1606
Langmuir 2009, 25, 1606-1611
Mechanism of DNA Strand Exchange at Liposome Surfaces Investigated Using Mismatched DNA Karolin Frykholm,† Bengt Norde´n,† and Fredrik Westerlund*,†,‡ Department of Chemical and Biological Engineering, Chemistry and Biochemistry, Chalmers UniVersity of Technology, SE-412 96 Gothenburg, Sweden, and Nano-Science Center and Department of Chemistry, UniVersity of Copenhagen, DK-2100 Copenhagen E, Denmark ReceiVed October 3, 2008. ReVised Manuscript ReceiVed NoVember 20, 2008 DNA strand exchange is of great importance in vivo for genetic recombination and DNA repair. The detailed mechanism of strand exchange is not understood in full detail despite extensive studies. Simplistic model systems in which molecular parameters can be varied independently are therefore of interest to study. We chose the surface of a positively charged liposome as a scaffold, which we recently demonstrated to be able to catalyze the exchange of fully complementary DNA oligonucleotides. We here study how single base pair mismatches affect the rate of strand exchange on the liposome surface. Interestingly, the rate of the exchange does not simply follow the stability of the duplex in solution, as determined by melting temperatures, but also depends sensitively on the position of the mismatch. For duplexes with similar melting temperatures, the exchange is much faster for a mismatch close to the end than for a mismatch in the middle of the sequence. Our results suggest that the single strands are stabilized by the liposome surface; therefore, the duplex is fraying more and the DNA opens up in a zipperlike fashion on the surface, increasing the probability of strand exchange. We also show that the competition between greater stability (higher Tm in solution) and higher concentration is important for the final composition of the duplex when a large excess of single strands is added to a complementary double-stranded DNA. Finally, the similar exchange rate constants for fully base-paired duplexes on the liposome surface when adding fully matched single strands or single strands with a mismatched base indicate that the rate is governed largely by separation of the initial duplex and not by the formation of the product duplex.
Introduction DNA strand exchange is an important process for DNA repair and for keeping the genetic diversity. The RecA protein has a central role in the reaction in bacterial cells in ViVo1,2 and the eukaryotic homologue is Rad51.3,4 The mechanism for the strand exchange promoted by these proteins has been extensively studied, but is not understood in full detail. Therefore studies of simple model systems are of interest to investigate the importance of the individual molecular parameters involved one at a time. A well studied in Vitro system for DNA strand exchange is the comb-type copolymers developed and investigated by the group of Maruyama.5-7 The polymer consists of a positively charged poly(L-lysine) backbone to electrostatically accumulate the DNA and dextrane graft chains claimed to stabilize a triplex intermediate. Lipid bilayers are among the most important self-assembled structures in nature, and lipid/DNA interactions are of importance in several research areas. One example is nonviral gene delivery where DNA bound to positively charged lipid vesicles, so-called lipoplexes, is extensively used, and studies regarding their stability * Corresponding author. E-mail:
[email protected]. Tel: +45-35320483. Fax: +45-35320406. † Chalmers University of Technology. ‡ University of Copenhagen.
(1) Takahashi, M.; Norden, B. AdV. Biophys. 1994, 30, 1. (2) Cox, M. M. Prog. Nucleic Acid Res. Mol. Biol. 1999, 63, 311. (3) Bianco, P. R.; Tracy, R. B.; Kowalczykowski, S. C. Front. Biosci. 1998, 3, 570. (4) West, S. C. Nat. ReV. Mol. Cell Biol. 2003, 4, 435. (5) Kim, W. J.; Akaike, T.; Maruyama, A. J. Am. Chem. Soc. 2002, 124, 12676. (6) Kim, W. J.; Sato, Y.; Akaike, T.; Maruyama, A. Nat. Mater. 2003, 2, 815. (7) Choi, S. W.; Kano, A.; Maruyama, A. Nucleic Acids Res. 2008, 36, 342.
and structure are of great interest.8-11 Interactions between DNA and lipids are also important in vivo because the interior of the cell nucleus consists, to a significant extent, of lipids (5-14%)12,13 that have been suggested to interact with DNA, histones, and chromatin.14,15 Furthermore, Corsi et al. demonstrated that DNA can be transcribed in the liquid-crystalline phase of phospholipids.16 Lipid surface scaffolds are considered for numerous nanotechnology applications, and the development of a system that can “self-correct” mismatches would be an important breakthrough in the buildup of, for example, the addressable hexagonal DNA networks developed by our group.17-19 The liposome surface is attractive for studying DNA strand exchange because it allows for a large amount of variety in composition. Hence, the individual molecular parameters governing the rate of exchange can be investigated independently by introducing lipids with head groups that can have specific (8) 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. (9) Chesnoy, S.; Huang, L. Annu. ReV. Biophys. Biomol. Struct. 2000, 29, 27. (10) Safinya, C. R. Curr. Opin. Struct. Biol. 2001, 11, 440. (11) Rubanyi, G. M. Mol. Aspects Med. 2001, 22, 113. (12) Hunt, A. N.; Clark, G. T.; Attard, G. S.; Postle, A. D. J. Biol. Chem. 2001, 276, 8492. (13) Irvine, R. F. Nat. ReV. Mol. Cell Biol. 2003, 4, 349. (14) Jones, D. R.; Divecha, N. Curr. Opin. Genet. DeV. 2004, 14, 196. (15) Kuvichkin, V. V. Bioelectrochemistry 2002, 58, 3. (16) Corsi, J.; Dymond, M. K.; Ces, O.; Muck, J.; Zink, D.; Attard, G. S. Chem. Commun. 2008, 2307. (17) Tumpane, J.; Sandin, P.; Kumar, R.; Powers, V. E. C.; Lundberg, E. P.; Gale, N.; Baglioni, P.; Lehn, J.-M.; Albinsson, B.; Lincoln, P.; Wilhelmsson, L. M.; Brown, T.; Norden, B. Chem. Phys. Lett. 2007, 440, 125. (18) Tumpane, J.; Kumar, R.; Lundberg, E. P.; Sandin, P.; Gale, N.; Nandhakumar, I. S.; Albinsson, B.; Lincoln, P.; Wilhelmsson, L. M.; Brown, T.; Norden, B. Nano Lett. 2007, 7, 3832. (19) Banchelli, M.; Betti, F.; Berti, D.; Caminati, G.; Baldelli Bombelli, F.; Brown, T.; Wilhelmsson, L. M.; Norden, B.; Baglioni, P. J. Phys. Chem. B 2008, 112, 10942.
10.1021/la8032513 CCC: $40.75 2009 American Chemical Society Published on Web 01/05/2009
Mechanism of DNA Strand Exchange
Langmuir, Vol. 25, No. 3, 2009 1607
Table 1. Sequence of the Template Strand (t), the Fully Complementary Strand (c), and the Single Base Mismatched Strands (mX), with the Mismatch at Position Xa code
sequence
t c m3 m4 m5 m10
3′-GCA GTT GTA TGT ATA GTG GT-5′ 5′-CGT CAA CAT ACA TAT CAC CA-3′ 5′-CGG CAA CAT ACA TAT CAC CA-3′ 5′-CGT AAA CAT ACA TAT CAC CA-3′ 5′-CGT CCA CAT ACA TAT CAC CA-3′ 5′-CGT CAA CAT CCA TAT CAC CA-3′
t* c* m3*
3′-GCA TGT GTA TGT ATA GTG GT-5′ 5′-CGT ACA CAT ACA TAT CAC CA-3′ 5′-CGG ACA CAT ACA TAT CAC CA-3′
a Included are the extra strands for melting studies (t*, c*, and m3*), with the switched bases in italics.
interactions with DNA such as hydrophobic interactions or hydrogen bonds. We have recently reported that DNA strand exchange is significantly enhanced on the surface of positively charged liposomes compared to that in bulk solution.20 We found that 35% cationic lipid content is optimal for fast, efficient exchange. We also suggested that the rate-determining step under the optimal conditions is the destabilization of the initial duplex. We here investigate the effect of single base mismatches on the rate of DNA strand exchange on the surface of positively charged liposomes. By inserting mismatches at different positions in a 20-mer oligonucleotide, we want to investigate how the position of the mismatch correlates with the exchange rate and if the duplex stability on the liposome surface is the same as in the bulk. We observe that the exchange rate is much faster for a mismatch close to the end of a 20-mer duplex than for a mismatch in the middle of the duplex. This suggests that the ends of the duplex are fraying significantly more on the liposome surface than in bulk solution and that the duplex opens up in a zipperlike fashion. We also investigate the delicate competition between stability and concentration that governs the composition of the final product when adding a large excess of single strands to DNA duplexes. Finally, we confirm our earlier reported observation that the rate is mainly governed by the dissociation of the initial duplex and not the formation of the product duplex.
Materials and Methods Materials. DNA oligonucleotides in Table 1, template strands (t and t*) with and without fluorescein 5′-modification, and complementary strands (c, c*) and strands with single base mismatches mX (m3, m4, m5, m10, and m3*, with X denoting the position of the mismatch starting from the end of the strand opposite the dye) with and without TAMRA 3′-modification were purchased from ATDbio. To form duplexes [t + c] and [t + mX], equimolar amounts of template and complementary strands were heated to 90 °C and annealed slowly over 6 h. 1,2-Dioleoyl-sn-glycero-3-phosphatidylcholine (DOPC) was purchased from Larodan and N-[1-(2,3-dioleoyloxy)propyl]-N,N,Ntrimethylammonium chloride (DOTAP) from Sigma-Aldrich. The molecular structures of the lipids are shown in Figure 1. Preparation of Lipid Vesicles. To prepare large unilamellar vesicles (LUVs), lipids dissolved in chloroform were mixed in a 35:65 DOTAP/DOPC molar ratio, and solvent was then evaporated under reduced pressure using a rotary evaporator. To remove any remaining traces of solvent, the lipid film was placed under vacuum for at least 2 h or overnight. The film was dissolved in aqueous buffer (50 mM sodium phosphate, pH 7.5, containing 1 mM EDTA) by vortex mixing and subjected to five to six cycles of freeze-thaw (liquid nitrogen/45 °C heat block), with the final thawing at room (20) Frykholm, K.; Baldelli Bombelli, F.; Norden, B.; Westerlund, F. Soft Matter 2008, 4, 2500.
Figure 1. Molecular structures of the lipids used. (Top) Zwitterionic DOPC. (Bottom) Cationic DOTAP.
temperature. The solution was then extruded 21 times through polycarbonate filters with a pore size of 100 nm using a hand-held syringe extruder. The size of the liposomes was approximately 110-140 nm, as determined by dynamic light scattering (DLS) using a Malvern Instruments Zetasizer Nano ZS at a wavelength of 633 nm. FRET Experiments. The fluorescence resonance energy transfer (FRET) experiments were performed on a Cary Eclipse fluorescence spectrophotometer (Varian). The samples were excited at 495 nm, and the emission was collected at 516 nm as a function of time. All strand-exchange experiments were performed at 25 °C. In the FRET experiment, doubly labeled double-stranded DNA was mixed with lipid vesicles at a concentration of 50 mM lipids. The measurement was started, and after approximately 60 s, the measurement was paused, unlabeled single-stranded DNA was added in 5-fold excess compared to double-stranded DNA, and the sample was thoroughly mixed before the measurement was restarted, typically after approximately 200 s. DNA was added to the liposomes at a lipid/DNA charge ratio of 1:0.5, calculated as the ratio between positive charges from the total amount of lipids in the sample and negative charges from the total amount of double- and single-stranded DNA. When the TAMRA-labeled strand of the duplex is exchanged with a nonlabeled strand, the FRET pair is separated, and the fluorescein emission is increased. The kinetic traces were satisfactorily fitted to a monoexponential expression. All experiments were repeated at least three times, and the inverse rate constants were averaged, giving the inverse rate constants and standard deviations reported in the Tables. The strand-exchange yield of each sample was related to the yield in a reference sample, representing the maximum possible amount of strand exchange. A yield of 1 indicates that the sample at equilibrium has the same composition as a temperature-annealed sample where the final product has the thermodynamically favored composition. (See ref 20 for details.) DNA Melting Temperatures (Tm). DNA melting curves were measured by UV spectroscopy on a Cary 4000 UV-vis spectrophotometer (Varian) equipped with a programmable multicell temperature block. Samples of duplex DNA, degassed under reduced pressure before measurement, were heated from 10 to 85 °C at a rate of 0.5 °C min-1 and cooled again to 10 °C at the same rate. The samples were kept at 10 °C for 5 min, and then the temperature ramping was repeated at a rate of 0.25 °C min-1. The absorption at 260 nm was measured with a temperature interval of 1 °C. The maximum of the first derivative of the curves was used to estimate the melting temperature, Tm, of the duplex. Each measurement resulted in four curves from which the melting temperature could be determined. Similar Tm values were obtained from the heating and cooling curves and from the different rates of temperature ramping. Each measurement was repeated twice, and the average of the eight data points is reported in the Tables.
Results To test the effect of single base mismatches on the rate of DNA strand exchange on liposome surfaces and to determine how the position of the mismatch correlates with the exchange rate, a series of oligonucleotides were designed (sequences in
1608 Langmuir, Vol. 25, No. 3, 2009
Frykholm et al.
Table 2. Inverse Rate Constants for Strand Exchange of a Doubly Labeled Fully Complementary Duplex [t + c] or a Singly Mismatched Duplex [t + mX] (with Mismatch in Position X) When Adding a 5-Fold Excess of the Unlabeled Fully Complementary Strand (+ c) or a Strand with an Identical Mismatch (+ mX)a
Table 3. Melting Temperatures, Tm, for the Fully Complementary Duplex [t + c] and the Single Base Mismatched Duplexes [t + mX] with the Mismatch at Position Xa Tm/°C duplex
ssDNA
[t [t [t [t [t
τ/s dsDNA [t [t [t [t [t a
+ + + + +
c] m3] m4] m5] m10]
+c
+ mX
1030 ( 80 480 ( 90
