Subscriber access provided by MT ROYAL COLLEGE
Article
Experimental determination of drug diffusion coefficients in unstirred aqueous environments by temporally resolved concentration measurements Massimiliano Pio di Cagno, Fabrizio Clarelli, Jon Våbenø, Christina Lesley, Sokar Darsim Rahman, Jennifer Cauzzo, Erica Franceschinis, Nicola Realdon, and Paul C. Stein Mol. Pharmaceutics, Just Accepted Manuscript • DOI: 10.1021/acs.molpharmaceut.7b01053 • Publication Date (Web): 20 Feb 2018 Downloaded from http://pubs.acs.org on February 28, 2018
Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.
Molecular Pharmaceutics is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.
Page 1 of 25 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Molecular Pharmaceutics
1
Experimental determination of drug diffusion coefficients in unstirred aqueous
2
environments by temporally resolved concentration measurements
3 4
Massimiliano Pio di Cagnoa,*, Fabrizio Clarellib, Jon Våbenøc, Christina Lesleyd, Sokar
5
Darsim Rahmand, Jennifer Cauzzoa,e, Erica Franceschinise, Nicola Realdone and Paul C.
6
Steind
7 8
a
9
Arctic University of Norway, Tromsø, Norway
Drug Transport and Delivery Research Group, Department of Pharmacy, University of Tromsø, The
10
b
11
University of Norway, Tromsø, Norway
12
c
13
d
14
Denmark
15
e
Computational Pharmacology Group, Department of Pharmacy, University of Tromsø, The Arctic
Helgeland Hospital Trust, Sandnessjøen, Norway Department of Physics, Chemistry and Pharmacy, University of Southern Denmark, Odense,
Department of Pharmaceutical Sciences, University of Padova, Padova, Italy
16 17 18
*Corresponding author: Massimiliano Pio di Cagno; Tel: +47 776 45301; e-mail:
19
[email protected] 20
1 ACS Paragon Plus Environment
Molecular Pharmaceutics 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
21
Table of contents graphics
22
2 ACS Paragon Plus Environment
Page 2 of 25
Page 3 of 25 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Molecular Pharmaceutics
23
Abstract
24
The diffusion coefficient (also known as diffusivity) of an active pharmaceutical ingredient
25
(API) is a fundamental physicochemical parameter that affects passive diffusion through
26
biological barriers and, as a consequence, bioavailability and biodistribution. However, this
27
parameter is often neglected, and it is quite difficult to find diffusion coefficients of small
28
molecules of pharmaceutical relevance in the literature. The available methods to measure
29
diffusion coefficients of drugs all suffer from limitations that range from poor sensitivity to
30
high selectivity of the measurements or the need for dedicated instrumentation. In this work, a
31
simple but reliable method based on time resolved concentration measurements by UV-visible
32
spectroscopy in an unstirred aqueous environment was developed. This method is based on
33
spectroscopic measurement of the variation of the local concentration of a substance during
34
spontaneous migration of molecules, followed by standard mathematical treatment of the data
35
in order to solve Fick’s law of diffusion. This method is extremely sensitive and results in
36
highly reproducible data. The technique was also employed to verify the influence of the
37
environmental characteristics (i.e. ionic strength and presence of complexing agents) on the
38
diffusivity. The method can be employed in any research laboratory equipped with a standard
39
UV-visible spectrophotometer, and could become a useful and straightforward tool in order to
40
characterize diffusion coefficients in physiological conditions and help to better understand
41
the drug permeability process.
42
43
Keywords: Diffusion coefficients, UV-visible spectroscopy, Fick’s laws of diffusion, data
44
fitting, molecular modelling, ion strength
45
3 ACS Paragon Plus Environment
Molecular Pharmaceutics 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Page 4 of 25
46
1. Introduction
47
Diffusion is the spontaneous migration of molecules of a substance from regions of higher
48
chemical activity (high concentration) to regions of lower chemical activity (low
49
concentration) through one or several isotropic environments. This process is of capital
50
importance in pharmaceutics since passive transport of active pharmaceutical ingredients
51
(APIs) remains a significant (and often primary) mechanism of biological membrane
52
permeation (1) and therefore the diffusion process is crucial for reaching an appropriate
53
bioavailability. According to Stokes-Einstein equation (Eq. 1), the diffusivity D (cm2/sec) is
54
directly proportional to the absolute temperature T (K) and Boltzmann constant Kb
55
(cm2*g/s2*K) but inversely proportional to the viscosity of the media η (g/cm*sec) and the
56
hydrodynamic radius r (cm) of the molecules.
=
6
Equation 1
57
The relation between diffusion and flux of a drug substance in aqueous environment can be
58
described by Fick’s first law as:
= −
Equation 2
59
Where j represents the net flux (mol/cm2*sec) of drug substance through the media, dc/dx
60
represents the concentration gradient and D is the diffusion coefficient. D is an essential
61
parameter to understand pharmaceutically relevant processes such as mass transport through
62
barriers (i.e. absorption) and bioavailability of APIs. However, pharmaceutical research has
63
for years overlooked the importance of this parameter in favor of the measurement of the
64
apparent permeability coefficient (Papp, 2-4). Since water is the primary component of the
65
human body, a proper quantification of D values for APIs should be of crucial importance for
66
a proper evaluation of the permeability process. Reliable D values, together with a proper
67
evaluation of partition coefficients (LogP), could help to elucidate the role of unstirred water 4 ACS Paragon Plus Environment
Page 5 of 25 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Molecular Pharmaceutics
68
layers (UWL) (5) and interface transitions in passive transport through barriers (6). The
69
scientific literature is significantly lacking in diffusivity values for APIs in aqueous
70
environments, and methods to measure them in a simple but reliable manner. Currently,
71
several techniques for measuring diffusion coefficients are available, including holographic
72
laser interferometry (HLI (7)), electron speckle pattern interferometry (ESPI (8)), mass
73
transfer through barrier studies (9), Taylor dispersion analysis (TDA (10,11)), nuclear
74
magnetic resonance (NMR (12)), dual-focus fluorescence correlation spectroscopy (FCA
75
(13)) and fluorescence recovery after photobleaching (FRAP (14)) just to mention the most
76
employed ones. Just a few of these techniques (TDA, NMR and mass transfer studies) have
77
been applied to estimate diffusivity of APIs, the reason being that all the techniques listed
78
above show limitations when facing pharmaceutically relevant problems such as poor
79
sensitivity (e.g. NMR), high specificity (e.g. FRAP) and the need for dedicated
80
instrumentation (e.g. HLI, ESPI, FCA, TDA and NMR which are not standard
81
instrumentations in most pharmaceutical research laboratories).
82
Another approach to estimate diffusivities of small molecules in aqueous environment is to
83
predict them from molecular properties such as molecular volume, molar volume or by
84
molecular simulation (15). However, all these in silico approaches are prone to uncertainty
85
and do not provide a complete interpretation of diffusion in the case of environmental changes
86
such as ion strength or presence of complexing agents. In this work, a simple, versatile and
87
easily implemented spectroscopic method (16, 17) for diffusion coefficient measurement is
88
introduced. This versatile method is based on the measurement of variation of local
89
concentration of a compound in aqueous media and subsequent temporal resolution of
90
diffusion equation by means of analytical as well as numerical mathematical methods. By
91
comparing different approaches to the solution of Fick’s diffusion equation we show that
92
simple non-linear least-squares fitting is sufficient to extract the diffusion coefficient. The
5 ACS Paragon Plus Environment
Molecular Pharmaceutics 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
93
experiments are performed employing a regular quartz cuvette and a standard UV-visible
94
spectrometer. This method is simple, reliable, reproducible and extremely sensitive.
95
Moreover, it has the significant advantage that it could be employed in any research
96
laboratory equipped with a standard UV-visible spectrophotometer and very small changes
97
related to environmental alterations can be measured.
98
2. Experimental section
99
2.1 Materials
100
Sodium-di-hydrogen-phosphate monohydrate (NaH2PO4·H2O), di-sodium hydrogen-
101
phosphate dodeca-hydrate (Na2HPO4·12H2O), sodium chloride (NaCl), sodium hydroxide
102
(NaOH), hydrochloric acid (HCl, ≥ 37%), caffeine, calcein, chloramphenicol, ketoprofen,
103
nitrofurantoin, tetracycline hydrochloride, trimethoprim, vancomycin, and β-cyclodextrin
104
(βCD) were all purchased from Sigma Aldrich Chemie GmbH (Steinheim, Germany).
105
Paracetamol was purchased from Norsk Medisinaldepot AS (Oslo, Norway). Penicillin G
106
employed was USP Reference Standard (Timbrook, USA). All chemicals were of analytical
107
grade.
108
2.2. UV–visible localized spectroscopy
109
2.2.1 Preparation of the solutions
110
Phosphate buffer saline (PBS) 73 mM was prepared mixing a solution of NaH2PO4·H2O
111
(2.2% W/V) with a solution of Na2HPO4·12H2O (1.8 % W/V) in a ratio 1 to 5. The pH of the
112
PBS was adjusted to 7.3-7.4 (pH meter Lab 744, Metrohm AG, Herisau, Switzerland) by the
113
addition of NaOH solid pellets. Tonicity of the PBS solution was adjusted to 280-290 mOsm
114
(Semi-Micro Osmometer K-7400, Knauer, Berlin, Germany) by the addition of NaCl. Each of
115
the compounds was than adequately weighted and dissolved in the PBS in order to achieve a
116
concentration of approximately 0.5-1.5 mM. The molecular weights of the compounds and the
117
characteristics of the donor solutions (compound concentrations, buffer concentrations,
6 ACS Paragon Plus Environment
Page 6 of 25
Page 7 of 25 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Molecular Pharmaceutics
118
solution densities, wavelengths of maximal absorption (λMAX) and specific absorptivities (ε))
119
are reported in Table 1.
120
Table 1. General characteristics and composition of the aqueous solutions used as donor
121
environments. The diffusional medium was composed of plain distilled water (DW, density of
122
1.0015 g/mL). Each of the compounds was detected at its maximum wavelength of absorption
123
(λMAX) and the local concentration was calculated using its specific absorptivity (ε). Nominal conc. (mM)
Buffer conc. (mM)
Density*
λMAX
ε
name
Molecular Weight (g/mol)
(g/mL)
(nm)
(cm2/µmol)
DW
Caffeine
194.2
1.23
72.8
1.0086
269
11.3
DW
Calcein
622.6
1.23
72.8
1.0083
294
11.3
DW
Chloramphenicol
323.2
1.20
72.8
1.0095
277
10.0
DW
Ketoprofen
254.3
1.17
72.8
1.0064
257
18.4
DW
Nitrofurantoin
238.2
0.77
72.8
1.0113
274
13.7
DW
Paracetamol
151.2
1.42
72.8
1.0079
236
13.2
DW
Penicillin G
334.4
0.60
72.8
1.0071
207
19.9
DW
Tetracycline
444.4
0.83
72.8
1.0197
361
14.8
DW
Trimethoprim
290.3
1.19
72.8
1.0085
278
5.6
DW Vancomycin 1449.3 1.20 72.8 *Data are reported as mean of three measurements. SD below 0.5%
1.0011
280
6.7
124 125
Solutions of βCD (used in the diffusion experiments to investigate complexation effects) were
126
prepared by adding solid powder of the dextrin derivative (3.4-4 mg) directly into 2 mL of
127
ketoprofen neutral and acidic solution (pH 1.0, obtained by mixing 1.9 mL of neutral
128
ketoprofen solution plus 0.1 mL of HCl) in order to achieve a molar ratio drug:βCD of
129
approximately 1:1.
130
2.2.2 Experimental procedure
131
For the spectrophotometric measurements, a double array VWR (VWR International, Radnor,
132
USA) UV-visible spectrophotometer (model UV-6300 PC) equipped with a Hellma®
133
Suprasil® (Sigma-Aldrich) quartz absorption cuvettes (chamber volume of 700 µL and path
134
length of 10 mm) was employed. The experimental setup is schematized in Fig. 1.
Diffusional medium
Compound
7 ACS Paragon Plus Environment
Molecular Pharmaceutics 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
135 136
Figure 1. Schematic representation of localized UV-visible spectroscopy experimental setup.
137
Both reference and sample cuvette were filled with 675 µl of distilled water and placed in the
138
respective compartment of the spectrophotometer. At time (t) = 0 sec (starting of the
139
experiment), 25 µL of drug solution was gently injected in the bottom of the sample cuvette
140
by a needle syringe. The reference and sample cuvette were both sealed with parafilm in order
141
to avoid evaporation of the solvent. Absorbance readings were recorded at fixed wavelength
142
(Table 1) at regular time intervals (120 sec) for 24 hours at room temperature (23-24°C).
143
Absorbance was recorded at a specific position (x1) corresponding to 0.51 cm from the bottom
144
of the cuvette (Figure 2). This was obtained by keeping the sample cuvette lifted 0.60 cm
145
(Hard Head calliper, Jula Norge AS, Lørenskog, Norway) from the bottom of the cuvette
146
holder.
8 ACS Paragon Plus Environment
Page 8 of 25
Page 9 of 25 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Molecular Pharmaceutics
147 148
Figure 2. Adaptation of the mono-dimensional diffusional model used to the cuvette
149
geometry and related relevant spatial points (initial point x0, measurement points x1 and x2
150
and total length of the diffusion chamber L).
151
For each of the API studied, two independent experiments were conducted (n=2) and the final
152
diffusional profile (employed for the mathematical data treatment and diffusivity calculations)
153
was obtained by averaging the respective absorbances at each time point. The average
154
variance (in percentage, %) between the absorbance registered at each time point between the
155
two parallel experiments was used to estimate the experimental error (≤ 5 %).
156
2.2.3 Mathematical model and UV-visible spectroscopy data interpretation
157
The mathematical approach used in this work to fit the experimental data obtained by UV-
158
visible spectroscopy to calculate the diffusion coefficients was based on the standard diffusion
159
equation (Eq. 3) and well-known classical physical models (9, 18). Due to the geometry of the
160
experimental setup (Figure 2), the diffusion process was approximated with a one-
161
dimensional model. The diffusion equation (Eq. 3) describes the concentration of a solute
162
over time and space in a homogenous medium (assuming constant diffusivity) as:
9 ACS Paragon Plus Environment
Molecular Pharmaceutics 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
, , =
Page 10 of 25
Equation 3
163
Where c represents the concentration of the substance, t the time, x the position and D the
164
diffusivity. In theory, any x between 0.51 cm and 1.11 cm can be selected (Fig. 2) by lifting
165
the cuvette from its standard recording position. In this work, the recording point was
166
primarily set to 0.51 cm (see experimental procedure section).
167
The first mathematical approach used to find numerical solution of D was analytical. This
168
approach is based on the assumption that for times t, t