermal Drug Delivery, R.O. Potts and R.H. Guy Eds., Marcel Dekker Inc., New York, 1997, pp. 267-290. IMANIDIS G., LUETOLF P. - An extended model based on the modified Nernst-Planck equation for describing transdermal iontophoresis of weak electrolytes. - J Pharm Sci, in press. PIKAL M.J. - Transport mechanisms in iontophoresis. I. A theoretical model for the effect of electroosmotic flow on flux enhancement in transdermal iontophoresis. - Pharm Res, 7, 118-126, 1990. AGUILELLA V., KONTTURI K., MURTOMAEKI L., RAMIREZ P. - Estimation of the pore size and charge density in human cadaver skin. - J Control Release, 32, 249-257, 1994. KOCHHAR C. and IMANIDIS G. - In vitro transdermal iontophoretic delivery of leuprolide under constant current application. - J Control Release, 98, 25-35, 2004.

ACKNOWLEDGMENTS We wish to thank Ralf Schoch and Prof. M.J. Mihatsch (Institute of Pathology, University Hospital, Basel, Switzerland) for their continuous support by way of donation of human cadaver skin.

MANUSCRIPT Received 27 October 2005, accepted for publication 8 February 2006.

98