Transport properties and association behaviour of the zwitterionic drug 5-aminolevulinic acid in water

European Journal of Pharmaceutical Sciences 21 (2004) 347–350

Transport properties and association behaviour of the zwitterionic drug 5-aminolevulinic acid in water A precision conductometric study Nadia Merclin a,∗ , Per Beronius b a

Department of Pharmacy, Physical Pharmaceutical Chemistry, Uppsala Biomedical Centre, Uppsala University, P.O. Box 580, SE-751 23 Uppsala, Sweden b Department of Analytical Pharmaceutical Chemistry, Uppsala Biomedical Centre, Uppsala University, P.O. Box 574, SE-751 23 Uppsala, Sweden Received 22 July 2003; received in revised form 10 October 2003; accepted 30 October 2003

Abstract The behavior of the hydrochloride salt of 5-aminolevulinic acid (ALA-HCl) with respect to transport properties and dissociation in aqueous solution at 25 ◦ C has been studied using precision conductometry within the concentration range 0.24–5.17 mM. The conductivity data are interpreted according to elaborated conductance theory. The carboxyl group appears to be, in practice, undissociated. The dissociation constant, Ka , of the NH3 + form of the amino acid molecules is determined to 6.78 × 10−5 (molarity scale); pKa = 4.17. The limiting molar conductivity of the ALA-H+ ion, λ0 = 33.5 cm2 −1 mol−1 ; electric mobility u = 3.47 × 10−4 cm2 V−1 s−1 , is close to the electric mobilites of the acetate and benzoic ions. © 2003 Elsevier B.V. All rights reserved. Keywords: 5-Aminolevulinic acid hydrochloride (ALA-HCl); Zwitterionic drugs; Precision conductometry; Electric ionic mobility; Iontophoretic conditions (prerequisites)

1. Introduction Photodynamic therapy (PDT) is a new type of treatment for basal cell carcinoma (BCC). It involves exogenous administration of 5-aminolevulinic acid, H2 N–CH2 – CO–CH2 –CH2 –COOH, abbreviated ALA, to enhance the endogenous synthesis of protoporphyrin IX (PpIX), (Rhodes et al., 1997). Unfortunately, the penetration depth into deep tissue layers is shallow resulting in insufficient local bioavailability of the drug in order to achieve full therapeutic effect. Within the frame of a program intended to improve the treatment of basal cell cancer (a common type of skin cancer) by iontophoretic administration of ALA, a precision conductometric investigation of the hydrochloride of this drug, aminolevulinic acid hydrochloride (ALA-HCl), in aqueous solution was undertaken. The main purpose was to obtain information about transport properties and association behaviour of this zwitterionic drug in pure water as solvent medium. ∗ Corresponding author. Tel.: +46-18-471-4371; fax: +46-18-471-4377. E-mail address: [email protected] (N. Merclin).

0928-0987/$ – see front matter © 2003 Elsevier B.V. All rights reserved. doi:10.1016/j.ejps.2003.10.027

The formula of ALA above indicates that the end groups can form COO− and NH3 + ions, the extent of which depends on the pH of the solution. As reviewed in the monograph of Mysels (1959) the carboxyl group tends to ionize at pH values above about 4 and the amino group below pH about 10. Hence, in acidic solution the ALA molecule would be positively charged because of the NH3 + and COOH groups, and in basic solution negatively charged because of the NH2 and COO− groups. In the evaluation of the conductance data we tentatively neglected the effect of the slight dissociation of the carboxyl group. We therefore started with the assumption that the only equilibrium to have to be considered is that between ALA-H+ cations, electrically neutral ALA molecules, and H+ ions.

2. Experimental 2.1. Reagents ALA-HCl (approximately 98%) was obtained from Sigma, St. Louis, MO, USA. The conductivity, κ, of the Milli-Q water used as solvent, determined as described fur-

348

N. Merclin, P. Beronius / European Journal of Pharmaceutical Sciences 21 (2004) 347–350 77.2

R/

77

76.8

76.6

76.4 0

0.0002

0.0004 -1

/Hz

0.0006

-1

Fig. 2. Cell resistance vs. reciprocal frequency of a 6 mM aqueous ALA-HCL solution at 25 ◦ C.

ther, varied between 2.77×10−6 and 3.01×10−6 −1 cm−1 . Potassium chloride (Merck, suprapur), used to determine the cell constant, was dried for 3 h at 130 ◦ C. Solutions were prepared by weight and corrections for the buoyancy effect of the air applied. Molalities were converted to molarities using the density 0.99707 g cm−3 of water at 25 ◦ C. Because of the low concentrations of ALA-HCl in the measurements this density was used also in calculating molarities of solutions. 2.2. Equipment Conductivities were determined using a Leeds and Northrup 4666 conductivity bridge connected to a Princeton Applied Research Model 129 A two phase lock-in amplifier, a Hewlett Packard Model 201C audio oscillator, and a hp 11473 B balancing transformer. A detailed account of the principle of the method can be found e.g. in Chapter 5 of the monograph of Robinson and Stokes (1965). The conductometric cell (Daggett et al., 1951) of 400 ml capacity, was fitted with platinized platinum electrodes (Fig. 1). The cell constant, determined by five calibrations with aqueous potassium chloride, was 0.056199 (0.047% S.D.) cm−1 .

68 67.5

R = -1,2177t + 95,4869 r 2 = 0,9987

67 66.5

R/ Ω

Fig. 1. Schematic representation of the conductivity cell used. The resistance of the ALA-HCL solution between two platinum electrodes (a), 40 mm in diameter and separated 1 cm from each other, is measured. The electrodes, connected via platinum wires (not shown in the figure) to the conductivity bridge, are located to the bottom of the cell.

Approximately 300 ml of water was transferred to the cell and 5–20 ml portions of a stock ALA-HCl solution added using a Metrohm Herisau calibrated Dosimat E 535 buret kept at 25 ± 0.5 ◦ C. The cell solution was agitated by means of a magnetic stirrer. The temperature of the solution, which changes slowly during the measurements, was determined using a mercury-in-glass certified thermometer graduated to 0.01 ◦ C. After each addition the cell resistance was determined at five different frequencies between 2 and 5 kHz to enable extrapolation of the resistance to infinite frequency. Fig. 2 shows a typical graph of the dependence of resistance, R, on the reciprocal frequency, ν. To enable resistance values to be corrected to 25.00 ◦ C the temperature dependence of a 6 mM ALA-HCl solution was measured at several temperatures within the range 23–25 ◦ C. According to Fig. 3 the resistance depends linearly on temperature within this range (correlation coefficient r ≈ 1); temperature coefficient 1.87%/◦ C at 25 ◦ C. After correcting the conductivity of the solution for the solvent conductivity (maximun correction 5.3%), conductivities, κ, were converted to molar conductivities, Λ.

66 65.5 65 64.5

2.3. Conductivity measurements

22.5

23

23.5

24

24.5

25

25.5

t/˚C

Measurements of conductivites were performed at several ALA-HCl concentrations in the range 0.24–5.17 mM.

Fig. 3. Resistance–temperature graph of a 6 mM ALA-HCL aqueous solution; temperature coefficient 1.87%/◦ C at 25 ◦ C.

N. Merclin, P. Beronius / European Journal of Pharmaceutical Sciences 21 (2004) 347–350 Table 1 Dependence of molar conductivity of ALA-HCl in water on concentration at 25.0 ◦ C

349

Denoting the thermodynamic dissociation constant, Ka , of ALA-H+ , we have, a(ALA)a(H+ ) a(ALA-H+ )

104 (c/M)

Λ/(cm2 −1 mol−1 )

Ka =

Series 1: lower concentration range 2.4021 2.9559 3.4517 4.1779 4.8969 6.1830

235.23 226.91 220.69 213.20 207.27 198.57

where a stands for activity. Assuming acitivity coefficients of these univalent ions to be equal and that of the electrically neutral ALA molecule to be equal to unity, the dissociation constant can be expressed in terms of concentrations: Ka =

Series 2: higher concentration range 4.7034 9.2953 13.701 17.973 22.322 27.145 31.839 37.971 44.012 51.737

207.96 183.69 171.41 163.46 157.51 153.01 148.75 144.73 141.60 140.96

(2)

c(ALA)c(H+ ) cα2 = c(ALA-H+ ) (1 − α)

(3)

where c is the analytical ALA-HCl concentration and α the degree of dissociation of ALA-H+ . The molar conductivity is expressed, Λ = m[λ0 (H+ )α + λ0 (ALA-H+ )(1 − α) + λ0 (Cl− )]

(4)

The factor, m, is introduced to correct for the change in ionic mobility with the charge density of the solution. According to the “FHFP” equation (Fuoss and Hsia, 1967, 1968, Fernandez-Prini, 1969) the dependence of this correction factor on the concentration, ci , of free ions of a univalent electrolyte can be expressed,

3. Calculations and results In Table 1 molar conductivities of two series of measurements have been compiled. A graphic representation of these data, are shown in Fig. 4, where the molar conductivity has been plotted as a function of the ALA-HCl concentration. 3.1. Interpretation of the conductance curve As suggested above we will interpret our conductance data on the assumption that the only equilibrium necessary to be considered is that between ALA-H+ ions, electrically neutral ALA molecules and protons, i.e. ALA-H+ ⇔ ALA + H+

1/2

m=

Λ0 − Sci

3/2

+ Eci log ci + J1 ci − J2 ci Λ0

The coefficients S and E depend on the limiting molar conductivity, Λ0 , the temperature, and solvent properties (dielectric constant, ε, and viscosity, η). The coefficients J1 and J2 depend, in addition, on the maximum distance between the charges of paired ions. This parameter was set equal to the Bjerrum radius, which for univalent electrolytes in aqueous solution at 25 ◦ C is equal to 0.357 nm. Values of limiting molar ionic conductivity, dielectric constant, and viscosity of the solvent at 25 ◦ C were taken from the monograph of Robinson and Stokes, 1965 λ0 (H+ ) =

(1) 1.8

450

1.4

σ(Λ)/%

400

Λ/cm2 Ω -1 mol -1

350

1

300 250

0.6 200 150

0.2 26

100

28

30

32

34

36

38

40

42

λ0(ALAH+)/cm2Ω-1 mol-1

50 0

0.001

0.002

0.003 c/M

0.004

0.005

0.006

Fig. 4. Molar conductivity–concentration graph of aqueous ALA-HCL at 25 ◦ C. Best fit for λ0 (ALA-H+ ) = 33.5 cm2 −1 mol−1 and pKa = 4.17.

Fig. 5. Dependence of relative standard deviation in Λ on λ0 (ALA-H+ ) for different choices of acid dissociation constant, Ka , of ALA-H+ . From the left in the graph the curves correspond to the following choices of 106 Ka : 74, 72, 70, 67.8, 66, 64, 62, respectively.

350

N. Merclin, P. Beronius / European Journal of Pharmaceutical Sciences 21 (2004) 347–350

Table 2 pKa for ␦-aminolevulinic acid hydrochloride in aqueous solution according to the present study Method used for pKa determination

pKa

Reference

Precision conductometry Potentiometry Spectrophotometry

4.17 3.90 4.05

This study Elfsson et al. (1998) Novo et al. (1996)

Potentiometry data of Elfsson et al. (1998) and spectrophotometry data of Novo et al. (1996) are included for comparison.

349.81; λ0 (Cl− ) = 76.35 cm2 −1 mol−1 ; ε = 78.3; η = 0.008903 P. Using a computer program, constructed in Excel, the values of Ka and λ0 (ALA-H+ ) resulting in the best fit of Eq. (4) to the experimental data in Table 1 were calculated. The best fit, represented by the fulldrawn curve in Fig. 4, corresponds to the dissociation constant, Ka (ALA-H+ ) = 6.78 × 10−5 (pKa = 4.17) and λ0 (ALA-H+ ) = 33.5 cm2 −1 mol−1 , Fig. 5, in which the relative standard deviation in molar conductivity, σ(Λ), has been plotted as a function of λ0 (ALA-H+ ) for a series of Ka -values.

for the assumption of negligible dissociation of the carboxyl group at the pH concerned. The electric mobility of the ALA-H+ ion, u = 3.45 × 10−4 cm2 V−1 s−1 , derived from the limiting ionic conductivity, λ0 (ALA-H+ ) = 33.5 cm2 −1 mol−1 , appears to be quite reasonable; it is slightly less than the mobility of the acetate ion, 4.24×10−4 cm2 V−1 s−1 and close to the mobility of the benzoate ion, 3.36 × 10−4 cm2 V−1 s−1 , Robinson and Stokes (1965). The agreement between the conductivity data according to the two, in part independent series of conductivity measurements in the lower and higher concentration range (series 1 and 2, respectively, in Table 1) may be taken as evidence for the reliability of the experimental procedure here employed. As to the determination of pKa values of zwitterionic compounds it may be of interest to mention that spectrophotometry was used by Ross and Riley (1990, 1992), in a study of the relationship between pKa and structure of fluoroquinolone antimicrobials and conductometry by Hatanaka et al., 2000, in an investigation of ion pair skin transport of the zwitterionic drug cephalexin. However, the literature survey did not reveal any previous publications of precision conductometry in combination with advanced conductance theory to determine pKa values of zwitterionic compounds.

4. Conclusions A condition for iontophoretic delivery of drugs is that the drug molecule is in an ionized state (positively or negatively charged). If a molecule is neutral the transport can be achieved either by introducing a charge into the molecule or by electroosmosis. The mobility is affected by several factors: concentration, polarity of the solvent used, interactions between the ionic species themselves, just to mention a few. The main purpose of our current research is to improve the treatment of basal cell cancer using iontophoresis to deliver ALA more efficiently into the skin. Extended knowledge of the parameters affecting the mobility of the drug is necessary to optimize the conditions in the iontophoretic process. The pKa here obtained, 4.17, agrees quite well with the potentiometrically determined value according to Elfsson et al., 1998, pKa = 3.90 and the spectrophotometrically one obtained by Novo et al., 1996, pKa = 4.05 (see Table 2). More extensive sets of potentiometric, spectrophotometric, and conductivity data would be necessary to obtain an indication if the slight differences in pKa might be interpreted as indicating a significant discrepancy according to these methods of pKa determinations. As can be seen in Fig. 4, the assumption that the only equilibrium to be considered is that of the dissociation of NH3 + at the amino end of the ALA molecule results in an excellent fit of Eq. (4) to the experimental conductivity data. The relative standard deviation of the single Λ-value, σ(Λ) = 0.60%. No systematic trend between experimental and computed conductivity values with the concentration of ALA-H+ was observed. This observation may be taken as further support

Acknowledgements The authors wish to thank Professor Hans Ehrsson, Karolinska Pharmacy, Stockholm, for his kind support of this research and for many valuable discussions. References Daggett Jr., H.M., Bair, E.J., Kraus, C.A., 1951. Properties of electrolytic solutions. XLVII. Conductance of some quaternary ammonium and other salts in water at low concentration. J. Am. Chem. Soc. 73, 799–803. Elfsson, B., Wallin, I., Eksborg, S., Rudaeus, K., Ros, A.M., Ehrsson, H., 1998. Stability of 5-aminolevulinic acid in aqueous solution. Eur. J. Pharm. Sci. 7, 87–91. Fernandez-Prini, R., 1969. Conductance of electrolyte solutions. A modified expression for its concentration dependence. Trans. Faraday Soc. 65, 3311–3313. Fuoss, R.M., Hsia, K.-L., 1967. Association of 1-1 salts in water. Proc. Natl. Acad. Sci. USA 57, 1550–1557. Fuoss, R.M., Hsia, K.-L., 1968. Association of 1-1 salts in water. Proc. Natl. Acad. Sci. USA 58, 1818. Hatanaka, T., Kamon, T., Morigaki, S., Katayama, K., Koizumi, T., 2000. J. Contr. Rel. 66, 63–71. Mysels, K.J., 1959. Introduction to Colloid Chemistry, Interscience Publishers Inc., New York. Novo, M., Hüttman, G., Diddens, H., 1996. J. Photochem. Photobiol. B Biol. 34, 143–148. Rhodes, L.E., Tsoukas, M.M., Anderson, R.R., Kollias, N., 1997. J. Invest. Dermatol. 108, 87–91. Robinson, R.A., Stokes, R.H., 1965. Electrolyte Solutions, second ed. Butterworths, London. Ross, D.L., Riley, C.M., 1992. Int. J. Pharm. 83, 267–272. Ross, D.L., Riley, C.M., 1990. Int. J. Pharm. 63, 237–250.