Langmuir 2006, 22, 2935-2937
2935
DNA-Mediated Delivery of Lipophilic Molecules via Hybridization to DNA-Based Vesicular Aggregates Paul M. Dentinger,* Blake A. Simmons, Evelyn Cruz, and Matthew Sprague Sandia National Laboratories, P.O. Box 969, LiVermore, California 94550 ReceiVed NoVember 8, 2005. In Final Form: February 7, 2006 A scheme is presented for stabilizing hydrophobic molecules and releasing them into aqueous solution via DNA hybridization. A tetradecyl hydrophobic tail is covalently attached to synthetic oligomers, and the resulting amphiphilic molecules take up substantial amounts of orange OT and pyrene dyes in aqueous environments. The resulting structures do not affect the surface tension and are predominantly spherical as shown by light scattering and TEM, and the pyrene fluorescence is consistent with a hydrophobic environment. It is concluded that the amphiphilic DNA creates vesicular domains upon which the hydrophobic dyes reside and are stabilized in solution. Upon exposure to the complementary strand, the pyrene dye is released from the structures, showing that the scheme can be used for unlabeled or DNAmediated drug delivery.
There is considerable interest in DNA-mediated delivery of substances and in unlabeled DNA detection. For the delivery of substances, vesicle-forming molecules have been used for some time.1-4 Vesicles are stable species that can deliver lipophilic drugs to targeted sites if the vesicles have some form of recognition1,4-8 or can encapsulate hydrophilic moieties and deliver them upon disruption.2,6,8-10 In addition to traditional lipids, various bilayer-forming molecules such as amphiphilic poly-L-lysine,9 polymeric chitosan,7,10 alkyl glycerols,11 and derivatized peptide nucleic acids12 have been shown to form vesicles, and targeted delivery has been shown with these nonphospholipid schemes.7 DNA itself has been associated noncovalently with surfactants and lipids in bilayer structures.4,8,13 However, we are unaware of any reports of DNA-based structures that respond and release their encapsulated contents upon DNADNA recognition, a tool that could be useful in the targeting of liposome/vesicle-based formulations or in DNA-hybridizationspecific detection schemes. In addition, sequence-specific hybridization is the basis of many assays for the unlabeled detection of DNA. Typically, the strategy employs a novel DNA-based construct or material that, upon hybridization, is disrupted, resulting in a signal. Frutos et al. showed increased fluorescence based on the hybridization* Corresponding author. E-mail:
[email protected]. (1) Presant, C. A.; Proffitt, R. T.; Teplitz, R. L.; Williams, L. E.; Tin, G. W. Method of Delivering Micellar Particles Encapsulating Chemotherapeutic Agents to Tumors in a Body. U.S. Patent 5,441,754 A, August 15, 1995. (2) Perrie, Y.; Barralet, J. E.; McNiel, S.; Vangala, A. Int. J. Pharm. 2004, 284, 31-41. (3) Bouwstra, J. A. Drugs Pharm. Sci. 2000, 105, 271-302. (4) Mady, M. M.; Ghannam, M. M.; Khalil, W. A.; Repp, R.; Markus, M.; Rascher, W.; Muller, M.; Fahr, A. J. Drug Targeting 2004, 12, 11-18. (5) Takuya, S.; Ohami, Y.; Fukuda, A.; Hayashi, M.; Kawakubo, A.; Kato, K. Cytotechnology 2001, 36, 93-99. (6) Zhou, W.; Yuan, X.; Wilson, A.; Yang, L.; Mokotoff, M.; Pitt, B.; Li, S. Bioconjugate Chem. 2002, 13, 1220-1225. (7) Dufes, C.; Muller, J.-M.; Couet, W.; Olivier, J.-C.; Uchegbu, I. F.; Schaetzlein, A. G. Pharm. Res. 2004, 21, 101-107. (8) Pantos, A.; Tsiourvas, D.; Nounesis, G.; Paleos, C. M. Langmuir 2005, 21, 7483-7490. (9) Wang, W.; Tetly, L.; Uchegbu, I. F. J. Colloid Interface Sci. 2001, 237, 200-207. (10) Uchegbu, I. F.; Schatzlein, A. G.; Tetly, L.; Gray, A. I.; Sludden, J.; Siddique, S.; Erasto, M. J. Pharm. Pharmacol. 1998, 50, 453-458. (11) Gopinath, D.; Ravi, D.; Rao, B. R.; Apte, S. S.; Rambhau, D. Int. J. Pharm. 2002, 246, 187-197. (12) Marques, B. F.; Schneider, J. W. Langmuir 2005, 21, 2488-2494. (13) Dias, R. S.; Lindman, B.; Miguel, M. G. J. Phys. Chem. B 2002, 106, 12600-12607.
Scheme 1
induced disruption of bimolecular beacons,14 whereas Gaylord et al. used PNA/conjugated polymer constructs15 and Elganian used Au-nanoparticle/DNA constructs.16 In general, it is the clever assembly of a material that is disrupted upon hybridization that can lead to the sensitive detection of DNA. In this letter, we describe a DNA-based vesicular structure or amphiphilic aggregate that readily takes up lipophilic molecules (in this case, dyes) into aqueous solutions with high efficiency. Complementary strands of DNA disrupt the aggregates and release the dyes into solution whereas a nonbinding strand does not. We show that the dyes are stable in the aggregates until addressing the complementary strands. In addition, we show that the number of DNA molecules needed to release approximately 40% of the lipophilic contents is approximately 10 times less than the amount of DNA utilized to encapsulate the dyes, showing that the general scheme can be used, in some sense, as an amplification technique. The scheme could potentially be used in the unlabeled detection of DNA or potentially in the DNA-mediated delivery of drugs via stable vesicles disrupted by complementary strands. Scheme 1 shows the basics of the concept. A 16-mer oligomer has a 5′C10 carboxylate linker attached during the synthesis scheme. An alkylamine is reacted to form a hydrophobic tail on the oligomer. The resulting amphiphilic oligomers form ag(14) Frutos, A. G.; Pal, S.; Quesada, M.; Lahiri, J. J. Am. Chem. Soc. 2002, 124, 2396-2397. (15) Gaylord, B. S.; Heeger, A. J.; Bazan, G. C. Proc. Natl. Acad. Sci. U.S.A. 2002, 99, 10954-10957. (16) Elghanian, R.; Storhoff, J. J.; Mucic, R. C.; Letsinger, R. L.; Mirkin, C. A. Science 1997, 277, 1078-1079.
10.1021/la053005o CCC: $33.50 © 2006 American Chemical Society Published on Web 03/07/2006
2936 Langmuir, Vol. 22, No. 7, 2006
gregates in phosphate-buffered saline (PBS) solutions. We refer to the resulting structures as vesicular aggregates. These vesicular aggregates take up substantial portions of lipophilic dyes such as orange OT17 and pyrene (Aldrich, Milwaukee, WI).18 Upon exposure to the complementary oligomer, the aggregates are disrupted, and the lipophilic dye is released into solution. All oligomers were purchased from Trilink Biotechnologies (San Diego, CA). The tetradecylamine-modified oligomer predominantly used was analyzed with PAGE and RP-HPLC and confirmed with ESI mass spectrometry at Trilink and reported as 5351 amu. The aggregates were typically prepared in PBS 7.4 and 10-100 mM concentration of buffer and shaken vigorously by hand and vortexed prior to surface tension, light scattering, TEM, and dye uptake. However, when the oligomers were utilized in high concentrations for surface tension measurements, 150 mM PBS buffer was necessary such that the oligomer did not overwhelm the buffer and turn the solution acidic. Surface tension analysis was done with a Sensa-Dyne bubble formation pressure surface tensiometer (Chem-Dyne Research Corp.) with a bubble frequency of 0.5 Hz using a dry nitrogen source gas. Light scattering was performed on a Wyatt Technology (Santa Barbara, CA) Dawn EOS multiangle static light scattering instrument operated in batch mode at 25 °C. Freeze-fracture TEM was performed at Nano Analytical Laboratory (San Francisco, CA) by using a sandwich sampling technique and liquid-nitrogencooled propane to achieve a cooling rate of 10 000 K/s. Fracturing of the frozen samples was conducted in a JEOL JED-9000 freeze etcher. The exposed fracture planes were shadowed with Pt, followed by carbon deposition, and then cleaned with fuming HNO3 for 24 h, followed by rapid agitation and rinsing with a 1:1 (v/v) chloroform/methanol solution. TEM was performed using a JEOL 100 CX TEM.19 Aggregates were prepared with orange OT and pyrene at a concentration higher than the concentration limit (2 mg/mL), vortexed, and left to stand at least 24 h to obtain equilibrium concentrations. Samples were then centrifuged (2500 rpm × 10 min) to separate the excess dye, and the supernatant was removed by transfer pipet for UVvis analysis. It was initially thought that the amphiphilic oligomers may be surface-active and form micelles above set concentrations in solution. If the oligomers formed micelles, then they would force the use of millimolar concentrations of DNA, which are above typical concentrations utilized in most DNA detection or drugdelivery techniques, and would be of exceptional cost. To determine if the oligomers tended to form micelles, two different oligomers were preparedsone with the C14 alkane tail attached to the decyl carboxylate linker and the other with the 2-ethyl hexyl aliphatic tail commonly found on surfactants. At the highest concentrations utilized (8.8 mM for the C14 tail), neither of the hydrophobic-tagged oligomers appeared to affect the surface tension of the solution as determined by dynamic surface tensiometry. Similar concentrations are needed for many micelle systems, but we did not observe any effect of the oligomers on surface tension. Very preliminary experiments of the orange OT dye uptake in water and PBS appeared to favor the oligomer with the C14 tail, whereas the oligomer with the 2-ethylhexyl tail showed little activity in dye uptake, so the C14-tail oligonucleotide (C14ON) was chosen for the remainder of the experiments. Repeated measurements of the surface tension of (17) Schott, H. J. Phys. Chem. 1966, 70, 2966-2973. (18) Damas, C.; Naejus, R.; Coudert, R.; Frochot, C.; Brembilla, A.; Viriot, M.-L. J. Phys. Chem. B 1998, 102, 10917-10924. (19) Kan, P. L.; Papahadjopolous-Sternberg, B.; Wong, D.; Waigh, R. D.; Watson, D. G.; Gray, A. I.; McCarthy, D.; McAllister, M.; Schatzlein, A. G.; Uchegbu, I. F. J. Phys. Chem. B 2004, 108, 8129-8135.
Letters
Figure 1. FF-TEM image taken of a 3.4 mM solution of C14ON. Table 1. Freeze-Fracture TEM and Static Light Scattering Results on DNA Vesicular Aggregates radiusav (nm) std dev (nm)
FF-TEM
SLS
100 N/A
120 20
samples with concentrations up to 8.8 mM were performed on C14ON with no appreciable drop in surface tension, indicating that the amphiphilic oligomers were not forming micelles. Although the C14ON was not surface-active, it was thought that aggregates could exist in solution. To determine whether aggregates may exist in solution resulting from the amphiphilic DNA molecules, freeze-fracture transmission electron microscopy (FF-TEM) was employed. Figure 1 shows a typical FF-TEM of a 3.4 mM solution of the C14ON. Several morphological features existed in the FF-TEM, including oblong shapes with an aspect ratio of approximately 2:1, large-diameter species (greater than 800 nm), and smaller, 20-50 nm spheres. However, many 100 nm spherical objects populated the Figures. Given that no attempt was made to prepare exceptional vesicles at this point using techniques such as drying and rehydration or extrusion, it is possible that a more careful exploration of the concentration, solvent, and preparation space would produce more monodisperse samples. Table 1 lists the results of static light scattering experiments (SLS), and these are compared to freeze-fracture TEM (FFTEM). The static light scattering results indicate the presence of relatively monodisperse structures with an average radius of 120 nm. It is clear from light scattering and FF-TEM that aggregates exist in solution, and it also appears that these are spherical structures that are much larger than typical micelles (2-10 nm) and are in the size range of vesicle structures (50-500 nm). It is clear from the surface tension, light scattering, and TEM measurements that some form of aggregate is present in solution and from the spherical or ovoidal shape combined with the lack of surface activity that the structures could be vesicular. If they were bilayer structures, then they would be expected to absorb and stabilize highly hydrophobic dyes in aqueous solution. Figure 2 shows the UV-vis spectra of two dyes used to determine lipophilic environments within aqueous systemssorange OT and pyrene. It is clear that orange OT is present and stabilized within the C14ON solution at much higher concentrations than its solubility limit in water. C14ON stabilizes orange OT to an absorbency of 2.3 at 17.5 g/L of DNA. This is approximately half of the extrapolated capability of poly(ethylene oxide) dodecyl ether nonionic surfactants on a mass/liter basis but is similar in capability to the reported value of sodium dodecyl sulfonate.17 Control experiments with 5′CTA CTGCTGGGAACTC3′ without the aliphatic tail showed that this diazo dye could be stabilized in solution, but at a concentration of approximately 10-fold less
Letters
Figure 2. UV-vis spectra of orange OT (red) and pyrene (blue) solubilized by a 3.5 mM solution of C14ON.
than in experiments with the hydrophobic tail, indicating that the dye was not solely electrostatically bound to the anionic DNA under these conditions. To further confirm the ability of C14ON to absorb lipophilic dyes and to further confirm the presence of a lipophilic phase within the C14ON/PBS system, pyrene was used. In this case, not only is the pyrene-based dye stabilized in aqueous solution but also fluorescence peaks at I1 ) 373.2 and I3 ) 384.5 with a subsequent I1/I3 ratio of 1.16 (λexcit ) 320 nm) are present. The I1/I3 ratio falls between the values reported for perfluorohexanediol (1.36) and hexanediol (1.05), further confirming that this dye is experiencing a hydrophobic environment.18 It is clear from the data above that the C14ON utilized here forms a vesicular aggregate in solution and that this vesicular aggregate stabilizes lipophilic dyes. If the aggregates could release the dyes upon exposure to target molecules, then it would be expected that this concept could be used for either an assay or a drug-delivery system. Figure 3 shows the result of the addition of the target sequence on the pyrene/aggregate solution. Upon addition of the complementary sequence, the aggregates quickly release the pyrene dye. With only 1/10 the concentration of target (free oligomer) to probe (aggregated oligomer), the pyrene is reduced by approximately 40%. Figure 3 shows only about a 0.5 absorbance unit decrease for pyrene at these concentrations. Given the extinction coefficient for pyrene of approximately 35 000 M-1cm-1, the target oligomer at 0.35 mM displaced approximately 14 µM of the pyrene dye; therefore, roughly 24 DNA molecules are needed to displace a pyrene molecule. A nonbinding strand (identical oligomer sequence without the hydrophobic tail) was ineffective at producing an absorbance change, indicating that hybridization was at least necessary to
Langmuir, Vol. 22, No. 7, 2006 2937
Figure 3. Plot of UV-vis data taken from a 3.5 mM aqueous solution of the 1-pyrene oligomer before and after the addition of 0.35 mM target oligomer as a function of time.
effect the delivery of the hydrophobic molecule. One aspect of either drug delivery or DNA-based sensing applications would be the stability of the dyes within the aggregates. The solubilized dyes were measured 10 weeks after preparation with little observed change. Simply measuring the absorbance of the stabilized dye as a function of aggregate change due to hybridization is a complicated phenomena that involves both the ability of the aggregate to interact with the strand and also the solubility characteristics of the dye as the aggregates are rearranged or disrupted. It is expected that additional work on hydrophobic molecules relevant to either drug delivery or specific unlabeled detection schemes would result in a clarification of the mechanism and will be the subject of future work. We have shown that hydrophobic tails placed onto oligonucleotides form aggregates in solution and that these aggregates are not micelles but do substantially stabilize lipophilic dyes. In addition, it is shown that the vesicular aggregates can be disrupted upon hybridization to complementary strands, yielding potential uses in DNA-targeted drug delivery and the potential for unlabeled DNA detection schemes. Acknowledgment. Sandia is a multiprogram laboratory operated by Sandia Corporation, a Lockheed Martin Company, for the United States Department of Energy under contract no. DEACO4-94AL85000. LA053005O
