Article pubs.acs.org/Langmuir
Phase Transitions in DNA/Surfactant Adsorption Layers Vanda V. Lyadinskaya,† Shi-Yow Lin,† Alexander V. Michailov,‡ Alexey V. Povolotskiy,‡ and Boris A. Noskov*,‡ †
National Taiwan University of Science and Technology, Chemical Engineering Department, 43 Keelung Road, Section 4, 106 Taipei, Taiwan ‡ Institute of Chemistry, St. Petersburg State University, Universitetsky pr. 26, 198504 St. Petersburg, Russia S Supporting Information *
ABSTRACT: The adsorption layers of complexes between DNA and oppositely charged surfactants dodecyltrimethylammonium bromide (DTAB) and cetyltrimethylammonium bromide (CTAB) at the solution/air interface were studied with surface tensiometry, dilational surface rheology, atomic force microscopy, Brewster angle microscopy, infrared absorption−reflection spectroscopy, and ellipsometry. Measurements of the kinetic dependencies of the surface properties gave a possibility to discover the time intervals corresponding to the coexistence of two-dimensional phases. One can assume that the observed phase transition is of the first order, unlike the formation of microaggregates in the adsorption layers of mixed solutions of synthetic polyelectrolytes and surfactants. The multitechniques approach together with the calculations of the adsorption kinetics allowed the elucidation of the structure of coexisting surface phases and the distinguishing of four main steps of adsorption layer formation at the surface of DNA/ surfactant solutions.
■
INTRODUCTION The formation of complexes between DNA and oppositely charged amphiphile molecules has attracted significant attention because of their use in the course of DNA extraction,1 purification,2 the preparation of highly organized hybrid bioinorganic nanostructures,3 and the construction of DNA chips.4 The application of the complexes for nonviral gene transfer is probably of special importance.5−7 The corresponding technological and biomedical interest resulted in extensive investigations of DNA−surfactant interactions in aqueous solutions.7−14 These interactions at liquid−fluid interfaces have attracted less attention, although the layers of amphiphilic molecules at the boundary between two fluid phases are frequently used as simple physical models of biomembranes. In spite of the obvious difference in the two systems, this approach has a long history and continue to be important nowadays.15,16 Note that important steps of gene delivery into the cells can include DNA incorporation into liposomes and the interaction of DNA carriers with cellular membranes. A relatively large number of studies are devoted only to DNA adsorption from solutions onto spread monolayers of insoluble lipids or surfactants.17−24 These model systems give a possibility to choose the components and optimal composition of the complex gene vector.23,24 The simultaneous adsorption of DNA and cationic surfactants at the liquid−gas interface has been studied only by a few authors until now.25−32 At the same time, the surface properties of mixed solutions of synthetic polyelectrolytes and oppositely charged surfactant have been intensively studied during the last two decades.33−36 These efforts led to relatively detailed models of the adsorption © 2016 American Chemical Society
layers of the complexes between some synthetic polyelectrolytes and surfactants. In the case of DNA/surfactant adsorption layers, the problem proved to be more complicated, and some conclusions on the layer structure are still based to a significant extent on surface tension measurements.30,32 The application of the neutron reflection method by two groups led to somewhat different conclusions probably because of the strong dependence of the surface properties on the DNA molecular weight, the solution ionic strength, and the possible influence of residual proteins in the sample.27,28 Moreover, different values of the adsorption layer thickness determined by the ellipsometry and the neutron reflection method forced the authors to assume that the latter technique was not sensitive enough to “highly hydrated bulk complexes adsorbing at the surface” because of unfavorable contrast conditions.28 It is noteworthy that almost all of the obtained results on DNA/surfactant adsorption layers relate to systems almost at equilibrium. Very recently, Moradi et al. have reported experimental results on the dynamic dilational surface elasticity of DNA solutions in mixtures with azobenzene-containing cationic surfactant but do not give any information on the kinetic dependencies of this quantity.37 Only McLoughlin and Langevin measured the kinetic dependencies of the surface tension and adsorbed amount at a few concentrations.26 At the same time, it has been shown recently that the kinetic dependencies of the dynamic surface elasticity of polyelecReceived: September 14, 2016 Revised: November 23, 2016 Published: November 23, 2016 13435
DOI: 10.1021/acs.langmuir.6b03396 Langmuir 2016, 32, 13435−13445
Article
Langmuir
of the oscillations were kept constant and equalled 0.1 Hz and 3%, respectively. The induced surface tension oscillations were recorded by the Wilhelmy plate method. The complex dynamic surface elasticity ε was then calculated according to the relation ε(ω) = δγ/δ ln A, where δγ and δA are the increments of the surface tension and surface area, respectively. If the phase shift between the oscillations of the surface area and surface tension is known, then it is possible to determine the real and imaginary parts of the dynamic surface elasticity, which is a complex quantity. All of the measurements of surface properties were started after purification of the surface (t = 0) using a movable barrier on the surface of the Langmuir trough and a Pasteur pipet connected to a pump.41 All of the measurements of the dynamic surface properties were carried out at 20 ± 0.5 °C. Atomic Force Microscopy. The equilibrium adsorption DNA/ CTAB films were transferred from the solution surface onto a freshly cleaved mica plate by the Langmuir−Blodgett technique and investigated by AFM using an NTEGRA Prima setup (NT-MDT, Russia) with a tip curvature radius of 10 nm. The constant surface tension was a criterion of the formation of an equilibrium adsorption layer at the liquid surface. The subphase was gently replaced by 20 mM NaCl/10 mM Tris-HCl solution in 3 h after the formation of a new liquid surface. The mica plate with the transferred DNA/CTAB layer was stored for 1 day in a desiccator at 4 °C. The semicontact regime was used for all of the measurements. NOVA software was used in the course of microscope operation and image analysis. Brewster Angle Microscopy. The morphology of DNA/surfactant solutions was investigated in situ by the BAM1 Brewster angle microscope (NFT, Göttingen, Germany) with a spatial resolution of 4 μm and equipped with a 10 mW He−Ne laser. The solution under investigation was placed in the Langmuir trough and equilibrated at room temperature for 24 h before images were obtained. Ellipsometry. A Multiskop null ellipsometer (Optrel GBR, Germany) at a single wavelength of 632.8 nm was applied to estimate the adsorbed amount using a fixed compensator (45°) and a two-zone averaging nulling scheme. All of the ellipsometric measurements were performed at an angle of incidence of 50°, close to the Brewster angle. Elliptically polarized light consists of two components with the electric vectors oscillating parallel and perpendicular to the plane of incidence. The reflection of light at an interface results in different changes in the phase and amplitude of these two components. These changes depend on optical properties of the interface and can be characterized by two ellipsometric angles, a relative amplitude change Ψ, and a relative phase shift Δ, which are related to reflection coefficients rp and rs of the radiation polarized in plane and perpendicular to the plane of incidence, respectively,
trolyte/surfactant solutions can have local maxima corresponding to different steps in the adsorption layer formation.35,38,39 This effect allowed us to obtain information not only on the adsorption mechanism but also on some details of the adsorption layer structure. The first attempts to apply this approach to DNA/surfactant solutions discovered strong distinctions from the properties of the mixed solutions of synthetic polyelectrolytes and oppositely charged surfactants and thereby required the use of a larger set of experimental techniques. The main aims of this study are to explain the observed peculiarities of the dynamic surface properties, to connect them with the characteristic features of the structure of the DNA/surfactant adsorption layer, and to elucidate the mechanism of layer formation. For these purposes, we applied ellipsometry, infrared reflection adsorption spectroscopy (IRRAS), and atomic force microscopy (AFM) together with surface dilational rheology and surface tensiometry.
■
EXPERIMENTAL SECTION
Materials. Single-chain cationic surfactants DTAB (Merck, Germany) and CTAB (Sigma-Aldrich, Germany) were recrystallized twice from an ethyl acetate/ethanol mixture. Calf thymus DNA (DNA, Mw = 107 Da, 16 kbp) containing 6% sodium salt was purchased from Sigma-Aldrich (D1501) and used without further purification. The DNA concentration was determined by UV spectroscopy (UV3600 spectrophotometer, Shimadzu) by considering the molar extinction coefficient to be ε260 = 6600 M−1 cm−1. Absorbance ratios A260/A230 and A260/A280 were measured in the range of 1.8−1.9, thereby confirming a double-helix structure of DNA and the absence of protein residuals and organic contaminants.40 Sodium chloride (NaCl) was preliminarily heated in a muffle furnace at about 750 °C for the elimination of organic impurities. Trizma base was purchased from Sigma and used as supplied. All stock solutions were prepared in Millipore water (resistivity 18.2 MΩ cm) produced by a Milli-Q purification system. Sample Preparation. DNA fibers were dissolved by weight in 10 mM Tris-HCl buffer solution at pH 7.6 containing 20 mM NaCl in order to keep the double helix intact. The stock DNA solution was stored at 2−4 °C for no longer than 2 weeks after preparation. The overlap concentration of 3.3 × 10−4 M (in nucleotide units) for this DNA sample was not exceeded in our experiments.31 The final concentration of DNA was 50 μM (in nucleotide units) unless stated otherwise. The concentration of cationic surfactants was varied over a wide range of 1 × 10−7−3 × 10−3 M. The sample preparation protocol adopted during all of the measurements of the dynamic surface properties was as follows. The DNA stock solution was diluted to the desired concentration just before the measurements. The surfactant stock solution was prepared in Millipore water. The solutions of DNA and surfactant of the double required concentration were then mixed in a flask and swirled gently for a few minutes before the start of measurements. The prepared solution was transferred to a Teflon trough, which was cleaned with a chromic mixture and washed carefully with Millipore water before use. Methods. Surface Tension and Dilational Surface Elasticity. The surface tension was measured by the Wilhelmy plate method using a rectangular glass plate connected to an electronic microbalance and sandblasted to ensure complete wetting. The plate was cleaned with a freshly prepared chromic acid mixture, washed with Millipore water, and dried before measurements. Just before the Wilhelmy plate touched the solution, the liquid surface was gently cleaned by a pipet attached to a water pump. The surface tension was recorded during 5 h for each sample. The reproducibility of the surface tension was about ±1 mN/m. A handmade setup for measurements of the surface dilational rheological properties by the oscillating barrier method was described in detail elsewhere.41 The Teflon barrier moved back and forth along polished brims of a Langmuir trough, creating periodic expansions/ compressions of the liquid surface area. The frequency and amplitude
rp rs
= tan Ψe iΔ
For a sharp boundary between two homogeneous phases, the reflection coefficients and hence Ψ and Δ are determined by the Fresnel reflection coefficients. One can derive corrections to the Fresnel relations by taking into account a thin layer between the two phases with a thickness d that is much less than the wavelength of radiation λ. For nonabsorbing media, a correction to Δ is of first order in a small parameter d/λ, but Ψ has the correction of second order in this parameter. We will consider below the angle Δ only because the light absorbance at the wavelength of 632.8 nm is negligible in the system under investigation. In the model of a thin isotropic layer of uniform density between the liquid and gas phases, the difference Δsurf between the ellipsometric angle Δ for the investigated solution and that of pure water Δ0 is proportional to the adsorbed amount Γ.42 Therefore, we present below only data on the ellipsometric angle Δ to characterize the adsorption kinetics. Note that in a simple case of a single solute one can calculate the adsorbed amount Γ from ellipsometric results using the expression of de Feijter et al.43 13436
DOI: 10.1021/acs.langmuir.6b03396 Langmuir 2016, 32, 13435−13445
Article
Langmuir Γ=
d(nx − n0) dn/dc
(1)
where n is the refractive index of the solution with concentration c and n0 and nx are the refractive indexes of the substrate and of the adsorbed layer, respectively. Infrared Reflection Absorption Spectroscopy. IRRAS spectra were recorded using a Nicolet 8700 FTIR spectrometer (Thermo Scientific, USA) equipped with a Tabletop optical module (TOM). The IR beam in the TOM was focused on the surface of the sample at an angle of incidence equal to 40° and was polarized perpendicular to the plane of incidence using a wire grid polarizer on a silicon substrate (ThorLabs, USA). FTIR spectra were collected with resolution of 8 cm−1 using a liquid-nitrogen-cooled MCT-D* detector and averaged over 512 scans. Dynamic Light Scattering. The size distributions of DNA coils and DNA/CTAB aggregates in bulk aqueous solutions were determined by dynamic light scattering (DLS) using a laser analyzer of particle size SZ 100 (Horiba Jobin Ivon, Japan). The measurements were carried out at 25 ± 0.1 °C using a 10 mm quartz cuvette at a scattering angle of 173°. The refractive index, 1.333, and viscosity, 0.897 cP, of water were used as the dispersant properties. Autocorrelation functions were collected for 1 min for each sample, and the reproducibility was checked using 10 measurements. The hydrodynamic diameter was calculated according to the Stokes−Einstein relation assuming a spherical shape of the particles. The particle size distributions were characterized by the scattering light intensity with a standard form of the polydisperse distribution.
Figure 1. Surface tension as a function of surfactant concentration: pure CTAB in water (black squares), CTAB in a mixed solution of 10 mM Tris-HCl and 20 mM NaCl (red diamonds), and CTAB/ DNA(50 μM) in a solution of 10 mM Tris-HCl and 20 mM NaCl solution (blue circles). The lines are guides for the eye.
The dynamic surface elasticity of pure CTAB solutions close to equilibrium does not exceed the error limits at concentrations of less than 2 × 10−6 M and reaches the maximal value of about 20 mN/m (Figure 2). At the same time,
■
RESULTS AND DISCUSSION DNA does not display any surface activity in dilute solutions and reduces the surface tension only slightly at high concentrations (∼10 g/L).44,45 The addition of oppositely charged surfactant to DNA solutions decreases the surface tension at low surfactant concentrations (
