pubs.acs.org/Langmuir © 2010 American Chemical Society
Use of Total Reflection X-ray Fluorescence (TRXF) for the Quantification of DNA Binding to Lipid Monolayers at the Air-Water Interface Vladimir L. Shapovalov,*,† Matthias Dittrich,‡ Oleg V. Konovalov,§ and Gerald Brezesinski‡ †
N. N. Semenov Institute of Chemical Physics RAS, Kosygina 4, 119991 Moscow, Russia, ‡Max Planck Institute of Colloids and Interfaces, Science Park Potsdam-Golm, Am Muehlenberg 1, 14476 Potsdam, Germany, and § European Synchrotron Radiation Facility, BP-220, E-38043 Grenoble, France Received June 17, 2010. Revised Manuscript Received July 30, 2010
To use the TRXF technique for the quantification of DNA binding to monolayers at the air-water interface, DNA from salmon testes was labeled by covalently bound bromine. For this purpose, an analytical procedure for the quantification of bromine in labeled DNA with a detection limit of 10-20 μg was developed. It was found that the pH of the solution has a strong influence on the yield of brominated DNA (BrDNA) when Br2 is used as a reagent. Much higher degrees of bromination can be achieved at pH 5 than at pH 7. A degree of bromination above a threshold of 2 to 3% (bromine per base) leads to the cross linking of BrDNA with the formation of an insoluble gel during the precipitation procedure. Finally, a reaction scheme with N-bromosuccinimide (NBS) that avoids precipitation has been established. Succinimide and some bromide ions remain in the solution as byproducts. However, these bromide ions are not competitive with BrDNA for binding at positively charged monolayers. Therefore, a new method for binding studies of model DNA to Langmuir monolayers at the air-water interface has been established. An important result of these studies is the finding that higher salt concentrations (representing physiological conditions) lead to an increased amount of adsorbed DNA. This can be explained by the decrease in the effective charge of the DNA molecules with decreasing Debye screening length.
Introduction DNA together with cationic lipids condenses and builds lipoplexes, which are efficient agents for the transfection of eukaryotic cells.1-6 Divalent cations mediate the complex formation with zwitterionic lipids that have also been used as nonviral gene delivery vectors of late.7 To improve the transfection of these systems constantly, it is important to understand the processes of binding and condensation in detail8 and correlate them to transfection efficiencies. Langmuir monolayers provide an excellent tool for elementary binding studies of DNA to lipids at the air-water interface.9-11 Structural information can be obtained by the use of several methods as grazing incidence X-ray diffraction (GIXD), X-ray reflectivity (XR),12 or infrared reflection absorption spectroscopy (IRRAS).13 However, the possibility to quantify the amount of *Corresponding author. E-mail:
[email protected].
(1) Felgner, P. L.; Ringold, G. M. Nature 1989, 337, 387. (2) Felgner, P. L.; Gadek, T. R.; Holm, M.; Roman, R.; Chan, H. W.; Wenz, M.; Northrop, J. P.; Ringold, G. M.; Danielsen, M. Proc. Natl. Acad. Sci. U.S.A. 1987, 84, 7413. (3) Ahmad, A.; Evans, H. M.; Ewert, K.; George, C. X.; Samuel, C. E.; Safinya, C. R. J. Gene Med. 2005, 7, 739. (4) Scarzello, M.; Smisterova, J.; Wagenaar, A.; Stuart, M. C. A.; Hoekstra, D.; Engberts, J. B. F. N.; Hulst, R. J. Am. Chem. Soc. 2005, 127, 10420. (5) Zhdanov, R. I.; Podobed, O. V.; Vlassov, V. V. Bioelectrochemistry 2002, 58. (6) Wasungu, L.; Hoekstra, D. J. Controlled Release 2006, 116, 255. (7) Bruni, P.; Pisani, M.; Amici, A.; Marchini, C.; Montani, M.; Francescangeli, O. Appl. Phys. Lett. 2006, 88, 1. (8) Bloomfield, V. A. Curr. Opin. Struct. Biol. 1996, 6, 334. (9) Gromelski, S.; Brezesinski, G. Langmuir 2006, 22, 6293. (10) Antipina, M. N.; Schulze, I.; Dobner, B.; Langner, A.; Brezesinski, G. Langmuir 2007, 23. (11) Antipina, M. N.; Schulze, I.; Heinze, M.; Dobner, B.; Langner, A.; Brezesinski, G. ChemPhysChem 2009, 10, 2471. (12) Als-Nielsen, J.; Jacquemain, D.; Kjaer, K.; Leveiller, F.; Lahav, M.; Leiserowitz, L. Phys. Rep. 1994, 246, 251. (13) Mendelsohn, R.; Brauner, J. W.; Gericke, A. Annu. Rev. Phys. Chem. 1995, 46, 305.
14766 DOI: 10.1021/la102472u
DNA bound to a lipid monolayer is quite restricted. In the case of most cationic lipids, IRRAS can be applied to observe at least the absorptions of the phosphate backbone of DNA strands.14 However, the signal strength is low (unpublished data), and in ternary systems of DNA, zwitterionic phospholipids, and divalent cations, the bands of the DNA backbone interfere with those of the phosphate groups of the lipid molecules.9 In contrast to that, total reflection X-ray fluorescence (TRXF) is a powerful method of quantifying a variety of systems at the air-water interface.15 In particular, the number of ions binding to charged monolayers can be determined down to very low values.16,17 However, the application of TRXF to DNA systems turns out to be difficult. Despite the complexity of the DNA molecule, it contains only a few different types of light elements that the lipid monolayers also contain. Therefore, labeling with a foreign element that possesses a high absorption cross section and strong K emission lines is inevitable. Fulfilling the mentioned requirements, bromine is a reasonable marker. It can be inserted by an electrophilic addition into the 5,6-double bond of pyrimidine bases and pyrimidine nucleotides.18,19 Purine bases are not as readily brominated, especially adenine and adenosine.20 In addition to the single DNA components, the bromination of intact macromolecular DNA and RNA structures has also been successfully performed. Calf thymus DNA was brominated with Br2, and the tobacco mosaic virus-RNA was brominated with (14) Banyay, M.; Sarkar, M.; Gr€aslund, A. Biophys. Chem. 2003, 104, 477. (15) Wobrauschek, P. X-Ray Spectrom. 2007, 36, 289. (16) Antipina, M. N.; Dobner, B.; Konovalov, O. V.; Shapovalov, V. L.; Brezesinski, G. J. Phys. Chem. B 2007, 111, 1384. (17) Shapovalov, V. L.; Ryskin, M. E.; Konovalov, O. V.; Hermelink, A.; Brezesinski, G. J. Phys. Chem. B 2007, 111. (18) Wang, S. Y. Nature 1957, 180, 91. (19) Ross, S. A.; Burrows, C. J. Tetrahedron Lett. 1997, 38, 2805. (20) Jones, A. S.; Woodhouse, D. L. Nature 1959, 183, 1603.
Published on Web 08/27/2010
Langmuir 2010, 26(18), 14766–14773
Shapovalov et al.
Article
N-bromosuccinimide (NBS).21-23 It was also reported that the integrity of DNA remains largely intact even at high degrees of bromination of up to one bromine atom per four DNA bases.24 To compare model systems of DNA binding to Langmuir monolayers with systems applied in biological experiments, the intactness of the investigated biomolecule is of great importance. In the present work, different approaches to DNA bromination are investigated and the applicability of modified BrDNA for binding studies with the use of TRXF is shown.
Chemicals For all measurements and sample preparations, Milli-Q Millipore water with a specific resistance of 18.2 MΩ 3 cm was used. DNA from salmon sperm (product number D-1626), bromine (>99.5%), and N-bromosuccinimide (NBS) (>99%) were purchased from Sigma-Aldrich. Dioctadecyldimethylammonium bromide (DODAB) and arachidyl sulfate (AS) were purchased from Fluka. All other chemicals were of analytical grade and used without further purification. NaCl and KCl for DNA solutions were heated to 600 C to prevent potential organic impurities. DNA solutions were freshly prepared and experiments were performed at 20 C to minimize the risk of denaturation.
Monolayer Experiments For monolayer experiments, 1 mM stock solutions of surfactants were prepared in chloroform (for DODAB) or in chloroform/methanol 3:1 (w/w) (for AS) and spread on the subphase using a 100 μL microsyringe. After a waiting time of 10 min for the evaporation of the solvent, the pressure/area isotherms were recorded on a computer-interfaced Langmuir trough (R&K, Potsdam, Germany) that is equipped with a Wilhelmy balance. The compression speed was 5 A˚2 3 molecule-1 3 min-1. The temperature was kept constant at 20 C to an accuracy of 0.1 C.
Total Reflection X-ray Fluorescence (TRXF) TRXF measurements were carried out at the ID10B (Troika II) high-brilliance undulator beamline of the European Synchrotron Radiation Facility (ESRF). The setup includes a Langmuir trough with dimensions of 170 438 mm2 (short side in the beam direction) equipped with a single movable barrier. The surface pressure π was measured with a NIMA PS4 tensiometer and was kept constant during all measurements. All experiments were performed in air at a constant room temperature of 20 C. The monochromatic synchrotron X-ray beam was fixed to a photon energy of 22 keV. A downstream mirror was bent to focus the X-ray beam onto the sample (6 m from the mirror). The deflected beam touched the liquid surface at a grazing angle of 0.040 that is ca. 70% of the critical angle of total reflection for the water surface. A scintillation (NaI) detector for the reflected X-ray beam was used for the height adjustment of the liquid surface. The fluorescence signal was measured by a thermoelectrically cooled VortexEM silicon drift X-ray detector with an entrance window placed parallel to the liquid surface at a fixed distance of 40 mm. After the spreading and compressing of the monolayer, the vertical position of the trough, with the Vortex-EM detector attached to the lid of the trough container, was adjusted to bring the center of the footprint of the incident beam to the middle of the trough and at (21) Tung, F.; Tsai, K. H.; Marfey, P. Biochim. Biophys. Acta 1972, 277, 117. (22) Recondo, A.-M. D.; Londos-Gagliardi, D.; Aubel-Sardon, G. FEBS Lett. 1971, 14, 149. (23) Brammer, K. W. Biochim. Biophys. Acta 1963, 8267, 217. (24) Lindahl, T.; Nyberg, B. Biochemistry 1972, 11, 3610.
Langmuir 2010, 26(18), 14766–14773
the same time to the middle of the view angle of the fluorescence detector.
Labeling of DNA: General Procedures For the atomic labeling of DNA, bromine was chosen. It fulfils the mentioned TRXF requirements and was expected to be easily linked covalently to DNA. DNA solutions (usually 10 mM when referring to a “monomer” consisting of one nitrogen base, a sugar, and a phosphate moiety, FW ca. 360) were prepared in water under constant stirring at room temperature. After the completion of the bromination reaction, CH3COONa (1 M aqueous solution or solid salt) was added to the reaction mixture to a final concentration of 300 mM. Then BrDNA was precipitated by the addition of 1.2 volumes of 96% ethanol. After intensive shaking, the fibrous precipitate “medusa” was filtered by the use of a stainless steel grid and washed twice with 70% ethanol, followed by filtration. The remaining product was dried by lyophilization or exposed to air at room temperature.
Quantification of Bromine Bound Covalently to DNA To determine the amount of DNA bound to model monolayers in a truly quantitative manner, the independent determination of the Br content in the labeled DNA is inevitable. Unfortunately, modern analytical methods such as NMR, IR, UV-vis spectroscopy, and HPLC were useless for our purposes. On one hand, it was necessary to determine the total content of Br in brominated DNA irrespective of what nitrogen bases are brominated and what positions are occupied by Br. On the other hand, Br atoms or C-Br bonds produce no highly specific and easily observable (analytical) signals. NMR is hardly applicable to the study of DNA because of line broadening due to the hindered rotation of the huge DNA molecules. The total decomposition of DNA with subsequent chemical modification of nitrogen bases followed by GC/MS25 seems excessively complicated and time-consuming to obtain the single necessary numerical value (Br content). Besides, multiple minor products (that cannot be discovered in such a way) can contain a significant fraction of the total amount of bromine. To determine the bromine content of BrDNA, the 150-year-old method of total mineralization of Carius26 (which is still in use27-29) was combined with a slightly more modern titration of bromide with AgNO3 using an adsorption indicator. Dry samples of BrDNA (ca. 10 mg) were heated to 300 C in a sealed glass ampule (i.d. = 8 mm, o.d. = 11 mm, volume = ca. 12 mL) containing 0.1 mL of fuming HNO3 (nearly 100%) and 2 to 3 mg of AgNO3. After at least 3 h of heating, the sample was given 1 h to cool. High-pressure gases (CO2 and nitrogen oxides) were liberated by heating the wall of the ampule near the sealed edge with the sharp flame of a propane torch until the glass softened and melted. After this, the ampule could be cut safely in the usual way. After the addition of ca. 1 mL of water to the reaction mixture, a few submillimeter agglomerates of AgBr can be seen (depending on the Br content). The precipitate (both agglomerates and the sediment on the walls) was washed three to four times with water in a special way: the liquid was sucked off with a thin glass tube with a cellulose membrane filter (pore diameter 0.2 μm) attached to the open end. This procedure guarantees no loss of the AgBr precipitate. Then, bromide was reliberated by the reduction of (25) Razskazovskii, Y.; Swarts, S. G.; Falcone, J. M.; Taylor, C.; Sevilla, M. D. J. Chem. Phys. B 1997, 101, 1460. (26) Carius, L. Ann. Chem. 1860, 116, 1. (27) Emich, F.; Donau, J. Monatsh. Chem./Chem. Mon. 1909, 30, 745. (28) Makineni, S.; McCorkindale, W. M.; Syme, A. C. J. Appl. Chem. 1958, 8, 310. (29) Howard, M. E.; Vocke, R. D. J. Anal. Atom. Spectrom. 2004, 19, 1423.
DOI: 10.1021/la102472u
14767
Article
AgBr to Ag and NaBr with 0.1 mL of a mixture of 0.2 M NaBH4 and 0.1 M NaOH and then neutralized with 0.1 mL of 1 M acetic acid (causing the decomposition of excess NaBH4 as well). During this procedure, the suction device was immersed in the solution in order to treat the residual AgBr on the filter as well. For titration, the acetate-buffered bromide solution was transferred to a clean vial by the use of the same suction device. At this stage, black reduced Ag can be easily seen on the filter. The resultant solution (approximately 1 mL) was titrated with a 5 mM AgNO3 solution (under permanent stirring with a magnetic stirrer). A minimum amount of cationic fluorescent dye Rhodamine 6G (0.01 mL of a 0.1 mM solution) was used as an adsorption indicator for the disclosure of the equivalence point. Under blue light excitation (blue LED with a violet cutoff filter), the emission of the dye in solution is bright yellow. In the presence of negatively charged colloidal AgBr (while there is excess Br-), the cationic dye adsorbs onto the particles forming aggregates and its emission is significantly quenched. As a result, at the beginning of the titration the yellow emission vanishes and reappears only after the equivalence point is reached. Overtitration leads to positively charged colloidal AgBr (excess Agþ), inducing the detachment of the dye into solution again. The emission of Rhodamine 6G was monitored visually by the use of an orange cutoff filter crossed with a violet one (see above). Because of increasing light scattering of the AgBr colloid during the course of the titration, the use of crossed filters is essential. The method described above allows the detection of a minimum amount of bromine of 10-20 μg per sample (e.g., 0.5-1% of DNA bromination) giving an accuracy of about 20% (the increase in the Br content results in better accuracy). It can be supposedly improved by the replacement of the titration procedure for a direct determination of bromide by means of ionic chromatography.
Bromination with Br2 The first attempt at bromination was performed using a methanolic solution of bromine as the brominating agent. For bromination with Br2, it is necessary to defend the DNA molecules from an acidic pH24 as being a result of the reaction. Therefore, a buffer is needed. A methanolic solution of bromine was added dropwise to a phosphate-buffered (50 mM phosphate concentration, pH 7), ice-cooled DNA solution under constant stirring. The calculated degree of bromination was 10%. After precipitation and washing, a 0.1 mM solution of the product was used as a subphase for a DODAB monolayer in a TRXF test experiment. The bromine signal was 10-100 times smaller than reasonable preliminary estimates suggested and therefore much too low for the quantitative analysis of the binding of DNA. In the following attempt, the mild bromination procedure of Razskazovskii et al.25 was chosen. Namely, DNA was brominated by Br2 transfer via the gas phase. The whole setup (Figure 1) was made of glass except for the titanium wire holding the inner vessel. (Titanium is perfectly resistant to the simultaneous action of Br2, HBr, and high humidity.) After preliminary ice cooling of the outer vessel containing the DNA solution, a calculated amount of 0.1 M aqueous bromine was added to the inner vessel. Afterwards, the outer vessel was immediately closed tightly. The DNA solution was stirred under ice cooling until the color of the bromine completely disappeared plus one more hour. Using the above-described quantification procedure for bromine, it was found that variations of the pH (near neutral) lead to very different degrees of bromination. The theoretically calculated amount of bromine was 10% in all reactions. At pH 7 14768 DOI: 10.1021/la102472u
Shapovalov et al.
Figure 1. Setup used for the bromination. Br2 is transferred through the gas phase out of compartment A into compartment B containing the DNA solution.
(50 mM phosphate buffer), the actual degree of bromination is found to be not more than 1% whereas bromination at pH 5 (30 mM acetate buffer) leads to a 6 to 7% degree of bromination. Without buffer, the degree of bromination is 2 to 3%. Test experiments, in which Br2 is added to the buffer solution without DNA, were also performed. If a phosphate buffer (pH 7) was used, then the result was a quick discoloration of the formerly orange-brown solution. The change in color is likely based on the well-known disproportionation of Br2: Br2 þ 3H2 O h Br - þ BrO - þ 2H3 Oþ The use of a buffer shifts the equilibrium to the right by neutralizing the hydronium ions (bromine dissolved in water is stable). Thus at pH 7, a high degree of disproportionation of Br2 will occur. In contrast to that, Br2 in an acetate buffer (pH 5) remains stable and keeps its original color. Therefore, the test experiments can explain the different degrees of bromination obtained with different buffers qualitatively. Preliminary estimates suggested that DNA with a degree of bromination of 5-10% is enough to get a sufficiently intense TRXF signal for quantitative studies of DNA binding. This degree of bromination can be achieved at pH 5 by the procedure described above. From here on, it was necessary to optimize the whole synthesis including the final purification and drying process in order to get a product that can be easily stored and rapidly dissolved. Surprisingly, it was found that dry BrDNA (6% bromination) is not truly soluble (in contrast to the original DNA) irrespective of the drying procedure (lyophilization or drying in the open air). Several hours of intense stirring, even combined with mild heating and preliminary overnight incubation in water, produces a solution that is mechanically and optically inhomogeneous. The samples consist of differently sized pieces of a gel-like material dispersed in the liquid, which has a rather low viscosity. The gel of BrDNA sustains heating up to 80 C and even treatment with acids and bases. It was found that freshly precipitated BrDNA (not dried at all) cannot be fully redissolved. On the contrary, original DNA retains its solubility during all steps of treatment (except for the addition of bromine). It should be noted that (i) original DNA fully dissolves within 1 to 2 h of mild stirring giving a perfectly homogeneous, highly viscous solution and (ii) DNA solutions do not change their appearance during the bromination stage. Furthermore, a degree of bromination of 1 to Langmuir 2010, 26(18), 14766–14773
Shapovalov et al.
2% does not affect the solubility behavior of DNA (after precipitation and drying) noticeably. All of these observations clearly suggest that the bromination of DNA results in chemical cross linking with the formation of a 3D network during precipitation and that this effect has a threshold. Chemical binding of few DNA molecules in solution and further formation of a “physical gel” in the precipitation stage was excluded by a special test with a 2 mM DNA solution (instead of 10 mM). A lower DNA concentration should decrease the probability of linkage in solution, but no effect was observed. (The linkage in the precipitate is independent of the DNA concentration in solution.) At this stage, the least unfavorable assumption is that the “active center(s)” responsible for chemical cross linking is (are) a highly reactive minor product(s) of the bromination reaction. Cross linking takes place within the precipitate because there the DNA molecules are much closer to each other than in solution. In such a case, at least three options to avoid the cross linking and retain a high degree of bromination are feasible. The first option is to shorten the initial DNA chains to reduce the probability to get an active center within a single DNA molecule. The initial DNA solution was treated with intense ultrasound until a substantial viscosity decrease in the solution was obtained. Surprisingly, the solubility of the BrDNA precipitate was not improved but rather got worse. There has been no explanation of this result until now. The second option was to change the brominating reagent/ procedure in order to affect the product distribution and possibly to avoid the formation of highly reactive products. (The latter is a matter of luck.) N-Bromosuccinimide (NBS) was reported23 to be a convenient brominating reagent for DNA. Bromination with NBS was performed by adding a freshly prepared aqueous solution of NBS (5 mM) dropwise to a buffered 10 mM DNA solution at room temperature under permanent stirring. The mixture was stirred 30 min, followed by the precipitation of BrDNA. Phosphate-buffered (50 mM, pH 7) and acetate-buffered (30 mM, pH 5) solutions with a calculated degree of bromination of 5-20% were used. Only one sample (pH 7, calculated degree of bromination 5%) appeared to be completely soluble. The degree of bromination of these samples was not determined because the experiments were performed with small amounts of DNA only. In another attempt, strong bromination agent CF3COOBr used for the bromination of inactive aromatic molecules30 but never reported as a bromination reagent for DNA was tried at pH 5. The precipitated BrDNA (calculated degrees of bromination of 10 and 20%) appeared distinctly yellow and insoluble. Additionally, the bromination of solid DNA with gaseous bromine was tried (exchange of reaction solvent from water to none) but led to an insoluble product once again. Thus, the second option also gave no satisfactory results. The third option was to add a chemical (“false target”) to the reaction mixture that can be readily attacked by the cross-linking agent at the stage of precipitation. This idea has long been used, and we propose only one more application. Indeed, if there are many false targets in the surroundings, then DNA will be very unlikely attacked by the cross-linking agent. At this stage, it was important to get an idea of the nature of the cross-linking agent and what target it attacks. A plausible assumption was that the cross-linking agent contains labile Br atoms and attacks an N atom of a nitrogen base in a neighboring DNA molecule, hence forming an intermolecular C-N bond. Accordingly, watersoluble nonvolatile amines triethanolamine (TEA) and monoethanolamine (MEA) were used as alternative targets for the (30) Barnett, J. R.; Andrews, L. J.; Keefer, R. M. J. Am. Chem. Soc. 1972, 94, 6129.
Langmuir 2010, 26(18), 14766–14773
Article
cross-linking agent. The addition of TEA after bromination with Br2 (calculated degree of bromination 20%) at pH 5 but before precipitation leads to BrDNA that can easily be dissolved. However, the concentration of TEA has to be surprisingly high;equal to or higher than that of DNA (10 mM). Stronger base MEA was expected to be a more efficient false target, and in fact 2 mM MEA is already enough to keep BrDNA soluble (calculated degree of bromination 20%). Unfortunately, the analysis revealed that the degree of bromination in both soluble products is only ca. 2%. An additional experiment with increased pH at the stage of precipitation (to keep amines deprotonated) and the use of antioxidative agents (NaHSO3 and hydroquinone) was unsuccessful and will not be described here. Because the introduction of false targets (TEA and MEA) not only prevents cross linking but also reduces the degree of bromination drastically, it is likely that the cross-linking agent is a main product of the bromination reaction. Apparently, the only remaining possibility for obtaining soluble BrDNA with the desired high degree of bromination is to avoid the precipitation procedure that is shown to be crucial to cross linking. Bromination with bromine produces significant quantities of bromide ions and requires the use of a buffer. These ions can be, e.g., the result of the electrophilic addition of Br2 to a thymine residue of a DNA molecule, followed by the substitution of OH- for Br-.31
Therefore, an isolation/purification step is needed in any case. Because significant amounts of solution have to be prepared in a short time for TRXF experiments, dialysis seems to be poorly applicable in our particular case. Fortunately, the reaction with NBS can be performed in water without the addition of a buffer because acid is not formed during the basic reaction: C4 H4 NO2 Br þ base H f C4 H4 NO2 H þ base Br After the reaction, the solution will contain brominated DNA and highly water soluble succinimide. However, an oxidative reaction can lead to small amounts of inorganic bromide: C4 H4 NO2 Br þ base H þ OH- f C4 H4 NO2 H þ base OH þ BrAdditionally, NBS can react with a thymine residue in the same manner as Br2, leading to Br- ions as well.31 The analysis of BrDNA obtained by the bromination with NBS without the use of a buffer (the calculated degree of bromination is 10%) reveals a degree of bromination of 6%. For the analysis, BrDNA was precipitated and air dried in the usual way. Without precipitation, the solution stays perfectly homogeneous and clear. It can be diluted with water without showing any signs of alteration. It should be again noted that BrDNA is insoluble after the precipitation step. Because the bromination yield is 60%, it has to be taken into account that the highest bromide ion content of the solution will be 40% of the amount of NBS used. In fact, this amount of total bromine was found in the filtrate after precipitation. The analysis (31) Dalton, D. R.; Dutta, V. P.; Jones, D. C. J. Am. Chem. Soc. 1968, 90, 5498.
DOI: 10.1021/la102472u
14769
Article
Shapovalov et al.
of the bromide content was performed by drying the filtrate, followed by heating until CH3COONa was melted. Afterwards, the product was dissolved and filtered from the carbonlike precipitate. Then bromide was precipitated as AgBr and centrifuged. The final steps in the analysis were the same as for BrDNA (see above). Higher degrees of bromination with NBS are presumably possible but have not been tried because the properties and the secondary structure of DNA might change noticeably.
Binding of BrDNA to the Model Cationic Monolayer Determination of Surface Concentration. The total reflection X-ray fluorescence (TRXF) technique allows the selective determination of various chemical elements within a thin surface layer near the air-water (or air-monolayer) interface, ignoring the same elements located in the bulk.17 Whereas the angle of incidence of the X-ray beam is below the critical angle of total external reflection, only a thin surface layer is affected by an evanescent wave. It decays exponentially with increase of a distance z from the interface: Iex ¼ I0 expð- z=dÞ
ð1Þ
d is the so-called penetration depth, which is practically independent of the X-ray wavelength. It depends on the total electron density of the bulk and increases dramatically in close proximity to the critical angle. For the particular case of water as a bulk material and an angle of incidence that is 30-70% of the critical one, d is in the range of 5-7 nm. The experimentally observed intensity of a fluorescence line j of an element i with a concentration profile ci(z) (along the direction normal to the interface) can be written as Z ð2Þ Iif, j ¼ bi, j Iex ðzÞ ci ðzÞ dz where Iex(z) is given by eq 1 and bi,j is a constant. This constant depends on multiple parameters of the experimental setup, the X-ray absorbance of the particular element, and the fluorescence yield for a particular line but not on the structure and composition of the surface layer (monolayer and adjacent subphase). The negligible absorbance of the X-ray fluorescence within the thin surface layer is ignored in eq 2. The absorbance in air on the way to the detector (essential for soft emission lines) can be included in bi,j. If the element of interest is a counterion in the electrical double layer (EDL) of a charged monolayer, then it is mostly located within a very thin layer (80% and ∼6% for 50% and 1% bromide in the bulk, respectively. Because inorganic bromide can be detected in the reaction mixture of BrDNA after bromination with NBS, the following step was used to check whether the ion is able to bind to a DODAB monolayer in the presence of DNA. In the classical EDL model (Gouy and Chapman), the concentration of ions is described by the Boltzmann equation ci ¼ ci, ¥ expð - zi eφ=kTÞ
ð5Þ
where ci and ci,¥ are local concentrations of ion i in the region with the electric potential j and in the bulk, respectively; zi is the (32) Cavalli, A.; Dynarowicz-Latka, P.; Oliveira, O. N.; Feitosa, E. Chem. Phys. Lett. 2001, 338, 88. (33) Ionov, R.; El-Abed, A.; Goldmann, M. Eur. Biophys. J. 2009, 38, 229. (34) Ahuja, R. C.; Caruso, P.-L.; M€obius, D. Thin Solid Films 1994, 242, 195.
Langmuir 2010, 26(18), 14766–14773
Shapovalov et al.
Figure 2. Isotherms of DODAB on subphases containing different concentrations of NaCl and NaBr at 20 C. The inset shows the phase-transition pressures vs the bromide fraction. (The line is a guide for the eye.)
Figure 3. X-ray fluorescence spectra of DODAB at 30 mN 3 m-1
on different subphases at 20 C on natural (bottom left) and expanded vertical scales (bottom right). The total experimental spectrum for a 1 mM NaBr subphase with the indication of the basic lines is shown at the top.
charge of that ion, and e is the elementary charge. This equation is valid for point charges and is obviously not (strictly) applicable to huge DNA molecules bearing multiple charges (>3000 in our case). Nevertheless, rough estimates are possible. Taking into account that highly charged DNA should decrease the potential of the EDL to almost zero, the local concentration of bromide should be as small as in the bulk. In contrast to that, the amount of DNA in the EDL should be high enough to compensate for the charge of the monolayer headgroups, which suggests the close packing of DNA molecules. A rough estimate of the concentration (referenced to a monomer) of closely packed DNA gives 2 to 3 M (i.e., some orders of magnitude higher than that of bromide as the byproduct of the DNA bromination reaction). Langmuir 2010, 26(18), 14766–14773
Article
The experimental result (Figure 3) is not as impressive as that estimate but fully satisfies our requirements as follows. The Br fluorescence signals of two systems were compared. The first consisted of a DODAB monolayer on top of a subphase containing 1 mM NaCl, 0.1 mM NaBr (substitute for the bromide byproduct), and 0.1 mM DNA (substitute for BrDNA). A Br signal corresponding to 0.006 Br atom per DODAB molecule was obtained. In the second system, the same monolayer was kept at the same surface pressure of 30 mN 3 m-1 (and therefore the same packing density) on a subphase containing 1 mM NaCl and 0.1 mM BrDNA (6% degree of bromination). This system reveals 0.10 Br atom per DODAB molecule, which is ca. 15 times more than in the first one. Taking into account that 0.1 mM BrDNA contains a maximum of 0.04 mM bromide (see above), the contribution of bromide as a reaction byproduct to the total Br signal (mostly coming from the DNA Br labels) is less than 3%. The concentration of inorganic electrolytes (mostly NaCl) in biological liquids is about 0.15 M. Therefore, systems intended for gene transfection should sustain high salt concentrations. Taking this into account, TRXF was used to investigate DNA binding to a DODAB monolayer within the range of 1-100 mM salt concentration. In this set of experiments, the NaCl electrolyte was advisedly exchanged with KCl because the absorbance of Na is too weak at 22 keV and the emitted X-ray lines (20 mN 3 m-1), the isotherms reveal nonmonotonic behavior. The molecular area is largest for the subphase containing 10 mM KCl and smallest for the subphase containing 100 mM KCl. Additionally, a plateau surprisingly (re)appears in the isotherm at 100 mM KCl. Further experiments are needed to explain these effects, but this is not the aim of this study. TRXF spectra for the above systems (taken at a constant surface pressure of 30 mN 3 m-1) are shown in Figure 5. Relatively strong Br lines that progressively increase with increasing KCl concentration are observed. The K-line intensities are increased dramatically in the same manner. The latter is partially a result of the increasing signal from the bulk. (See the thin line for a 100 mM KCl subphase without DODAB.) However, the K signal from the monolayer is also essential. To quantify the surface concentration of K, one more calibration sample was used: an anionic monolayer of AS (strong acid) on top of a 1 mM KCl subphase was kept in the condensed state at 30 mN 3 m-1. (The molecular area is 25 A˚2 for this pressure according to ref 35.) The results of the TRXF data treatment are presented in Table 1. Surface concentrations of elements c0i are recalculated to the dimensionless element per headgroup ratio r = c0i A for convenience. (A molecular area of 70 A˚2 was used for all samples.) The amount of DNA is estimated from the number of Br labels in accordance with a degree of bromination of 6%. It is worth noting that signals originating from the bulk are subtracted during the data treatment procedure. (35) Shapovalov, V. L.; Brezesinski, G. J. Phys. Chem. B 2006, 110, 10032.
DOI: 10.1021/la102472u
14771
Article
Shapovalov et al.
Figure 6. Schematic illustration of the binding of BrDNA strands to a positively charged DODAB monolayer. At low ionic strength, the negative charges of the BrDNA molecules are screened by only a few counterions (right). At high ionic strength, the effective charge of BrDNA is more strongly reduced and a larger amount of BrDNA is needed to compensate for the overall charge of the monolayer (left).
Figure 4. Isotherms of DODAB on subphases containing 0.1 mM BrDNA and different concentrations of KCl at 20 C. The isotherm on 10 mM KCl is given as a reference for orientation.
Figure 5. X-ray fluorescence spectra of DODAB at 30 mN 3 m-1 on a subphase containing 0.1 mM BrDNA and different concentrations of KCl at 20 C. The thin line refers to a 100 mM KCl subphase without DODAB. Table 1. Numbers of Ions, Br Labels, and DNA Monomers per DODAB Headgroup for Different KCl Concentrations [KCl], mM
Cl-
Kþ
Br
DNA
1 3 10 100
