journal of
ELSEVIER
controlled release Journal of Controlled Release 38 (1996) 1-10
Transport of a charged molecule across the human epidermis due to electroporation Uwe Pliquett, James C. Weaver * Harvard-MIT Division of Health Sciences and Technology, Massachusetts Institute of Technology, Cambridge, MA 02139, USA
Received 16 April 1994; accepted 16 June 1995
Abstract
Transport of a charged molecule (calcein; z = - 4 ; 623 g/mol) across the epidermis can be caused and controlled by an electric pulse protocol. Our interpretation is that the stratum corneum (SC) is altered by the field, such that a series of regularly spaced exponential electric field pulses (time constant Zpu~se=1.1 ms) resulting in Usk~n=80--230 V causes a tremendous enhancement in molecular transport. Upon beginning a pulse protocol, the flux increases from negligible values, and exponentially approaches a quasi-steady state flux with a lag time constant, ~'~ag,that depends on the pulse spacing, but not the transdermal voltage magnitude. Above a threshold of Usk~,= 80 V across the skin for the pulse conditions used here, molecular transport increases almost linearly with Usk~n,and then levels off at higher voltages ( U~k~,> 250 V) or shorter spacing ( < 10 s). When pulsing is stopped, the flux decreases by an order of magnitude within 1 min. Once the quasi-steady state is reached, it is stable for at least 6 h. For a particular specimen, the maximum flux depends on the peak voltage across the skin, the pulse time constant and the pulse spacing. Measurements of the passive electrical properties were carried out simultaneously in order to independently characterize the skin in terms of its altered ability to transport small ions. Keywords: Transdermal drug delivery; Electroporation; Molecular transport; Passive electrical properties; Human skin; Stratum corneum
1. B a c k g r o u n d
The challenge of drug delivery is increasingly recognized to be of major importance to medicine for both existing and new therapeutic agents [ 1-5]. One new approach to improved transdermal drug delivery is to consider the use of 'high-voltage' treatment, with the goal of simultaneously creating new aqueous pathways ( ' p o r e s ' ) and driving molecular transport through these pathways [ 6 - 1 3 ] . The motivation of this new approach is the possibility of causing electroporation in the lipid-containing barriers within the stratum cor* Corresponding author. 1996 Elsevier Science B.V. SSDI 0168-3659 ( 95 ) 00089-5
neum ( S C ) of the skin, in order to increase both the amount of drug delivery and its temporal control. Electroporation is a phenomena which involves the creation of aqueous pathways across lipid-containing barriers. A considerable literature for electroporation of artificial planar bilayer membranes and cell membranes has been developed, with the basic idea being that for individual bilayer membranes transient aqueous pores are formed through the joint action of thermal fluctuations (stochastic ' k T ' energy) and electrostatic energy within the membrane (deterministic field energy) [ 12,14-16]. The combined energy is stochastic, and a range of pores sizes if therefore expected. The transient aqueous pore model has been successful in explaining much of the electrical and mechanical
U. Pliquett, J.C. Weaver~Journal of Controlled Release 38 (1996) 1-10 0
behavior of planar membranes, and may be able to explain the transport of small charged molecules across cell membranes. In this case, the hypothesis is that the increased local electric field across the membrane not only creates pores, it also provides a local driving force for charged molecules [ 17-20]. In the case of transdermal drug delivery, the skin's outermost layer, the stratum corneum (SC), is the main barrier. If this barrier is destroyed, for example by mechanically abrading the stratum corneum, then the protection vanishes, and there is no great difficulty in introducing water soluble molecules. In order to accomplish the latter without significant damage, the idea of using 'high-voltage' pulses across the skin is to physically open the stratum corneum on a microscopic scale, to temporarily transport the drug, and to allow natural recovery processes to re-establish the barrier. Although the stratum corneum is much more complicated than a single phospholipid bilayer, (it is a multilamellar structure containing corneocytes) it seems possible to create new aqueous pathways within this structure [8]. If the goal is to control drug delivery as a function of time, then knowledge of the flux kinetics is important. With this in mind, the aim of the present experiments has been to determine the flux as a function of the treatment, in this case a series of pulses that resulted in significant transdermal voltages, Usrj,. The parameters of interest were both Us~,, and the spacing of the pulses, tint. The pulses used here were nominally exponential, i.e. the voltage across the chamber was approximately U = Uoe -'/~"~"°. Because the measurement method employed in the present study involved determination only of the net flux [21], a distinction between the transport through these new pathways and electrically activated pathways through appendages (hair follicles and sweat ducts) is not possible.
2. Material and methods Human skin specimens were prepared by heatstripping the stratum corneum and the first layer of the epidermis with a thickness of about 50/zm [22]. This preparation was clamped in a side-by-side permeation chamber [ 8 ]. For the measurement of molecular transport due to high-voltage pulses, a flow through system was used [ 21 ]. The donor compartment of the chamber contacted the SC, and was filled with phosphate buf-
(HOCCH2)2NH2C
CH2N(CH2COH)2
,( Y
COOH
Fig. 1. Structure of calcein, an established polar tracer which is generally inert. It has intense green fluorescence emission, a charge z = - 4, and a size of 623 g / m o l [ 33 ].
fered saline (PBS 7.2, 150 mM) containing 1 mM calcein (charge q = - 4 e with e = 1.6X 1 0 - 1 9 C , fluorescence emission maximum at 515 nm [ G F = green fluorescence], M = 623 g/mol; Fig. 1). The receptor compartment solution volume was continuously replaced with fresh PBS by using a flow-through stream, which was driven by a peristaltic pump in order to provide a calibrated flow rate ranging from 0.01 to 0.03 ml s 1. A fluorescence measurement system (spectrofluorimeter, Fluorolog 2, S P E X Industries) was located between the peristaltic pump and the receptor chamber. This allowed continuous determination of the amount of calcein in the receptor compartment. The GF emission intensity was calibrated in terms of calcein concentration. The calibration was carried out over the concentration range of 100 pM to 1/xM. We also took precautions relating to interactions which degrade the fluorescence of calcein. Specifically, in order to prevent bleaching of calcein by oxygen, and to avoid disturbance of the fluorescence measurements by air bubbles in the fluorimeter cuvette, the PBS was degassed by exposure to a reduced pressure (80 kPa; standard atmosphere = 100 kPa). The degassing was further enhanced by using ultrasound. The mathematical description of the flow through system and the deconvolution of the measured signal is described elsewhere [21]. The pre-experimental phase required about 2 h. During this preparation time the passive (no pulsing) calcein flux and the passive electrical properties of the skin were measured. Other than the absence of electrical pulsing, the experimental conditions were the same as those used for enhanced calcein flux measurements. This pre-experimental phase was necessary in order to
u. Pliquett, J.C. Weaver/Journal of Controlled Release 38 (1996) 1-10
establish with confidence a steady baseline condition. The flow rate caused by the peristaltic pump was found to be influenced by the height of the PBS in the reservoir supplying the receptor chamber. For this reason, the PBS supply level was controlled using level sensing and a small water pump, which provided PB S as needed from an independent reservoir. The chambers contained two electrode pairs: ( 1 ) inner electrodes located 1.1 cm from each side of the skin, and (2) outer electrodes at a distance of 2.7 cm from each side. Ag/AgCI electrodes (In VivoMetric, Healdsburg) were used throughout. The advantage of these electrodes is their low electrode polarization voltage, and their ability to scavenge chlorine when used with highvoltage pulses. This is very important because C12 is produced at the positive electrode, and this electrode is usually located in the receptor compartment in order to drive the negatively charged calcein. But the interaction of C12 and calcein lead to bleaching of a significant fraction of the calcein, with this fraction especially large at the low calcein concentrations found in the receptor compartment. In order to obtain time resolution adequate for characterization of the rapidly changing electrical behavior after an electroporating pulse [ 23 ], measurements of the skin's passive electrical properties were carried out in the time domain using a rectangular wave ( Vo = 150 mV; repetition at f = 1.25 kHz). Previous skin measurements in the frequency domain using a four-electrode interface revealed a frequency dispersion between 100 Hz and 100 kHz, but below 100 Hz no additional dispersion was found. For this reason, we calculated the passive electrical properties in the range 1.25 kHz and 1.25 MHz from the deformed excitation voltage trace, and then extrapolated the low frequency behavior. The entire experimental apparatus was controlled by a microcontroller (MiniCon52, Phytec, Mainz). A high-voltage controller switched between the pulses and the low voltage impedance measurement. Pulses with an approximately exponential time dependence, V= Voe -'/~pu~, were applied to the outer pair of electrodes. The inner pair of electrodes was used to monitor the smaller voltage that appeared across the skin, Us~a,. For this purpose a digital oscilloscope (HP54602) was used in the voltage triggering mode, and the trace stored in channel 2. The voltage developed across a 5 J'2 resistor in series with the chamber was measured at channel 1 in order to obtain a measure of
3
the total current through the chamber. Subsequently the waveforms were transferred to a computer for analysis and display. Within 3-6 ms after a high-voltage pulse, a high-voltage switch connected the inner pair of electrodes to the impedance measurement system, containing a generator (HP3412) a measuring resistor and a special amplifier to match the wiring and the oscilloscope, channel 1. The rectangular wave obtained from the generator is deformed by the measured system (here a chamber-mounted human skin preparation), and after amplification was recorded by the oscilloscope. The synchronization signal of the generator was directed to channel 3 of the oscilloscope to provide triggering. The data were then automatically transferred to the computer. Although several effective resistances and capacitances can be identified by electrical measurements on unpulsed and pulsed skin [ 23 ], the skin resistance, R~kin,exhibits the largest change by far, and is also plausibly identified with changes in the aqueous pathways across the skin. For this reason, only Rskin w a s used for comparison with the calcein flux and pulse parameters. For experiments which addressed the voltage-flux relationship, a pulse spacing o f tin t = 5, 10, 20 and 60 s, and a pulse time constant of rpu~se= 1.1 ms (measured) was chosen. Pulsing was carried out for one hour. The calcein flux and passive electrical properties were measured simultaneously during the pulsing interval, and then continued for another hour post-pulsing. The high-voltage pulses were delivered from a BioRad Genepulser, which discharges a capacitor ( here 25/zF) through the electrical load. The parameter 'rpu~s~is well defined only if the load is purely resistive and fixed. In order to fix rpu~se, but also to prevent the pulser from overloading, an external voltage divider using high wattage (50 watt) resistors was used. For this purpose two resistors (40 and 10 J2) were connected in series, with the chamber connected in parallel to the 40 J'2 resistor. In order to find the sites of the pathways within which calcein transport occurred, the skin was washed after pulsing and placed on a microscope slide. Using a fluorescence microscope with 'blue excitation' illumination (approximately Aex = 435-490 nm), the calcein-containing regions within the SC could be visualized.
U. Pliquett, J.C. Weaver / Journal of Controlled Release 38 (1996) 1-10
4 0.25
It should be noted that RDy is not the ohmic resistance, but is instead a computed physical quantity with the dimension of resistance. RDy is the slope of the voltage-current relationship at a particular voltage• This means that RDy is applicable to both linear and nonlinear systems. RDy should not be confused with the skin resistance, Rskin , as the latter is valid only if the skin's response is linear, which generally occurs only for Usk~.< 1 V. The dynamic resistance exhibits impressive behavior, dropping to a few hundred ohms during the first few microseconds after application of an electroporating pulse, and further decreases over a few hundred microseconds reaching a minimum of about 150 g2. The time during which RDy reaches very v o l t a g e , Uskin(t).
230V =Ige It **=,l°llI°4=lt *l=*~ g B~lll~ oo
0.2
~n 0.15 :=L
0.1
"5 E
•
160V
• ,••• •,,,• "•• ••,-~•••. •,.~• •,• o•, ..,_.................%.~.....~..,.
120V
•
o.o~
...... o.•"
90V 05v
........................
CO"
20
40
60 time
80
1O0
120
[min]
Fig. 2. Time course of the calcein flux due to the pulsing protocol with the pulse voltage magnitude across the skin, U~k~..0, as a parameter. In the present study U~k~.,oranged from 65 to 230 V. These are average values for a particular pulsing protocol; for example, at the beginning of the pulse sequence Uskin w a s about 80 V, but as the time average skin resistance decreased U~k~.fell to slightly less than 65 V ( see Eq. 2). The first 3 min consisted of a control period, during which there was no pulsing. Pulsing at a rate offp,l~ = 1 m i n - l was applied during the first hour, but the calcein flux measurement was continued for a second hour.
(a)
0.4
0.35 a::
0.."
oE
0.2.=
~
0.-"
x
0.15
~ -5 E
0.1
3. Results Quasi steady state calcein flux as a function of time for several different pulse magnitudes is shown in Fig. 2. The voltage drop between the outer electrodes and the skin is significant• For this reason, the voltage was monitored during the pulse at the inner electrodes in order to calculate the real voltage across the skin, Uskin(t), using the chamber geometry and the resistivity, try, of the PBS, and the total current through the chamber during the pulse• The maximum value of the pulse voltage across the skin, U~kin,o, is used in presenting all of the results• To a first approximation the relationship between transported amount of calcein and U~kin,O is linear for a pulse spacing of 1 min and 5 0 < U s k i n < 2 3 0 V at 37°C. This is supported by another set of experiments, in which the same piece of skin was used with different voltages (Fig. 3a). An important derived quantity, the dynamic resistance,
was calculated from the measurements of I(t), the current through the entire system, and the transdermal
0.0~ 0
0 i
60
80
100
120
140
160
U~. [V]
180
200
220
2~0
(b)
olo/o
~
O.
0
0 0
0 0 40
6~)
8()
100
120
1~0
160
180
Uskin [Vl Fig. 3. Flux of calcein under quasi-steady state conditions with different pulsing rates (A) 1 ppm and (B) 12 ppm. In both cases a transition from iontophoresis to electroporation condition is evident. In the case of lower pulsing rate (B) a leveling off for higher voltages was found. The reason for the smaller voltage in (B) is less recovery before the next pulse, which results in a smaller transdermal resistance.
U. Pliquett, J.C. Weaver~Journal of Controlled Release 38 (1996) 1-10
small values is a function of the applied voltage, Uskin , and also the history of pulsing. For example, for subsequent pulses or for larger pulses this time interval becomes very short ( = 4 × 10 -5 s). For the most commonly employed protocol (tinter = 1 min; rouj~e--~ 1 ms) after about 10 pulses RDy reached its minimum value, about 100 J2. This value, Rskin.min, depended on the properties of the skin, and also on Usk~,- The calcein flux achieved a quasi-steady state, in which only the oscillations due to the pulse protocol are evident. The average calcein flux during this phase is a nearly linear function of the maximum transdermal voltage for this series of experiments ( Uskin,O = 65-230 V; averaged over the complete history of a particular pulsing protocol). For higher pulse voltages the rate of increase of the flux diminishes, and appears to saturate. After turning off the pulsing the flux decays immediately (within 1 min) by about one order of magnitude. After this further decay takes a much longer time. Hence, especially after larger pulse voltages (U~k~n.o> 200 V) the original background will not be reached again, even after times of up to a few hours (data not shown). The lag time and the amount transported during the quasi-steady state phase is also a function of the spacing the pulses. The dependence of the maximum flux on the pulsing rate (equivalently, on ti~t t ) is almost linear for the here used range of spacing from 5 to 60 s. The data in Fig. 4 give an example for U~k~n,o= 165 V. Not surprisingly, the slope of the pulse spacing/flux relationship depends on the particular piece of skin. The variability in this slope was found to be about 20% (because of the size of the error bars, they were omitted from Fig. 4). As shown in Fig. 2. achievement of the quasi-steady state occurs after the onset of about three to six pulses, with a nearly exponential relaxation, but varies somewhat from one skin preparation to another. This is the lag time, i.e. the time from a first detectable increase in the flux because of pulsing, to the time at which 63% of the quasi-steady state flux is achieved. The lag time seems to be a function only of the number of pulses, not of the voltage. If the experiments were characterized solely by the transdermal voltage the occurred on the first pulse of a pulsing protocol, the voltage range would be given as Uski,.o= 100--250 V. After the first pulse, however, the R~kJndrops significantly. The total resistance between the outer, pulsed electrodes is R~ham-
5
I.E ,,,,,'~
'7
®""
~-
......;
.............190V
.......
150V
"5 0.~ E
........230V
.....
' ........
~ ......... llOV
...................... ........ ® , , ...........
....... ~:::::iii.
x .......................x
...... x
a0v
o ..........................O .......................................................... o
0.bs
0'.i
o.~s
f pupso [ pulse per second ]
0'.2
Fig. 4. Dependence o f the quasi-steady state calcein flux on the pulse spacing (6,0. Because the transport occurs on a time scale much faster than the spacing o f the pulses, the total transport should be the product o f the transport per pulse multiplied by the number o f pulses. While in the quasi-steady state, therefore, the transport per time interval is proportional to the number o f pulses in that interval, which implies that flux should be linear in pulsing rate, fp = I/ti,,.
ber=Rbu~k+R~ki.. This means that a voltage divider effect must be considered, which gives
eskin
Uskin = UORskin q- Rbulk with Rbulk ~'~constant
(2)
where Uo is the amplitude of the exponential pulse measured across the (outer) chamber pulsing electrodes. This means that as Rskin decreases because of electroporation, and does not fully recover by the time the next pulse is applied, the voltage across the skin for subsequent pulses is smaller. As shown elsewhere [23], during the first pulse Rskin drops rapidly and significantly (by about two orders of magnitude for the larger pulses measured 1.8 ms after the pulse), but recovers only partially by the time the next pulse is applied. For this reason, the value of Uski, is smaller for all of the subsequent pulses even though the applied pulse magnitude is the same. With this in mind, it is better to characterize the behavior in terms of the average transdermal voltage that is achieved well into the pulse protocol, and this gives instead the range Uskin,O = 65-230 V. About 10-15 pulses are needed to reach the quasisteady state in the calcein flux, with the particular number of pulses depending on the specimen. The approach to the quasi-steady state is reasonably fit by an expo-
U. Pliquett, J. C. Weaver~Journal of Controlled Release 38 (1996) 1-10 2£ 22 18
14 12
~° 8
6
6 4 2
o
;
I'o
2'o
3'o
20
s'o
io
t~, [ s ]
Fig. 5. Plot of the time constant, ~'~, for the lag phase for calcein transport. The criterion is based on the use of an exponential approach to the quasi-steady state (Eq. 2) The number near each point indicates the number of pieces of skin used. The case of t~,~= 0 is that of continuous application of a voltage, i.e. iontophoresis conditions (typically provided by applying a constant current 1~ 100 k~A so that the current density was J = 1.4 A m 2
time from beginning of the pulsing sequence. The data were normalized by then dividing by the mean flux of the steady state in the pulsing sequence. Note that the pulses were repeated at a rate of one per min, and thus the time axis corresponds to the number of pulses. In Fig. 6b it can be seen that the increasing calcein flux requires about eight pulses to reach the quasi-steady state for the first pulsing series, but less for repeated protocols. After pulsing ceases, the decay of the flux depends slightly on the history. First, a flux decay of one order of magnitude was found for all different spacing. But after this the decay was faster for the longer spaced pulses. Also the extent of decay shows a dependence on the spacing, where for longer spacing a smaller decay was found, consistent with better recovery. For example, after a series of pulses with t~,, = 1 min the original background was almost achieved after 1 h. In
(a) 0.16
r
0.14
nential relaxation (Eq. 3). The time constant, ~-~,g,for exponential relaxation into the quasi-steady state is mainly influenced by the spacing of the pulses (Fig. 5). As criteria for %8 the achievement of 63% of the quasisteady state flux was chosen. A function of the form
x=
0.12 0.1 0.08
\
0.06
J~,l(t) = J~l,o[ 1 -- e - ' / ~ J
(3)
fits this data well. In case of the onset phase of previously pulsed skin the flux rises almost immediately, i.e. within one or two pulses to > 63% of the quasi-steady state value. Once a quasi-steady state was reached, the calcein flux closely tracked the pulsing protocol, and also showed a significant dependence on the spacing of the pulses (t~et), the transdermal voltage magnitude (U,k~e,O) and the pulse time constant (~-p,~s~). For the range investigated here, we interpret this behavior of the flux as indicating that the key attribute of the pulses is the electric charge moved across the skin. Fig. 6a provides an example of the time course of the calcein flux for three successive applications of a pulse train, with each followed by a recovery period after pulsing ceased. The behavior is the same, but for differences of the lag within the onset phase. This difference is shown more clearly in Fig. 6b, which uses an expanded scale, and plots the normalized calcein flux versus the
"6
0,04 0.02
time [h]
(b) 1.2
xXXx
~ + + ~
~+
0.8 o v
x
xxX
+ X 0.6
0.4
X at"
x 1. cycle 2.cycle
0
E
x XX
x
0.2
X
X
+ 3.cycle
X X
x
10 time [rain]
15
2'0
Fig. 6. (A) Three-fold repetition of U~k~.= 140 V pulsing protocol using a spacing of I h after 60 pulses separated by 1 rain. (B) Overlay of the above three onset phases.
u. Pliquett, J.C. Weaver~Journal of Controlled Release 38 (1996) 1-10
contrast, for tint= 5 S the flux decay decreased only to a value about 5-fold greater than the background. The behavior of small ion transport which underlies the electrical behavior of the skin is also of interest. As noted earlier, Rskin is the most significant parameter. The skin conductance, Gskin-- 1/Rskin, correlates well with the increase of the calcein flux. However, the immediate decrease after pulsing of the permeability of the SC to a highly charged molecule (calcein) is in marked contrast to the very slow recovery of Rskin. In earlier studies it was found that the recovery of RskJn after pulsing is characterized by several time constants [ 23 ]. A long-term recovery process was found to go on over hours, or, for particularly large amounts of transported charge (large voltages, many pulses).
4. Discussion It is widely accepted that high-voltage pulses are able to create additional pathways across cell membranes [ 12,14-16]. Typically short (100/zs to 3 ms) pulses are used, and a large increase in molecular transport is found if the transmembrane voltage reaches 0.5-1 V. It is well known that the SC barrier to charged species is mainly due to the approximately 100 bilayer membranes in series [24-28]. By applying a sufficiently large voltage ( = 50-100 V) to a multilayer lipid-containing structure such as the stratum comeum (SC), it is plausible that electroporation occurs. If so, both electrical and molecular transport behavior are expected, as the elevated voltage across the lipid-based barrier is believed to play the dual role of creating aqueous pathways (pores) and driving molecular transport through these pores. In the present experiments the driving force for calcein (charge q = - 4 e ; e = l . 6 × 1 0 -19 C) is interpreted by us as being electromigration during the pulse. The achievement of a quasi-steady state was found here to depend mainly on the number of pulses applied to the skin. It is not yet known when and where a steady state occurs in the case of widely spaced pulses, for which a greater recovery between pulses is expected. Moreover, it appears that all of the new pathways are created within the first 10-15 pulses, and the participating area for calcein transport does not increase with additional pulsing. This conclusion was reached by considering the appearance of intense GF regions
7
following different numbers of pulses using fluorescence microscopy. Transport of the calcein was found to occur mostly during pulses and to stop almost immediately after pulsing. This is consistent with a theoretical model which assumes that electroporation of the multilamellar bilayer structure within the SC occurs, and with simultaneous primary transport by electromigration and secondary transport (mainly concentration broadening) by diffusion [ 13]. It is also consistent with the interpretation of DNA introduction into cells being ascribed mainly to electrophoretic drift in the local electric field [ 18,19 ], and individual bilayer membrane theories that consider the transport of charged molecules through pores by the local electric field near and within the pores of a dynamic, heterogeneous pore population [ 17,20]. Because the transport decreases by about an order of magnitude within one min after pulsing (Fig. 2), diffusion seems incapable of accounting for this decrease. For small voltages across the SC the transport of a charged molecule should be a non-linear function of Uski,. This is fundamentally expected because of the energy cost to place charge near the low dielectric constant lipid, an effect which has been explicitly included in some models of planar bilayer membrane electroporation [17,20,29]. The exclusion of negatively charged calcein from aqueous pathways (pores) could be even greater if the pathways have interior negative charge, as is expected for pores in cell membranes. Qualitatively, therefore, calcein flux is expected to cease quickly after pulsing if the pathways are comparable in cross-sectional area to calcein. The continuous calcein flux measurements show that there is a small but significant flux even after a few minutes. However, this may be have to do with a loading of the epidermis while the pulses were applied, and the diffusion out of this layer after the electric field decayed. This may also explain why after using a pulsing protocol with previously pulsed skin, the very first pulses produce an apparent flux, but one which is less than that observed during the quasi-steady state phase. Fluorescence microscopy observations showed that there was significant lateral spreading (about x --40 to 80 /xm; arithmetic average of .f= 60 /xm) of the GF regions which are identified with calcein transport, which are subsequently studied in a separate series of experiments that are reported elsewhere [30]. Here fluorescence microscopy suggests that there is signifi-
8
U. Pliquett, J.C. Weaver / Journal of Controlled Release 38 (1996) 1-10 100
cant diffusion horizontally, as the effective diffusion constant is of order x 2 ( 6 × 10-5 m) 2 Deff=~t--~ ( 2 ) ( 1 0 2 s ) = 2 X 1 0 - 1 1 m 2 s - l
(4)
which is a plausible value for calcein moving within a charged matrix. After a few minutes without pulsing the concentration of the calcein in the localized surface regions has decreased. Upon resuming pulsing, the first pulses of a new pulsing series appear to be necessary to load the pathway region again with calcein, to achieve the concentration consistent with the quasisteady state. This hypothesis is supported by the measured changes in the passive electrical properties. Usually a decrease of R~kJn is observed during the lag time, and the decrease in R,ki, reaches a quasi-steady state at about the same time as the calcein flux. During the time of no pulsing Rskjn is mostly unchanged, which suggests that the new aqueous pathways (pores) still exist, and the pulses within the lag time are not needed to create them. For this reason, the calcein flux itself is not a suitable parameter for determining membrane recovery. This can be understood by noting that enhanced transport requires both additional aqueous pathways and an appropriate driving force. Electrical measurements respond relatively rapidly, because bulk charge relaxation times are of order 10- 9 s, and even the distributed tissue relaxation times (of order 10- 6 s) are extremely fast compared to molecular relaxation times. This means that measurement of electrical properties can monitor the presence of aqueous pathways that admit small ions such as Na + and C I - . Because of the 'Born energy repulsion' associated with the entry of charged species into pathways through lipid systems [ 17,20,31 ], it is a priori unlikely that all aqueous pathways that admit small ions will also admit molecules, although this may be the case for Small molecules such as calcein. When compared to the passive electrical properties, the average steady state flux tracks Gskin(t) = 1/Rskin, but the recovery of Rski, takes place on a time scale of minutes to hours, and in the case of very large pulses ( Usun,O> 200 V) insignificant recovery was observed. During the onset phase the flux increases with an exponentially decaying approach (time constant r~ag) until the quasi-steady state is reached. The lag time is
O O
'~E 10-1
=*
10-~ o
E
10
1.5
2
2.5
3
3.5
4
4.5
5
5.5
6
Gdy [mS]
Fig. 7. Calcein flux during the quasi-steady state presented as a function of the dynamicconductance,GDy= dl/dU~k~n,as determined by electrical measurements which began 100/xs after triggering the pulse.
a linear function of the pulse rate (fpulse = 1 Iti,,) for the pulse spacing used here (ti,t = 5 s to 1 min). That means it is possible to control ~'lagby changing the pulse spacing. The potential advantage for transdermal drug delivery is that by changing fpu~e rather than Uskm,O, reversible electroporation of the skin can be used. As reported by others [32], the amount of transported charged molecules is nearly proportional to the iontophoretic current. The following can be inferred from the dynamic skin conductance, GDy ~ d//dUski., which changes dramatically as Uski, increases. Typically GDy saturated at a value of 6 X 10 -3 S for our skin specimens with area = 7 X 10- ~ m -2, at higher voltages ( Uskin ~ 80 V ) is almost purely resistive. If these values are reached, the calcein flux becomes essentially independent of the dynamic conductance (Fig. 7). This means that a linear relationship between current and voltage has been established. In the voltage range investigated here for ti,t, the voltage-flux relationship was also determined, and found to be linear above a threshold of Uskin ~ 80 V. In summary, these results support the idea that electrical measurements and electrical control of the molecular transport of charged molecules are promising approaches for transdermal drug delivery. Control of two important quantities has been demonstrated: the magnitude of a charged molecule flux, and the lag time before a quasi-steady state is reached.
U. Pliquett, J.C. Weaver~Journal of Controlled Release 38 (1996) 1-10
Acknowledgements W e t h a n k M.R. P r a u s n i t z , R.O. Potts, R. L a n g e r , E.A. Gift, Y. C h i z m a d z h e v , a n d V . G . B o s e f o r m a n y h e l p f u l a n d s t i m u l a t i n g d i s c u s s i o n s , a n d also T.P. S i n g h a n d C. Liu for h e l p in p r e p a r i n g the skin, a n d N D R I for t i s s u e a c q u i s i t i o n , T h i s w o r k w a s s u p p o r t e d b y the Deutsche Forschungsgemeinschaft (U.P.), and NIH Grants GM34077 and ARH4921, Army Research Office G r a n t No. D A A L 0 3 - 9 0 - G - 0 2 1 8 , a n d a r e s e a r c h c o n t r a c t f r o m C y g n u s T h e r a p e u t i c S y s t e m s , Inc. to J.C.W.
References [1] J. Hadgraft and R.H. Guy, Transdermal drug delivery: Developmental Issues and Research Initiatives, Marcel Dekker, New York, 1989. [2] R. Langer, New methods of drug delivery, Science 249 (1990) 1527-1533. [ 3 ] C. Cullander and R.H. Guy, Transdermal delivery of peptides and proteins, Adv. Drug Deliv. Rev. 8 (1992) 291-329. [4 ] J.A. Wolff, Gene Therapeutics, Birkh~iuser, Boston, 1994. [5] H.I. Maibach and E.W. Smith, Percutaneous Penetration Enhancers, CRC Press, Boca Raton, FL, 1995. [6] M.R. Prausnitz, V.G. Bose, R.S. Langer and J.C. Weaver, Transdermal drug delivery by electroporation, Abstr. Proc. Int. Symp. Control. Release Bioact. Mater. 19, Controlled Release Society, July 26-29, Orlando, FL, 1992, pp. 232-233. [7] M.R. Prausnitz, V.G. Bose, R. Langer and J.C. Weaver, Transtissue molecular transport due to electroporation of skin, in: M. Blank (Ed.), Electricity and Magnetism in Biology and Medicine, San Francisco Press, 1993, pp. 122-124. [8] M.R. Prausnitz, V.G. Bose, R. Langer and J.C. Weaver, Electroporation of mammalian skin: A mechanism to enhance transdermal drug delivery, Proc. Natl. Acad. Sci. USA 90 (1993) 10504-10508. [9] M.R. Prausnitz, U. Pliquett, R. Langer and J.C. Weaver, Rapid temporal control of transdermal drug delivery by electroporation, Pharm. Res. 11 (1994) 1834-1837. [ 10] D. Bommannan, J. Tamada, L. Leung and R. Potts, Effect of electroporation on transdermal iontophoretic delivery of luteinizing hormone releasing hormone (LHRH) in vitro, Pharm. Res. 11 (1994) 1809-1814. [11] R. Vanbever, N. Lecouturier and V. Pr~at, Transdermal delivery of metoprolol by electroporation, Pharm. Res. 11 (1994) 1662-1665. [12] J.C. Weaver, Electroporation: A general phenomenon for manipulating cells and tissue, J. Cell. Biochem. 51 (1993) 426--435. [ 13 ] Yu.A. Chizrnadzhev, V. Zarnytsin, J.C. Weaver and R.O. Potts, Mechanism of electroinduced ionic species transport through a multilamellar lipid system, Biophys. J. 68 (1995) 749-765.
9
[ 14] T.Y. Tsong, Electroporation of cell membranes, Biophys. J. 60 ( 1991 ) 297-306. [ 15] S. Orlowski and L.M. Mir, Cell electropermeabilization: a new tool for biochemical and pharmacological studies, Biochim. Biophys. Acta 1154 (1993) 51-63. [ 16] J.C. Weaver, Electroporation in cells and tissues: A biophysical phenomenon due to electromagnetic fields, Radio Sci. 30 (1995) 205-221. [ 17] J.C. Weaver and A. Barnett, Progress towards a theoretical model of electroporation mechanism: Membrane electrical behavior and molecular transport, in: D.C. Chang, B.M.Chassy, J.A. Saunders and A.E. Sowers (Eds), Guide to Electroporation and Electrofusion, Academic Press, New York, 1992, pp.91-117. [ 18] V.A. Klenchin, S.I. Sukharev, S.M. Serov, L.V. Chernomordik and Yu.A. Chizmadzhev, Electrically induced DNA uptake by cells is a fast process involving DNA electrophoresis, Biophys. J. 60 (1991) 804-811. [ 19] S.I. Sukharev, V.A. Klenchin, S.M. Serov, L.V. Chernomordik and Y.A. Chizmadzhev, Electroporation and electrophoretic DNA transfer into ceils, Biophys. J. 63 (1992) 1320--1327. [20] M.A. Wang, S.A. Freeman, V.G. Bose, S. Dyer and J.C. Weaver, Theoretical modelling of electroporation: Electrical behavior and molecular transport, in: M. Blank (Ed.), Electricity and Magnetism in Biology and Medicine, San Francisco Press, 1993, pp. 138-140. [ 21 ] U. Pliquett, M.R. Prausnitz, Y. Chizmadzhev and J.C. Weaver, Measurement of rapid release kinetics for drug delivery, Pharm. Res. 4 (1995) 546--553. [22] C.L. Gummer, The in vitro evaluation of transdermal delivery, in: J. Hadgraft and R.H. Guy (Eds), Transdermal Drug Delivery: Developmental Issues and Research Initatives, Marcel Dekker, New York, 1989, pp. 177-186. [23] U. Pliquett, R. Langer and J.C. Weaver, The change in the passive electrical properties of skin due to electroporation, Biochim. Biophys. Acta, (1995) in press. [24] K.C. Madison, D.C. Swartzendruber, P.W. Wertz and D.T. Downing, Presence of intact intercellular lipid lamellae in the upper layers of the stratum comeum, J. Invest. Dermatol. 88 (1987) 714-718. [25] P.M. Elias and G.K. Menon, Structural and lipid biochemical correlates of the epidermal permeability barrier, Adv. Lipid. Res. 24 (1991) 1-26. [26] S.E. Friberg, L.B. Goldsmith, I. Kayali and H. Suhaimi, Lipid structure of stratum corneum, in: M. Bender (Ed.), Interfacial Phenomena in Biological Systems, Marcel Dekker, New York, 1991, pp. 3-32. [27] R.O. Potts, Ion transport through biomembranes, skin and other tissues: Applications to drug delivery, in: M. Blank (Ed.), Electricity and Magnetism in Biology and Medicine, San Francisco Press, 1993, pp. 32-37. [28] R. Lieckfeldt, Bedeutung von Lipidzusammensetzung und interner Morphologie des Stratum Corneums ftir seine Barriereeigenschaften, Inaugural-Dissetation zur Erlangung der Doktorwtirde der Naturwissenschaftlich Mathematischen Gesamtfakultat der Ruprecht-KarlsUniversitat Heidelberg, 1993.
10
U. Pliquett, J.C. Weaver~Journal of Controlled Release 38 (1996) 1-10
[29] S.A. Freeman, M.A. Wang and J.C. Weaver, Theory of electroporation for a planar bilayer membrane: Predictions of the fractional aqueous area, change in capacitance and porepore separation, Biophys. J. 67 (1994) 42-56. [ 30] U.F. Pliquett, T.E. Zewert, T. Chen, R. Langer and J.C. Weaver, Imaging of fluorescent molecule and small ion transport through human stratum corneum during high-voltage pulsing: Localized transport regions are involved, J. Biophys. Chem. (1995) in press.
[31] V.A. Parsegian, Energy of an ion crossing a low dielectric membrane: Solutions to four relevant electrostatic problems, Nature 221 (1969) 844-846. [32] R.R. Burnette, Iontophoresis, in: J. Hadgraft and R.H. Guy (Eds), Transdermal Drug Delivery: Developmental Issues and Research Initiatives, Marcel Dekker, New York, 1989, pp. 247-291. [ 33 ] R.P. Haugland, Handbook of Fluorescent Probes and Research Chemicals, Molecular Probes, Eugene, OR, 1992.
