Letter pubs.acs.org/macroletters
Direct Incorporation of Functional Peptides into M‑DNA through Ligand-to-Metal Charge Transfer Kwang Suk Lim,†,§ Daniel Y. Lee,†,‡,§ Gabriel M. Valencia,† David A. Bull,*,†,‡ and Young-Wook Won*,†,‡ †
Division of Cardio-Thoracic Surgery, Department of Surgery, University of Utah School of Medicine, Salt Lake City, Utah 84132, United States ‡ Department of Pharmaceutics and Pharmaceutical Chemistry, University of Utah, Salt Lake City, Utah 84112, United States S Supporting Information *
ABSTRACT: Conventional nonviral gene delivery methods suffer from the toxicity of the cationic nature of polymeric carriers. There is a significant need for a new method of gene delivery that overcomes the limitations and allows targeted gene delivery. In this study, we have developed a new method to incorporate functional peptides into DNA without the need for chemical conjugations by utilizing a ligand-to-metal charge transfer (LMCT) transition, which occurs between divalent metal ions and the sulfhydryl group in cysteine. To apply the LMCT transition to the incorporation of cysteine-containing targeting peptides into DNA, divalent metal ions must be first introduced to DNA. Zn2+ ions spontaneously intercalate into the DNA base pairs in the pH range of 7.0−8.5, resulting in the conversion of normal B-DNA to metal-bound DNA (M-DNA). We found that the Zn2+ ions present in M-DNA could interact with the sulfhydryl groups in cysteines of targeting peptides through the LMCT transition, and the M-DNA/peptide complex could specifically transfect the target cells.
S
self-assembly of DNA,13,16 this conformational change does not affect the activity of DNA. M-DNA was generated using Zn2+ ions, and the M-DNA was modified with a cancer-targeting peptide, C-RGD-C, or a cellpenetrating peptide (CPP), C-9R-C, through the LMCT transition. The Zn2+ ions present in the M-DNA interact with the sulfhydryl group of cysteine present in these peptides. The C-RGD-C peptide attached to the M-DNA enhanced gene transfection into the target cancer cells. Similarly, the LMCT transition and the electrostatic interaction simultaneously contributed to the enhanced binding of the CPPs to the MDNA, which in turn reduced the amount of the peptide necessary to achieve high levels of gene transfection. This method can serve as a means to modify DNA with any functional peptide. The M-DNA formation was verified through an ethidium fluorescence assay.14,17 The effects of the Zn2+ concentration and the time on the conversion of B-DNA to M-DNA were studied at pH 7.5 (Supporting Information Figure 1). M-DNA was partially generated by incubation with ≤3.66 mM Zn2+, while complete conversion to M-DNA was observed in the presence of ≥5.49 mM Zn2+. In the absence of Zn2+, no conformational change in the DNA was observed. It was
uccessful gene therapy is reliant upon the transfer of safe and efficient therapeutic nucleic acids to their target cells.1−3 Gene delivery vehicles can be divided into two categories: viral vectors and nonviral carriers.4,5 Each of these methods, however, suffers from the limitations of immunogenicity of the viral compartments or toxicity of the cationic nature of polymeric carriers.6−8 In particular, the development of nonviral polymer carriers utilizes one or more of the following principles: (1) electrostatic interaction; (2) encapsulation; and (3) absorption.3,4,9 For efficient gene transfection, these methods require the use of levels of cationic polymers, which can cause cytotoxicity.6,8 There is, therefore, a new method of nonviral gene delivery that reduces polymer use and allows targeted gene delivery without compromising transfection efficiency. We have focused on a ligand-to-metal charge transfer (LMCT) between Zn2+ ions and the sulfhydryl group in cysteine to directly bind functional peptides to nucleic acids.10,11 This LMCT transition between Zn2+ ions and the cysteinyl residues of peptides occurs at nanomolar and/or picomolar levels of affinity.12−15 To apply this LMCT transition to the modification of DNA with cysteine-containing peptides, Zn2+ ions must first be introduced to DNA. It is known that divalent metal ions, such as Zn2+, lead to the conversion of normal B-DNA to metal-bound DNA (M-DNA) through intercalation of the metal ions into the DNA base pairs in the pH range of 7.0−8.5.13,15 Although this binding can drive the © XXXX American Chemical Society
Received: November 14, 2016 Accepted: January 12, 2017
98
DOI: 10.1021/acsmacrolett.6b00865 ACS Macro Lett. 2017, 6, 98−102
Letter
ACS Macro Letters
decreased by the LMCT transition. This result further verifies the occurrence of the LMCT transition between the M-DNA and the C-9R-C peptide. We sought to determine the minimal amount of C-9R-C peptide necessary for the complete condensation of the MDNA. The migration of the M-DNA was completely retarded at a C-9R-C ratio of 0.08 (Figure 2a), whereas retardation of the B-DNA by the formation of the C-9R-C/B-DNA complex was observed at a C-9R-C ratio of 1.0 (Figure 2b). G-9R-G, which can bind to M-DNA solely through electrostatic interaction, retarded the migration of the M-DNA at a G-9R-G ratio of 1.0 (Figure 2c). This ratio is the same ratio at which the parent BDNA was condensed by G-9R-G (Figure 2d). The M-DNA was incubated with C-9R-C (Figure 2e) or G-9R-G (Figure 2f) at pH 5.0 or at pH 7.4 because the LMCT transition does not occur below pH 7.0,12 while the guanidine side groups of arginine retain their positive charge even below pH 7.0. While both peptides can interact with M-DNA through electrostatic interaction at pH below 7.0, C-9R-C binds to the M-DNA through both electrostatic interaction and the LMCT transition at pH 7.4. C-9R-C retarded migration of the M-DNA at a C9R-C ratio of 0.05 at pH 7.4. This retardation of the M-DNA migration was not observed at pH 5.0 (Figure 2e). There was no difference in migration, regardless of the pH, when the MDNA was condensed by G-9R-G (Figure 2f). The transfection efficiency (Figure 2h) and the cellular uptake (Supporting Information Figure 5) of the C-9R-C/MDNA complex were in direct proportion to the amount of C9R-C added and were consistently greater than that of the C9R-C/B-DNA complex without an increase in cytotoxicity (Figure 2i). Interestingly, the C-9R-C/M-DNA complex prepared at a C-9R-C ratio of 0.1 exhibited the same level of gene expression as the C-9R-C/B-DNA complex prepared at the C-9R-C ratio of 2.0. In addition, we compared the transfection efficiency of the C-9R-C/M-DNA complex to that of the G-9R-G/M-DNA complex prepared at pH 7.4 or at pH 5.0. The transfection efficiency of the C-9R-C/M-DNA complex relative to that of the G-9R-G/M-DNA complex increased continuously with an increase in the ratio of C-9R-C at pH 7.4, while no difference in the relative transfection was observed at pH 5.0 (Figure 2g). These results indicate that the C-9R-C peptide capable of interacting with the M-DNA through the LMCT transition and electrostatic interaction reduces the amount of peptide necessary for the high gene expression to less than 5% of the peptide amount required for polyplex formation with a cationic peptide. In order for the LMCT transition to be broadly applicable to gene delivery, there must be the capacity for targeted gene transfection. Conjugation of targeting peptides to nonviral vectors is one of the most established approaches for the development of targeted gene delivery.22,23 Since the direct conjugation of targeting moieties to nucleic acids can lead to the loss of nucleic acid activity,24,25 the direct introduction of targeting peptides to nucleic acids through conventional conjugation-based methodologies is not a viable approach. Moreover, negatively or slightly positively charged targeting peptides are unable to directly interact with nucleic acids through electrostatic interaction. We therefore hypothesized that the LMCT transition could allow the incorporation of any peptides containing cysteine to M-DNA without the need for either chemical conjugation or a positively charged mediator. To prove our hypothesis, a noncyclic C-RGD-C peptide, the
verified that the M-DNA generated in this study was nontoxic to cells (Supporting Information Figure 2). To compare our new method utilizing the LMCT transition to conventional electrostatic interaction, the M-DNA was first condensed by the most commonly used cationic polymer: branched polyethylenimine 25 kDa (PEI). While the formation of M-DNA prior to the addition of PEI can serve as a means to reduce the amount of polymer required for complete DNA condensation and efficient gene transfection (Supporting Information Figures 3 and 4), the PEI:M-DNA prepared at the ratio of 0.5:1.0 achieved similar levels of gene transfection compared to the PEI/B-DNA polyplex at the 1:1 (w/w) ratio, indicating that electrostatic interaction alone is insufficient to dramatically reduce the amount of the polymer required for efficient gene transfection. The C-9R-C peptide, a well-known CPP,18−20 is expected to interact with M-DNA through the combination of the LMCT transition and electrostatic interaction, while the control CPP, G-9R-G, can bind to M-DNA electrostatically. The 9R moiety binds to M-DNA through electrostatic interaction, while the cysteines located at both ends of the C-9R-C interact with the Zn2+ ions present in the M-DNA via the LMCT transition.12,21 On the other hand, the G-9R-G peptide reacts with the MDNA only through a charge interaction, due to the absence of cysteine. Compared to the G-9R-G peptide, the C-9R-C peptide would therefore form a stronger bond with the MDNA. Scanning of the electronic absorption spectrum in the far-UV region was performed to verify the binding of Zn2+ to the thiol groups in the C-9R-C peptide. This scanning of the absorption spectrum demonstrated that the LMCT transition was centered near 230 nm (Figure 1a). A bathochromic shift of the center of
Figure 1. Spectrophotometric titration of C-9R-C with Zn2+. (a) UV spectra obtained by the incremental addition of Zn2+. The arrow indicates the bathochromic shift. (b) Free cysteine concentration remained after the addition of different amounts of C-9R-C peptide to the reaction buffer, the M-DNA, or the reaction buffer containing Zn2+. The M-DNA was prepared in the presence of 7.32 mM ZnCl, and the same concentration of ZnCl without DNA was used as a control.
the bands was also observed by the analysis of the UV spectra recorded with a fixed amount of C-9R-C throughout the titration of Zn2+ ions. An Ellman’s assay to determine the amount of free cysteine present demonstrated that there was a significant difference in the levels of free cysteine present between the concentration of cysteine added and the concentration of free cysteine remaining following interaction with the M-DNA (Figure 1b). The gradients for each plot in Figure 1b are 0.8531 for the C-9R-C only, 0.5936 for the C-9RC + Zn2+, and 0.6908 for the C-9R-C + the M-DNA, confirming that the amount of free cysteine present was 99
DOI: 10.1021/acsmacrolett.6b00865 ACS Macro Lett. 2017, 6, 98−102
Letter
ACS Macro Letters
Figure 2. C-9R-C/M-DNA complex. Agarose gel electrophoresis of the M-DNA condensed by (a) C-9R-C or (c) G-9R-G and B-DNA condensed by (b) C-9R-C or (d) G-9R-G. Gel retardation of (e) the C-9R-C/M-DNA and (f) the G-9R-G/M-DNA at pH 7.4 or at pH 5.0. (g) Transfection efficiency of the C-9R-C/M-DNA complex relative to that of the G-9R-G/M-DNA complex prepared at pH 7.4 or at pH 5.0. (h) Luciferase activity and (i) cell viability following transfection of the C-9R-C/B-DNA complex or the C-9R-C/M-DNA complex (**p < 0.01, M-DNA vs B-DNA). The numbers indicate the weight ratios of C-9R-C to DNA. The M-DNA was prepared in the presence of 7.32 mM ZnCl.
(Figure 3c). The absorption at 230 nm, where the bathochromic shift of the center of the bands was observed, was increased as a function of the Zn/Cys binding (Figure 3d). These results provide evidence for the binding of C-RGD-C to the M-DNA through the LMCT transition. Finally, we verified the capacity for targeted gene expression by the formation of the C-RGD-C/M-DNA complex in cancer cells. MDA-MB-231, a breast cancer cell line, and a human embryonic kidney 293 (HEK293) cell line, both of which have been used widely in general gene transfection studies, were transfected with one of the following groups: (1) the C-RGDC/M-DNA complex; (2) the parent B-DNA mixed with CRGD-C peptide; and (3) the M-DNA. The hydrodynamic diameters and the zeta potential of the M-DNA, the Zn/CRGD-C complex, and the C-RGD-C/M-DNA complex are shown in Supporting Information Table 1. In cancer cells, the levels of gene expression relative to the M-DNA were increased as the ratio of C-RGD-C to the M-DNA increased in the CRGD-C/M-DNA complex group, whereas addition of C-RGDC to the B-DNA did not lead to a meaningful increase in gene expression (Figure 4a). The viability of the cells was not affected by the transfection of the C-RGD-C/M-DNA complex (Figure 4c). This indicates that the gene expression is increased as a consequence of the formation of the C-RGD-C/M-DNA complex but not other factors that may arise from the addition of the C-RGD-C peptide to the B-DNA. The transfection studies in the HEK293 cells further confirm the targeting ability of the C-RGD-C/M-DNA complex. In the HEK293 cells, there was no difference in the levels of luciferase activity between the C-RGD-C/M-DNA complex and the M-DNA (Figure 4b,d), which means that the integration of the C-RGD-C peptide within the M-DNA had no effect on the gene transfection into the nontarget HEK293 cells. In addition, we compared the cellular uptake of the C-RGD-C/M-DNA complex to that of the C-RGD-C + B-DNA in MDA-MB-231 cells and HEK293 cells. The C-RGD-C/M-DNA complex internalized into the cancer cells more efficiently than the B-DNA mixed with CRGD-C, whereas no difference in the cellular uptake between the C-RGD-C/M-DNA and the C-RGD-C + B-DNA was observed in the normal cells (Supporting Information Figure 6). We have explored the feasibility of the LMCT transition as a novel means to directly incorporate functional peptides, regardless of their charge density, within DNA. Cationic peptides containing cysteine(s) can facilitate and strengthen the binding to M-DNA through the LMCT transition in
most validated targeting peptide for cancer, was introduced to M-DNA via the LMCT transition. The parent B-DNA or the M-DNA was mixed with the CRGD-C peptide at different weight ratios and then electrophoresed on an agarose gel. Migration of the B-DNA was not retarded even at the highest ratio of the peptide. Addition of CRGD-C to the M-DNA, however, partially retarded the MDNA migration as the amount of the peptide increased (Figure 3a). We recorded the electronic absorption spectrum in the far-
Figure 3. Generation of the C-RGD-C/M-DNA complex. (a) Agarose gel electrophoresis of the M-DNA or the B-DNA modified with the CRGD-C. The M-DNA prepared in the presence of 7.32 mM ZnCl was reacted with C-RGD-C peptide at varying weight ratios of the peptide to the M-DNA. (b) UV spectra of the C-RGD-C/M-DNA complex at the weight ratio of 0.5/1.0. (c) UV-difference spectra generated by subtracting the absorbance of the peptide from each spectrum recorded in the presence of Zn2+. (d) Absorbance changes at 230 nm as a function of the Zn/Cys ratio.
UV region to verify the LMCT transition between C-RGD-C peptide and Zn2+ ions present in the M-DNA, which is centered near 230 nm (Figure 3b). The UV spectra recorded with a fixed amount of C-RGD-C throughout the Zn2+ titration were analyzed to confirm the LMCT transition between CRGD-C and Zn2+. The incremental addition of Zn2+ to CRGD-C up to a Zn/Cys ratio of 0.5 led to an increase in the absorbance values near 230 nm. The absorbance spectra of the C-RGD-C/M-DNA complex acquired by subtracting the background absorbance from each spectrum show that the LMCT transition reaches steady state at a Zn/Cys ratio of 0.5 100
DOI: 10.1021/acsmacrolett.6b00865 ACS Macro Lett. 2017, 6, 98−102
Letter
ACS Macro Letters
at room temperature, the fluorescence intensity was measured using a fluorescence photometer at excitation and emission wavelengths of 516 and 598 nm, respectively, every 5 min up to 40 min. Preparation of the Peptide/M-DNA complex. Each peptide was dissolved in water and added to the M-DNA at various weight ratios of the peptide to DNA. The pH was adjusted to pH 7.4 or below pH 7.0. The mixture was reacted for 30 min at room temperature. Characterization. Five micrograms of C-RGD-C was added to 1 mL of 7.32 mM ZnCl solution containing or not containing 10 μg of DNA. The pH of the mixture was adjusted to pH 7.4 by adding NaOH and then incubated for 10 min. The size and the zeta potential of the Zn/C-RGD-C complex, the C-RGD-C/M-DNA complex, and the MDNA were determined by dynamic light scattering (Malvern Zetasizer Nano-ZS; Malvern Instruments) with three parallel measurements. UV Spectrophotometry. Absorption spectra of the peptides, the peptide/M-DNA complex, and the M-DNA in the range 200−450 nm were recorded at 25 °C with a Tecan Infinite 200 PRO NanoQuant spectrophotometer in 1 cm quartz cuvettes. Stock solutions of the peptides in 0.01 M HCl were diluted to final concentrations of 1 μM 50 mM borate, pH 7.4, in the presence of 50 μM TCEP. Transfection and Cellular Uptake. MDA-MB-231 and HEK293 cells were maintained according to the protocols provided by the ATCC. The MDA-MB-231 cell line was chosen to test the targeting ability of the C-RGD-C/M-DNA complex because this cell line is widely used in targeted gene delivery studies with the C-RGD-C peptide. HEK293 cells were used in all other transfection studies because this cell line has been used in general gene transfection studies. For the in vitro transfection studies, the M-DNA was prepared with plasmid DNA encoding luciferase. Cells were incubated for 48 h after treatment, and the luciferase activity in the cell lysates was determined according to the manufacturer’s protocol (Luciferase assay system; Promega). Yoyo-labeled plasmid DNA was prepared according to the manufacturer’s instruction (ThermoFisher) and processed to form M-DNA for the flow cytometry (FACSCanto; BD Bioscience, Heidelberg, Germany).
Figure 4. Targeted gene transfection. Luciferase activity and cell viability in (a, c) MDA-MB-231 cancer cells and (b, d) HEK293 normal cells. The M-DNA prepared in the presence of 7.32 mM ZnCl was reacted with C-RGD-C peptide at the weight ratio of the peptide to DNA 0.25 or 0.5. The B-DNA mixed with C-RGD-C peptide served as a negative control (*p < 0.05 vs M-DNA).
■
S Supporting Information *
combination with electrostatic interaction. Using this strategy, the minimal amount of C-9R-C peptide necessary for reliable gene expression was significantly decreased compared to conventional polyplex formation. The LMCT transition has been further employed to enable the direct incorporation of a targeting peptide, C-RGD-C, within the M-DNA. The C-RGDC/M-DNA complex increased gene transfection in the target cancer cells, indicating that the direct introduction of targeting peptides to M-DNA is feasible using our methodology. Consequently, the LMCT transition is a promising strategy to modify metal-bound nucleic acids without the requirement for chemical conjugation or electrostatic interaction.
■
ASSOCIATED CONTENT
The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acsmacrolett.6b00865. Generation of M-DNA, cytotoxicity of M-DNA, preparation of PEI-modified M-DNA, transfection of the PEI/M-DNA, cellular uptake of C-9R-C/M-DNA complex, cellular uptake of C-RGD-C/M-DNA complex, and characteristics of C-RGD-C/M-DNA complex (PDF)
■
AUTHOR INFORMATION
Corresponding Authors
EXPERIMENTAL SECTION
*E-mail:
[email protected]. Phone: 801-581-2870. *E-mail:
[email protected]. Phone: 801-581-5311.
Materials. ZnCl2, MTT, and PEI were purchased from SigmaAldrich (St. Louis, MO). Plasmid DNA expressing luciferase and a luciferase assay kit was obtained from Promega (Madison, WI). Peptides were synthesized by the DNA/peptide core at the University of Utah. All cell culture supplies were purchased from Life Technologies (Invitrogen, Grand Island, NY). Preparation of M-DNA. The M-DNA was prepared as described previously with some modifications.14,16,17 Ten micrograms of plasmid DNA was mixed with the ZnCl stock solution at the final Zn2+ concentrations ranging from 1.83 to 7.32 mM. Thereafter, the pH of the mixture was adjusted to pH 7.4 or below 7.0 by adding NaOH. This mixture was incubated for 20 min to generate M-DNA. Ethidium Fluorescence Assay. ZnCl was added to DNA dissolved in PBS in the presence of EtBr. The fluorescence intensity of the mixture was recorded over time. The M-DNA prepared as described above or its parent B-DNA was diluted in PBS containing ethidium at the final concentration of 4 μg/mL. During the incubation
ORCID
Young-Wook Won: 0000-0002-6424-2493 Author Contributions §
K.S. Lim and D.Y. Lee contributed equally to this work.
Notes
The authors declare no competing financial interest.
■
REFERENCES
(1) Medina-Kauwe, L. K.; Xie, J.; Hamm-Alvarez, S. Gene Ther. 2005, 12 (24), 1734−51. (2) Putnam, D. Nat. Mater. 2006, 5 (6), 439−51. (3) Whitehead, K. A.; Langer, R.; Anderson, D. G. Nat. Rev. Drug Discovery 2009, 8 (2), 129−38. 101
DOI: 10.1021/acsmacrolett.6b00865 ACS Macro Lett. 2017, 6, 98−102
Letter
ACS Macro Letters (4) Wong, S. Y.; Pelet, J. M.; Putnam, D. Prog. Polym. Sci. 2007, 32 (8−9), 799−837. (5) Park, T. G.; Jeong, J. H.; Kim, S. W. Adv. Drug Delivery Rev. 2006, 58 (4), 467−86. (6) Zhou, J.; Liu, J.; Cheng, C. J.; Patel, T. R.; Weller, C. E.; Piepmeier, J. M.; Jiang, Z.; Saltzman, W. M. Nat. Mater. 2012, 11 (1), 82−90. (7) Lee, J. B.; Hong, J.; Bonner, D. K.; Poon, Z.; Hammond, P. T. Nat. Mater. 2012, 11 (4), 316−22. (8) Lv, H.; Zhang, S.; Wang, B.; Cui, S.; Yan, J. J. Controlled Release 2006, 114 (1), 100−9. (9) Pack, D. W.; Hoffman, A. S.; Pun, S.; Stayton, P. S. Nat. Rev. Drug Discovery 2005, 4 (7), 581−93. (10) Henehan, C. J.; Pountney, D. L.; Zerbe, O.; Vasak, M. Protein Sci. 1993, 2 (10), 1756−64. (11) Koch, M.; Bhattacharya, S.; Kehl, T.; Gimona, M.; Vasak, M.; Chazin, W.; Heizmann, C. W.; Kroneck, P. M.; Fritz, G. Biochim. Biophys. Acta, Mol. Cell Res. 2007, 1773 (3), 457−70. (12) Krezel, A.; Maret, W. J. Am. Chem. Soc. 2007, 129 (35), 10911− 21. (13) Aich, P.; Labiuk, S. L.; Tari, L. W.; Delbaere, L. J.; Roesler, W. J.; Falk, K. J.; Steer, R. P.; Lee, J. S. J. Mol. Biol. 1999, 294 (2), 477−85. (14) Wood, D. O.; Dinsmore, M. J.; Bare, G. A.; Lee, J. S. Nucleic acids research 2002, 30 (10), 2244−2250. (15) Muntean, C. M.; Nalpantidis, K.; Feldmann, I.; Deckert, V. Spectroscopy 2009, 23 (3−4), 155−163. (16) Lim, K. S.; Lee, D. Y.; Valencia, G. M.; Won, Y.-W.; Bull, D. A. Adv. Funct. Mater. 2015, 25 (34), 5445−5451. (17) Lee, J. S.; Latimer, L. J.; Reid, R. S. Biochem. Cell Biol. 1993, 71 (3−4), 162−8. (18) Won, Y. W.; Kim, H. A.; Lee, M.; Kim, Y. H. Mol. Ther. 2010, 18 (4), 734−42. (19) Won, Y. W.; Yoon, S. M.; Lee, K. M.; Kim, Y. H. Mol. Ther. 2011, 19 (2), 372−80. (20) Won, Y. W.; Adhikary, P. P.; Lim, K. S.; Kim, H. J.; Kim, J. K.; Kim, Y. H. Nat. Mater. 2014, 13 (12), 1157−64. (21) Maret, W.; Li, Y. Chem. Rev. 2009, 109 (10), 4682−707. (22) Won, Y. W.; McGinn, A. N.; Lee, M.; Bull, D. A.; Kim, S. W. Mol. Pharmaceutics 2013, 10 (1), 378−85. (23) Won, Y. W.; Bull, D. A.; Kim, S. W. J. Controlled Release 2014, 195, 110−9. (24) Tanimoto, M.; Kamiya, H.; Minakawa, N.; Matsuda, A.; Harashima, H. Bioconjugate Chem. 2003, 14 (6), 1197−202. (25) van der Aa, M. A.; Koning, G. A.; d’Oliveira, C.; Oosting, R. S.; Wilschut, K. J.; Hennink, W. E.; Crommelin, D. J. J. Gene Med. 2005, 7 (2), 208−17.
102
DOI: 10.1021/acsmacrolett.6b00865 ACS Macro Lett. 2017, 6, 98−102
