Biomacromolecules 2002, 3, 841-845
841
Evaluation by Fluorescence Resonance Energy Transfer of the Stability of Nonviral Gene Delivery Vectors under Physiological Conditions Keiji Itaka,†,‡ Atsushi Harada,† Kozo Nakamura,‡ Hiroshi Kawaguchi,‡ and Kazunori Kataoka*,† Department of Materials Science, Graduate School of Engineering, The University of Tokyo, 7-3-1 Hongo, Bunkyo-ku, Tokyo 113-8656, Japan, and Department of Orthopaedic Surgery, Faculty of Medicine, The University of Tokyo, 7-3-1 Hongo, Bunkyo-ku, Tokyo 113-8655, Japan Received February 25, 2002; Revised Manuscript Received April 2, 2002
The stability in physiological medium of polyplex- and lipoplex-type nonviral gene vectors was evaluated by detecting the conformational change of complexed plasmid DNA (pDNA) labeled simultaneously with fluorescein (energy donor) and X-rhodamine (energy acceptor) through fluorescence resonance energy transfer (FRET). Upon mixing with cationic components, such as LipofectAMINE, poly(L-lysine), and poly(ethylene glycol)-poly(L-lysine) block copolymer (PEG-PLys), the fluorescence spectrum of doubly labeled pDNA underwent a drastic change due to the occurrence of FRET between the donor-acceptor pair on pDNA taking a globular conformation (condensed state) through complexation. The measurement was carried out also in the presence of 20% serum, under which conditions FRET from condensed pDNA was clearly monitored without interference from coexisting components in the medium, allowing evaluation of the condensed state of pDNA in nonviral gene vectors under physiological conditions. Serum addition immediately induced a sharp decrease in FRET for the LipofectAMINE/pDNA (lipoplex) system, which was consistent with the sharp decrease in the transfection efficiency of the lipoplex system in serum-containing medium. In contrast, the PEG-PLys/pDNA polyplex (polyion complex micelle) system maintained appreciable transfection efficiency even in serum-containing medium, and FRET efficiency remained constant for up to 12 h, indicating the high stability of the polyion complex micelle under physiological conditions. Introduction The demand for nonviral gene delivery systems which are safe and highly efficient has increased with the recent progress of clinical gene therapy. Transfection efficiency, one of the most important factors for such delivery systems, is closely correlated with the stability of the systems under physiological conditions. In fact, in most cases of in Vitro transfection using polymer and lipid-based gene vectors, the transfection efficiency decreases in the presence of serum,1,2 suggesting that the physicochemical properties of the vector particles are greatly influenced through interaction with serum components. To gain insight into the properties of vectors under physiological conditions, however, a novel analytical approach is required, because conventional methods for the analysis of nanometer size particles, such as light scattering measurement3-5 and ethidium bromide assay,6-9 usually require measurement conditions that cannot be achieved in physiological media which contain proteins. Methods based on fluorescence spectroscopy can be carried out under physiological conditions, since they are * To whom correspondence may be addressed. Phone: +81-3-58417138. Fax: +81-3-5841-7139.. E-mail:
[email protected]. † Department of Materials Science, Graduate School of Engineering, The University of Tokyo. ‡ Department of Orthopaedic Surgery, Faculty of Medicine, The University of Tokyo.
less sensitive to coexisting components. Indeed, to estimate the intracellular localization and dynamics of DNA complexes with cationic polymer (polyplexes), fluorescently labeled DNA and/or cationic polymer is often used.10-12 This method, however, still has several problems. For example, there is a resolution limit in observing the close proximity of labeled DNA/labeled polymer when analyzing the fluorescent images. Moreover, fluorescent labeling of a cationic polymer is often tedious and may cause changes in the physicochemical property of polymers. To overcome these problems, we investigated a novel fluorescent approach aimed at detecting conformational change of plasmid DNA (pDNA) in nonviral vector systems and the results of this study are reported here. Inside the complexes, pDNA molecules take a condensed conformation via interaction with the cationic partner, resulting in an increase in the stability and availability of these vectors. An adequate donor-acceptor pair of fluorescent dyes was attached to a single pDNA molecule, and the conformational change of pDNA, which leads to a change in the distance between the two fluorescent molecules, was detected by fluorescence resonance energy transfer (FRET). The study was carried out with different types of gene vectors, a lipoplex and a polyplex, to evaluate their stability. It was demonstrated that the polyplexes formed from pDNA and poly(ethylene glycol)-poly(L-lysine) block copolymer
10.1021/bm025527d CCC: $22.00 © 2002 American Chemical Society Published on Web 05/02/2002
842
Biomacromolecules, Vol. 3, No. 4, 2002
(polyion complex micelle) display remarkable stability in serum, while lipoplexes underwent rapid decomposition under identical conditions. Experimental Section Chemicals. Poly(L-lysine) with a degree of polymerization (DP) ) 80 was purchased from Sigma Chemical Co. Poly(ethylene glycol)-poly(L-lysine) block copolymer (PEG, MW ) 12 000; poly(L-lysine), DP ) 48) was prepared as described previously.13 Sodium dextran sulfate (MW ) 500 000) was purchased from Wako Pure Chemical Industries, Ltd. LipofectAMINE reagent and fetal bovine serum (FBS) without heat-inactivating treatment were purchased from GIBCO BRL. Hydroxychloroquine was purchased from Acros Organics. Plasmid DNA. pGL3-Luc (Promega) was used in all experiments. This plasmid was amplified in competent DH5R Escherichia coli and purified using EndoFree Plasmid Maxi or Mega Kits (QIAGEN). The DNA concentration was determined by UV absorbance measured at 260 nm. Fluorescent Labeling. Plasmid DNA (pDNA) was labeled using a Label IT Nucleic Acid Labeling Kit (Panvera, USA). This system promotes the covalent attachment of the specific fluorescent molecules to guanine residues in nucleic acids. Following a protocol provided by the manufacturer, slightly modified to allow double labeling of DNA, 5-50 µL of a pDNA solution (1 mg/mL) and the same amount of Label IT Reagent (for fluorescein or X-rhodamine) were mixed in 20 mM MOPS buffer (pH 7.5) and incubated at 37 °C for 2 h. For double labeling of fluorescein and X-rhodamine, the two reagents were added simultaneously to the pDNA solution. Unreacted labeling reagent was removed, and pDNA was purified by ethanol precipitation. The labeling efficiency to pDNA was evaluated by the method recommended by the manufacturer using the standard curve prepared by plotting concentration of a fluorescent molecule vs fluorescence intensity (fluorescein, excitation λ ) 492 nm and emission λ ) 520 nm; X-rhodamine, excitation λ ) 576 nm and emission λ ) 597 nm).14 The emission intensity of labeled pDNA was measured and the concentrations of fluorescein and X-rhodamine were then calculated using the standard curve. Then, the number of base pairs of pDNA per label was calculated from MW of fluorescent molecules (fluorescein, 1003 g/mol; X-rhodamine, 1030 g/mol) and the concentration of pDNA determined by UV absorbance at 260 nm. Formation of DNA-Loaded Complex Particles. Poly(ethylene glycol)-poly(L-lysine) block copolymer and pDNA were separately dissolved in 10 mM Tris-HCl buffer (pH 7.4). Both solutions were mixed at various charge ratios of the number of lysine units per nucleotide. The final DNA concentration of the mixture was adjusted to 33.3 µg/mL. The poly(L-lysine)/DNA complex (Plys-polyplex) was prepared similarly by mixing poly(L-lysine) and pDNA solution. The polycationic lipid/DNA complex (lipoplex) was prepared by mixing the pDNA solution and LipofectAMINE reagent following by the protocol provided by the manufacturer. The LipofectAMINE reagent (15 µL) was added to 100 µL of a
Itaka et al.
pDNA solution (50 µL/mL) in 10 mM Tris-HCl buffer (pH 7.4) and, finally, the DNA concentration was also adjusted to 33.3 µg/mL. Fluorescence Measurements. Emissions of free pDNA and DNA-loaded complex particles were measured at 37 °C using a spectrofluorometer (JASCO, FP-777). The excitation wavelength was 492 nm. Serum Stability Assay. To particles (PIC micelles, Plyspolyplex, and lipoplex) loaded with doubly labeled pDNA, fetal bovine serum (FBS) without heat-inactivating treatment was added and the mixture was incubated at 37 °C. Changes in the emission spectra with time were measured for an appropriate time period (120 min for lipoplex and 3 days for PIC micelles and Plys-polyplex). Transfection. 293 cells were seeded in six-well culture plates. After 48 h of incubation in medium containing 10% FBS, the cells were rinsed and 1000 µL of culture medium without FBS was added to each well. One-hundred microliters of PIC micelles or lipoplex solution was applied to each well. The concentration of pDNA was adjusted to 3 µg of DNA per well. For PIC micelles, 100 µM of hydroxychloroquine was included for enhancement of gene expression.15 Six hours later, the medium was removed and replaced by 10% FBS containing medium, and after 24 h of incubation, luciferase gene expression was measured. For preincubation study with serum, 20% volume of FBS was added to the solutions of PIC micelles or of lipoplex and incubated at 37 °C for 3-30 min prior to transfection. Results and Discussion Labeling Efficiency. The efficiency was approximately 30 bp DNA/label (33.3 labels/1000 bp). As the distance between two bases in DNA molecules is 3.3 Å, the average distance between two attached fluorescent molecules is approximately 100 Å. The Fo¨rster distance (R0) between fluorescein (donor) and rhodamine (acceptor) is 55 Å.16 The efficiency of energy transfer between dyes is negligible when pDNA is in the relaxed conformation. Change in Emission Spectra upon Micellization and Release of pDNA from Micelles. We first investigated whether changes in FRET efficiency in PIC micelle system correspond to pDNA compaction upon charge neutralization. It is known that totally water-soluble and narrowly distributed nanoparticular complexes (PIC micelle) surrounded by hydrophilic PEG chains are formed by mixing charged block copolymers and pDNA at a charge ratio r g 1, where r is defined as the ratio of the number of lysine unit per nucleotide.13,17,18 Poly(ethylene glycol)-poly(L-lysine) block copolymer was added to labeled pDNA in Tris-HCl buffer (pH 7.4) in various charge ratios, and the fluorescence was measured. As shown in Figure 1A, the fluorescence intensity at 520 nm, due to fluorescein emission, decreased upon addition of copolymer, with concomitant increase in the intensity at 597 nm, due to X-rhodamine emission. Correspondingly, the excitation spectra observed at 597 nm increased in the region corresponding to fluorescein excitation (data not shown).
Stability of Nonviral Gene Delivery Vectors
Biomacromolecules, Vol. 3, No. 4, 2002 843
Figure 2. Changes in the emission intensity ratio at 597/520 nm by the release of pDNA from PIC micelles, plotted as a function of [sulfate of dex-sul.]/[phosphate in nucleotide]. (9) and (b) indicated PIC micelles of r ) 1 and r ) 2, respectively.
Figure 1. (A) Emission spectra of doubly labeled free pDNA, and PIC micelles formed with poly(ethylene glycol)-poly(L-lysine) block copolymer and the labeled pDNA. The ratio of the number of lysine unit per nucleotide (r ) [lysine residue]/[nucleotide]) of the PIC micelles was 0.5 and 1.0. pDNA was labeled by both fluorescein and X-rhodamine, and the ratio of attached molecules (fluorescein:Xrhodamine) was 1:5. The excitation wavelength was 492 nm. (B) Change in the emission intensity ratio at 597/520 nm (b) and fluorescence intensity of ethidium bromide (9) as a function of r for PIC micelles.
These results indicate that the fluorescence of X-rhodamine is due to FRET between fluorescein (donor) and Xrhodamine (acceptor), conjugated to the condensed pDNA inside the particles. As shown in Figure 1A, the change in emission spectra derived from FRET became gradually apparent with increasing charge ratio. This behavior is emphasized in a plot of the changes in the ratio of emission intensities at 597/520 nm (Figure 1B).19 The ratio increased gradually with increasing in r value for r e 1. It leveled off at charge neutralization (r ) 1). This observation correlated well with the result of the ethidium bromide assay also shown in Figure 1B. We have shown previously that pDNA is released quantitatively upon addition of an equi-unimolar ratio of polysulfate to PIC micelles, due to an exchange reaction of pDNA and polyanion.5 Figure 2 shows that the release of pDNA from PIC micelles can be observed by FRET. Upon addition of an equivalent amount of dextran sulfate (MW ) 500 000) to PIC micelles containing doubly labeled pDNA, the ratio
Figure 3. Comparison of the emission intensity ratio at 597/520 nm in doubly labeled free pDNA, PIC micelles, poly(L-lysine)/DNA complex (Plys-polyplex), and polycationic lipid/DNA complex (lipoplex).
of emission intensities at 597/520 nm decreased to that of free pDNA. These results confirm that the globule to coil transition of pDNA which is triggered by the dissociation of the PIC micelles is detectable by FRET. As the most apparent change in emission spectra was observed when the fluorescein to X-rhodamine ratio was 1:5, this labeling ratio was used throughout this study. FRET Measurements for Plys-polyplex and Lipoplex. In PIC micelles and Plys-polyplex (r ) 2), the emission intensity ratio of 597/520 nm was about 1.8 (Figure 3). Interestingly, the ratio was higher in lipoplex than in PIC micelles and Plys-polyplex. This difference in the emission ratio may reflect the difference in the condensation state of pDNA in the complexes. It is likely that in Plys-polyplex and PIC micelles, the interwinding poly(L-lysine) strands may interfere with the contact between donor and acceptor fluorescent dyes along pDNA strands. However in lipoplex which is composed of pDNA and low molecular weight lipids, it is expected that the pDNA strands have more freedom of rearrangement. Further detail on the structural parameters of complex particles affecting FRET efficiency is the subject of a separate study to be reported elsewhere. Serum Stability. One of the most important advantages of the FRET technique is that it can be used in various physiological environments. Even in the presence of 20%
844
Biomacromolecules, Vol. 3, No. 4, 2002
Itaka et al.
Figure 4. FRET evaluation of the serum stability. (A) Lipoplex containing double labeled pDNA was incubated at 37 °C with FBS without heat-inactivating treatment. The concentrations of FBS were 20% (9), 5% (2), and 1% (b), respectively. Data are presented as the emission intensity ratio at 597/520 nm for each emission spectrum. (B) Emission intensity ratio of PIC micelles of r ) 1 (9), r ) 2 (b), and Plys-polyplex r ) 1 (2), r ) 2 (1), free pDNA ([), and lipoplex (0), respectively. (O) corresponds to control PIC micelles without FBS.
serum, FRET from condensed pDNA was clearly detected without interference from coexisting components in the medium. In lipoplex, upon coincubation with FBS without heatinactivating treatment at 37 °C, the ratio of emission intensities at 597/520 nm decreased rapidly within a few minutes. Then, the slope of the regression line gradually leveled off under 1% serum condition (Figure 4A). This suggests that lipoplex had two phases of dissociation in 1% serum. When the concentration of FBS was increased to 5% and further, the first phase of dissociation became prominent resulting in a lower emission intensity ratio within a shorter time period than in the medium with 1% serum. Serum protein can rapidly associate with lipid systems,20,21 which may result in rapid decrease of the emission intensity ratio. Interestingly, the ratio did not reach the value measured for free pDNA but remained slightly high (see Figures 3 and 4A). Therefore, at the initial stage, nonspecific interactions between lipids and serum components may cause a drastic loosening of condensed pDNA molecules in the lipoplex. The steep reduction in the ratio at the initial stage, regardless of the concentration of serum in the medium, strongly suggests that lipoplex promptly becomes unstable in serum. On the other hand, the ratio of emission intensities at 597/ 520 nm remained relatively constant for hours in PIC
micelles (r ) 1 and 2), and Plys-polyplex (r ) 2) (Figure 4B). In r ) 1 Plys-polyplex, the ratio started to decrease immediately after coincubation, in contrast with the r ) 2 polyplex, which had an induction period of about 4 h. Most notable, though, is the fact that in PIC micelles the emission intensity ratio remained constant for a very long period (∼12 h), indicating the superior serum stability of PIC micelles compared to other lipoplex and polyplex systems. It is interesting to note also that the ratio of the number of lysine unit per nucleotide (r) in PIC micelles had a negligible effect on the time-dependent decrease in the emission intensity ratio. The high stability of complexed pDNA in PIC micelles had been reported previously on the basis of nuclease resistance experiments.5,17 This high nuclease resistance ability indicates the stable nature of the complex in which the migration of the constituent polymer chain (PEG-Plys and pDNA) is likely to be restricted. Furthermore, stabilization of the complexed pDNA through compartmentalization into the PEG microenvironment may contribute to the superior stability under physiological conditions. Transfection. To confirm the serum stability from the standpoint of transfection efficiency, in Vitro transfection was performed after preincubation of PIC micelles and lipoplex with 20% serum. Within a few minutes of preincubation with serum, the lipoplex showed a drastic decrease in transfection
Stability of Nonviral Gene Delivery Vectors
Biomacromolecules, Vol. 3, No. 4, 2002 845
it directly: FRET between two different fluorophores attached to a single pDNA molecule. Upon coil to globule transition of pDNA triggered by the formation and dissociation of pDNA complex with polymers or surfactants, the fluorescent spectrum undergoes significant changes due to the occurrence of FRET. Even in the serum-containing medium, PIC micelles maintain appreciable transfection efficiency and FRET is observed consistently, indicating a highly improved serum stability. This result suggests a promising feasibility of the PIC micelles as gene vectors used for clinical gene therapy. Acknowledgment. The authors thank Professor F. M. Winnik of University of Montreal for critical reading of the manuscript. This work was supported by Grants-in-Aid for Scientific Research (no. 11167210 to K.K and no. 12877221 to H.K.) and Special Coordination Funds for Promoting Science and Technology from the Ministry of Education, Culture, Sports, Science and Technology of Japan. References and Notes
Figure 5. Influence of serum preincubation time on the transfection efficiency. Data are presented separately for lipoplex (A) and PIC micelles (B).
efficiency (Figure 5A). In contrast, PIC micelles retained sufficient efficiency even after a serum preincubation time of 30 min (Figure 5B). This clearly showed that PIC micelles maintained their ability for gene transfection even in 20% serum, vouching for the highly improved serum stability of PIC micelles compared to lipoplexes. The improved stability of PIC micelles may be ascribed to their characteristic core-shell structure, consisting of a dense PEG corona and a core serving as stable pDNA microreservoir. From the standpoint of developing a nonviral vector for clinical gene delivery, the stability of the vector in physiological environments is one of the most important factors. Indeed, we recently confirmed that the complexed pDNA in PIC micelles was stable enough to remain in the blood circulation for a substantial period.22 Conclusions In this report, we have described the first application of a FRET method to evaluate the condensation state of pDNA in lipoplexes and polyplexes in the presence of serum. To evaluate the actual conformation of pDNA in the medium, we designed here a simple yet sensitive technique to observe
(1) Escriou, V.; Ciolina, C.; Lacroix, F.; Byk, G.; Scherman, D.; Wils, P. Biochim. Biophys. Acta 1998, 1368, 276-88. (2) De Smedt, S. C.; Demeester, J.; Hennink, W. E. Pharm. Res. 2000, 17, 113-26. (3) Kataoka, K.; Togawa, H.; Harada, A.; Yasugi, K.; Matsumoto, T.; Katayose, S. Macromolecules 1996, 29, 8556-8557. (4) Lim, Y. B.; Han, S. O.; Kong, H. U.; Lee, Y.; Park, J. S.; Jeong, B.; Kim, S. W. Pharm. Res. 2000, 17, 811-6. (5) Katayose, S.; Kataoka, K. J. Pharm. Sci. 1998, 87, 160-3. (6) Toncheva, V.; Wolfert, M. A.; Dash, P. R.; Oupicky, D.; Ulbrich, K.; Seymour, L. W.; Schacht, E. H. Biochim. Biophys. Acta 1998, 1380, 354-68. (7) Howard, K. A.; Dash, P. R.; Read, M. L.; Ward, K.; Tomkins, L. M.; Nazarova, O.; Ulbrich, K.; Seymour, L. W. Biochim. Biophys. Acta 2000, 1475, 245-55. (8) Ruponen, M.; Yla-Herttuala, S.; Urtti, A. Biochim. Biophys. Acta 1999, 1415, 331-41. (9) Bronich, T. K.; Nguyen, H. K.; Eisenberg, A.; Kabanov, A. V. J. Am. Chem. Soc. 2000, 122, 8339-8343. (10) Godbey, W. T.; Wu, K. K.; Mikos, A. G. Proc. Natl. Acad. Sci. U.S.A. 1999, 96, 5177-81. (11) Zelphati, O.; Szoka, F. C., Jr. Proc. Natl. Acad. Sci. U.S.A. 1996, 93, 11493-8. (12) Stegmann, T.; Legendre, J. Y. Biochim. Biophys. Acta 1997, 1325, 71-9. (13) Harada, A.; Kataoka, K. Macromolecules 1995, 28, 5294-5299. (14) FAQ: Label IT Nucleic Acid Labeling Kits; Panvera Corporation: 1999. (15) Itaka, K.; Kataoka K. Proc. Int. Symp. Control Rel. Bioact. Mater. 2001, 28, 1157-8. (16) Haugland, R. P.; Handbook of Fluorescent Probes and Research Chemicals, 8th ed.; Molecular Probes: Eugene, OR, 1999. (17) Katayose, S.; Kataoka, K. Bioconjugate Chem. 1997, 8, 702-7. (18) Kataoka, K.; Harada, A.; Nagasaki, Y. AdV. Drug DeliVery ReV. 2001, 47, 113-31. (19) Herman, B. Methods in Cell Biology; Academic Press: New York, 1989. (20) Chonn, A.; Semple, S.; C.; Cullis, P.; R. J. Biol. Chem. 1992, 267, 18759-65. (21) Li, S.; Tseng, W. C.; Stolz, D. B.; Wu, S. P.; Watkins, S. C.; Huang, L. Gene Ther. 1999, 6, 585-94. (22) Harada-Shiba, M.; Yamauchi, K.; Harada, A.; Shimokado, K.; Kataoka, K. Gene Ther., in press.
BM025527D
