Design, Synthesis, and Cellular Uptake of Oligonucleotides Bearing

Letter Cite This: Org. Lett. XXXX, XXX, XXX−XXX

pubs.acs.org/OrgLett

Design, Synthesis, and Cellular Uptake of Oligonucleotides Bearing Glutathione-Labile Protecting Groups Hisao Saneyoshi,*,† Takayuki Ohta,† Yuki Hiyoshi,† Takeo Saneyoshi,‡ and Akira Ono*,† †

Department of Material and Life Chemistry, Faculty of Engineering, Kanagawa University, 3-27-1 Rokkakubashi, Kanagawa-ku, Yokohama 221-8686, Japan ‡ Department of Pharmacology, Kyoto University Graduate School of Medicine, Kyoto 606-8501, Japan

Downloaded via TULANE UNIV on February 4, 2019 at 13:40:44 (UTC). See https://pubs.acs.org/sharingguidelines for options on how to legitimately share published articles.

S Supporting Information *

ABSTRACT: Glutathione-labile protecting groups for phosphodiester moieties in oligonucleotides were designed, synthesized, and incorporated into oligonucleotides. The protecting groups on the phosphodiester moieties were cleaved in a buffer containing 10 mM glutathione, which was used as a model of intracellular fluid. Cellular uptake of oligonucleotides bearing glutathione-labile protecting groups was strongly affected by the location and number of the protecting groups.

O

concentration nor did they exhibit the desired deprotection rate for the release of the naked oligonucleotides.33,35 In this Letter, the design, synthesis, deprotection properties, and cellular uptake of GSH-activated oligonucleotides are reported. We designed a series of GSH-labile protecting groups (X, Y, and Z) consisting of a disulfide moiety and a linker built on a phenylpropyl skeleton, as shown in Figure 1. These protecting groups should become deprotected by intracellular GSH via cleavage at the disulfide moiety to yield a phenolic intermediate, followed by intramolecular attack by the hydroxyl group on the carbon near the phosphotriester to produce native oligonucleotides. A typical example (Y) for the synthesis of the phosphoramidite units bearing the GSH-labile protecting groups (X, Y, and Z) is shown in Scheme 1. Phenol 1 was treated with methyl 3,3-dimethyl acrylate in the presence of methanesulfonic acid to give lactone 2. Lactone was treated with LiAlH4 to give the alcohol 3.38 The phenolic hydroxyl group of 3 was treated with methylthiomethyl (MTM) chloride to give 4. The primary hydroxyl group was acetylated to give 5. Subsequently, the MTM group was converted into a disulfide moiety to afford 6 by following a previously reported procedure.39 Finally, the acetyl group was deprotected to give alcohol 7, which was coupled with a thymidine derivative22 to give the desired phosphoramidite unit 8. Other phosphoramidite units bearing X and Z were synthesized using similar protocols as described in Schemes S1 and S2. According to 31P NMR spectra, though the modified phosphoramidites contained some impurities, the phosphoramidites were used directly without further purification and

ligonucleotide-based therapeutics have emerged as the next generation of chemotherapeutics. Antisense oligonucleotides, anti-miRNA oligonucleotides, and siRNAs for RNA interference appear to be the most direct therapeutic strategy for approaching RNA targets in cells.1−3 However, efficient delivery of these oligonucleotide-based drugs to their sites of action remains a major challenge. The multianionic and hydrophilic nature of oligonucleotides prevents efficient cellular internalization. In addition, oligonucleotides are unstable because they are readily digested by nucleases present in biological fluids. Several strategies such as nanoparticle or liposome encapsulation, as well as conjugation with small molecules or macromolecules, have been investigated in attempts to mitigate these drawbacks.4−8 In addition, a prodrug approach for oligonucleotides was proposed in the 1990s9−11 for improving cell membrane permeability and nuclease resistance. This strategy employs biolabile protecting groups for neutralizing the anionic charges of phosphodiester moieties. Several reports have shown that these prodrug oligonucleotides are internalized into cells without the use of a transfection reagent and are typically resistant to nucleases in biological fluids.12−35 After cellular internalization, the reversible phosphotriester−phosphodiester chimera oligonucleotides are converted to the naked phosphodiester form within the cellular environment. In this study, we focused on the use of glutathione (GSH)labile protecting groups as part of a prodrug strategy. GSH exists at a higher concentration inside cells (∼10 mM) relative to the extracellular fluid (∼10 μM).36,37 Because of this GSH concentration gradient, GSH-labile protecting groups on prodrug oligonucleotides should be stable in the extracellular fluid and, after cellular uptake, should become deprotected by the abundant intracellular GSH. However, previous studies of GSH-labile protecting groups for phosphodiester moieties were neither sufficiently sensitive toward this difference in GSH © XXXX American Chemical Society

Received: November 2, 2018

A

DOI: 10.1021/acs.orglett.8b03501 Org. Lett. XXXX, XXX, XXX−XXX

Letter

Organic Letters

Figure 1. Expected activity of GSH toward the GSH-labile protecting groups on the oligonucleotides.

Scheme 1. Typical Route of Synthesis of the Phosphoramidites Bearing the GSH-Labile Protecting Groups (Y in This Example)

Scheme 2. GSH-Triggered Deprotection Was Performed on 10 μM Oligonucleotide Using GSH (10 μM or 10 mM) in 50 mM Sodium Phosphate Buffer (pH 7.0) containing 100 μM dT as an Internal Standard

deprotected in a 10 mM GSH solution as a model of cytosolic conditions. Figure 2A shows the reversed-phase HPLC profiles of the time course of GSH-triggered deprotection of ODN 2 in the presence of 10 mM GSH. After 1 h incubation with 10 mM GSH, the disulfide link was cleaved to give both the phenolic intermediate (ODN 2′: 5′-TTTTY′T-3′) and the fully deprotected product (5′-TTTTT-3′). After 4 h, the fully

incorporated into oligonucleotides (ODNs) as shown in Table S1. All CPG-supported oligonucleotides were treated with 28% ammonium hydroxide for release from the solid support and deprotection. Crude products were purified with reversed-phase high performance liquid chromatography (HPLC). The structures of the synthesized ODNs were confirmed by matrix-assisted laser desorption ionization-time-of-flight mass spectrometry (MALDI-TOF Mass) (Table S1), and the purity of the ODNs was determined by HPLC (Figure S1). GSH sensitivity was tested using the model oligonucleotide, ODN 2 (5′-TTTTTYT-3′), as shown in Scheme 2. As mentioned, GSH is a ubiquitous thiol in the cytosol where its concentration is very high (up to 10 mM), whereas the extracellular concentration is a thousand-fold lower.36 A desirable property of a protecting group for use in a prodrug strategy is that it remains intact in 10 μM GSH but is

Figure 2. Reversed-phase HPLC chromatograms over the time course of GSH-triggered deprotection (ODN 2 in this example). (A) A 10 mM GSH treatment was used to mimic intracellular GSH conditions. (B) A 10 μM GSH treatment was used to mimic extracellular GSH conditions. B

DOI: 10.1021/acs.orglett.8b03501 Org. Lett. XXXX, XXX, XXX−XXX

Letter

Organic Letters

of deprotection. Moreover, deprotection reactions of Y in ODN 4 (5′-TTTYTYT-3′) and ODN 5 (5′-TTYTTTYTTTYTTTYT3′) were also achieved (Figure S5). Furthermore, ODN 6 (5′Flu-TTYTTTYTTTYTTTYT-3′) showed good stability in human serum compared with that of unmodified ODN (Figure S6). Overall, Y was found to be a good candidate-protecting group for GSH-activated oligonucleotides. In addition, this protecting group showed the best deprotection rate and stability when compared with reported GSH-labile protecting groups.33,35 We next investigated the cellular uptake of GSH-activated oligonucleotides. We initially prepared three types of fluorescein-labeled isomeric oligonucleotides (Types 1, 2, and 3) as shown in Figure 4A. Type 1 oligonucleotides contained evenly spaced protecting groups along their length. Type 2 oligonucleotides contained consecutive protecting groups at the 3′ end. Type 3 ODNs were double-cluster type oligonucleotides containing a consecutive array of protecting groups at the 5′ and

deprotected product was observed. In contrast, ODN 2 was stable in the buffer containing 10 μM GSH, which corresponds to the extracellular concentration of GSH, as shown in Figure 2B. In the case of ODN 1 (5′-TTTTXT-3′), similar to ODN 2, the disulfide bond was quickly cleaved to yield a phenolic intermediate (ODN 1′: 5′-TTTTX′T-3′) (Figure S2). However, deprotection of ODN 1′ was very slow and a low yield of the deprotected product (3%) was obtained even after 8 h incubation. In the case of ODN 3 (5′-TTTTZT-3′), the disulfide bond was similarly cleaved but quickly deprotected to yield the fully deprotected product without observation of the phenolic intermediate (ODN 3′: 5′-TTTTZ′T-3′; Figure S3). However, ODN 3 was unstable in buffer containing 10 μM GSH. Time courses of the deprotection reactions of ODNs 1−3 are summarized in Figure 3. The fastest deprotection occurred for Z,

Figure 3. Time course of the conversion of ODNs 1−3 during GSHtriggered deprotection. *Conversion (%) was calculated as the area of deprotection product divided the total area of substrate, intermediate, and deprotected product.

followed by Y and then X (Z > Y ≫ X). According to the results in Figures 2, S2, and S3, in all protecting groups, the disulfide moiety was cleaved by GSH prior to deprotection. In the subsequent deprotection step, the fastest deprotection was observed for Z, followed in order by Y and X (Z > Y ≫ X). These results were explained by the Thorpe−Ingold effect (gemdimethyl effect),32,40−42 which can be used to regulate the speed of the deprotection step. In the presence of 10 μM GSH, ODN 1 and ODN 2 were intact, and only ODN 3 showed unexpected deprotection. To clarify the GSH dependency, the stability of ODNs 1, 2, and 3 in buffer without GSH was analyzed by HPLC (Figure S4). Only ODN 3 was partially deprotected under this condition. These data suggest that the unexpected deprotection was due to instability of the thioformacetal moiety, which was likely destabilized by the two methyl groups on the benzene ring. Therefore, based on these observations, the protecting group Y in ODN 2 showed appropriate properties for deprotection in terms of the desired level of sensitivity toward GSH and the rate

Figure 4. CLSM images of the cellular uptake of fluorescein-labeled oligonucleotides bearing GSH-labile protecting groups. (A) Four protecting groups (PGs) with different locations comprising the Type 1−3 ODNs. (B) Type 3 ODN with different numbers (4−6) of protecting groups (PGs). C

DOI: 10.1021/acs.orglett.8b03501 Org. Lett. XXXX, XXX, XXX−XXX

Organic Letters

■

3′ ends. In this experiment, the total number of protecting groups on all three types was fixed at four. In a second experiment, different numbers of protecting groups (4−6) were tested on the Type 3 ODNs shown in Figure 4B. As shown in Figure 4A, the Type 1 ODN showed lower cellular uptake when compared with Types 2 and 3 ODNs (ODNs 6−8). Type 2 produced the brightest apparent signal in cells. Similarly, the Type 3 ODN also produced a brighter signal in cells. A fluorescence signal in cells was also observed even when using the live cell imaging method (Figure S10−1). Flow cytometry analysis also suggested efficient uptake (Figure S8− 1). Next, we examined the effect of the number of protecting groups on cellular uptake using the Type 3 topology. Increasing the number of protecting groups from four to six (ODN 8−10) afforded improved uptake (Figure 4B). Changes in the topology from Type 1 to Type 3 increased cellular uptake even though the number of protecting groups was fixed at six (ODNs 10−12), as described in Figures S7 and S8−3. In the case of ODN 9 (five protecting groups), the green fluorescence signal was widespread throughout the cells. In the case of ODN 10 (six protecting groups), some bright spots of green fluorescence were observed, indicating focal localization points. Flow cytometry analysis showed more cellular uptake of ODN 10 compared with that of ODN 9 (Figure S8−2). Comparable data of colabeled HeLa cells including Hoechst 33258 and cell membrane specific dye for ODNs 6−12 are summarized in Figure S9. Live cell fluorescence microscopy was also performed to support actual cellular uptake without the fixation step (Figure S10). These data indicated that uptake of ODNs bearing GSH-labile protecting groups by cells was successful. Overall, a higher number of protecting groups were more effective in increasing cellular uptake, although the location of these groups along the oligonucleotide was a more important factor in determining cellular uptake. However, it remains unclear as to why the clustered groups are more effective. The lipophilicity of the protecting groups for phosphodiester moieties, which neutralizes the effect of the negative charges, may contribute to the cellular uptake of oligonucleotides. The recent discovery of a thiol−disulfide exchange mechanism for cellular uptake suggests another possible mechanism.43 Investigations to clarify the precise mechanism of cellular uptake will be reported in future studies. In summary, we have designed and synthesized oligonucleotides bearing GSH-labile protecting groups on internucleotide linkages using standard DNA synthesis procedures. The protecting group was responsive to 10 mM GSH but remained intact in 10 μM GSH. Moreover, the synthesized oligonucleotides showed different levels of cellular uptake depending on the location and number of protecting groups. These results will be useful for the future design of prodrug-type oligonucleotides. The application of GSH-labile protecting groups to the prodrug strategy of biologically active oligonucleotides is in progress in our laboratory.

■

Letter

AUTHOR INFORMATION

Corresponding Authors

*E-mail: [email protected]. *E-mail: [email protected]. ORCID

Hisao Saneyoshi: 0000-0003-4061-101X Notes

The authors declare no competing financial interest.

■ ■

ACKNOWLEDGMENTS This work was supported by JSPS KAKENHI Grant Number 17K01966 (to H.S.). REFERENCES

(1) Dias, N.; Stein, C. A. Mol. Cancer Ther 2002, 1, 347. (2) Lennox, K. A.; Behlke, M. A. Gene Ther. 2011, 18, 1111. (3) Elbashir, S. M.; Harborth, J.; Lendeckel, W.; Yalcin, A.; Weber, K.; Tuschl, T. Nature 2001, 411, 494. (4) Mintzer, M. A.; Simanek, E. E. Chem. Rev. 2009, 109, 259. (5) Ku, S. H.; Kim, K.; Choi, K.; Kim, S. H.; Kwon, I. C. Adv. Healthcare Mater. 2014, 3, 1182. (6) Lee, Y. S.; Kim, S. W. J. Controlled Release 2014, 190, 424. (7) Ozpolat, B.; Sood, A. K.; Lopez-Berestein, G. Adv. Drug Delivery Rev. 2014, 66, 110. (8) Yang, J.; Liu, H.; Zhang, X. Biotechnol. Adv. 2014, 32, 804. (9) Barber, I.; Rayner, B.; Imbach, J.-L. Bioorg. Med. Chem. Lett. 1995, 5, 563. (10) Tosquellas, G.; Alvarez, K.; Dell’Aquila, C.; Morvan, F.; Vasseur, J.-J.; Imbach, J.-L.; Rayner, B. Nucleic Acids Res. 1998, 26, 2069. (11) Bologna, J. C.; Vives, E.; Imbach, J. L.; Morvan, F. Antisense Nucleic Acid Drug Dev. 2002, 12, 33. (12) Iyer, R. P.; Yu, D.; Agrawal, S. Bioorg. Med. Chem. Lett. 1994, 4, 2471. (13) Iyer, R. P.; Yu, D.; Agrawal, S. Bioorg. Chem. 1995, 23, 1. (14) Iyer, R. P.; Yu, D.; Devlin, T.; Ho, N.-h.; Agrawal, S. Bioorg. Med. Chem. Lett. 1996, 6, 1917. (15) Krise, J. P.; Stella, V. J. Adv. Drug Delivery Rev. 1996, 19, 287. (16) Wilk, A.; Chmielewski, M. K.; Grajkowski, A.; Phillips, L. R.; Beaucage, S. L. J. Org. Chem. 2002, 67, 6430. (17) Cieślak, J.; Grajkowski, A.; Livengood, V.; Beaucage, S. L. J. Org. Chem. 2004, 69, 2509. (18) Ausin, C.; Grajkowski, A.; Cieslak, J.; Beaucage, S. L. Org. Lett. 2005, 7, 4201. (19) Poijärvi, P.; Heinonen, P.; Virta, P.; Lönnberg, H. Bioconjugate Chem. 2005, 16, 1564. (20) Shah, S.; Rangarajan, S.; Friedman, S. H. Angew. Chem., Int. Ed. 2005, 44, 1328. (21) Grajkowski, A.; Ausin, C.; Kauffman, J. S.; Snyder, J.; Hess, S.; Lloyd, J. R.; Beaucage, S. L. J. Org. Chem. 2007, 72, 805. (22) Grajkowski, A.; Cieslak, J.; Kauffman, J. S.; Duff, R. J.; Norris, S.; Freedberg, D. I.; Beaucage, S. L. Bioconjugate Chem. 2008, 19, 1696. (23) Ora, M.; Taherpour, S.; Linna, R.; Leisvuori, A.; Hietamäki, E.; Poijärvi-Virta, P.; Beigelman, L.; Lönnberg, H. J. Org. Chem. 2009, 74, 4992. (24) Zhang, N.; Tan, C.; Cai, P.; Zhang, P.; Zhao, Y.; Jiang, Y. Chem. Commun. (Cambridge, U. K.) 2009, 0, 3216. (25) Ausín, C.; Kauffman, J. S.; Duff, R. J.; Shivaprasad, S.; Beaucage, S. L. Tetrahedron 2010, 66, 68. (26) Grajkowski, A.; Cieslak, J.; Gapeev, A.; Beaucage, S. L. New J. Chem. 2010, 34, 880. (27) Kiuru, E.; Ora, M.; Beigelman, L.; Blatt, L.; Lonnberg, H. Chem. Biodiversity 2011, 8, 266. (28) Kiuru, E.; Ahmed, Z.; Lönnberg, H.; Beigelman, L.; Ora, M. J. Org. Chem. 2013, 78, 950. (29) Jain, H. V.; Takeda, K.; Tami, C.; Verthelyi, D.; Beaucage, S. L. Bioorg. Med. Chem. 2013, 21, 6224.

ASSOCIATED CONTENT

S Supporting Information *

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acs.orglett.8b03501. Schemes S1−S2, Table S1, and Figures S1−S10, and detailed experimental procedures, spectral data for all compounds, and 1H, 13C NMR, 31P spectra, and MALDITOF mass spectra (PDF) D

DOI: 10.1021/acs.orglett.8b03501 Org. Lett. XXXX, XXX, XXX−XXX

Letter

Organic Letters (30) Meade, B. R.; Gogoi, K.; Hamil, A. S.; Palm-Apergi, C.; Berg, A. v. d.; Hagopian, J. C.; Springer, A. D.; Eguchi, A.; Kacsinta, A. D.; Dowdy, C. F.; Presente, A.; Lonn, P.; Kaulich, M.; Yoshioka, N.; Gros, E.; Cui, X.-S.; Dowdy, S. F. Nat. Biotechnol. 2014, 32, 1256. (31) Leisvuori, A.; Lonnberg, H.; Ora, M. Eur. J. Org. Chem. 2014, 2014, 5816. (32) Saneyoshi, H.; Shimamura, K.; Sagawa, N.; Ando, Y.; Tomori, T.; Okamoto, I.; Ono, A. Bioorg. Med. Chem. Lett. 2015, 25, 2129. (33) Saneyoshi, H.; Kondo, K.; Sagawa, N.; Ono, A. Bioorg. Med. Chem. Lett. 2016, 26, 622. (34) Saneyoshi, H.; Iketani, K.; Kondo, K.; Saneyoshi, T.; Okamoto, I.; Ono, A. Bioconjugate Chem. 2016, 27, 2149. (35) Hayashi, J.; Samezawa, Y.; Ochi, Y.; Wada, S. I.; Urata, H. Bioorg. Med. Chem. Lett. 2017, 27, 3135. (36) Meister, A.; Anderson, M. E. Annu. Rev. Biochem. 1983, 52, 711. (37) Saito, G.; Swanson, J. A.; Lee, K. D. Adv. Drug Delivery Rev. 2003, 55, 199. (38) Amsberry, K. L.; Borchardt, R. T. J. Org. Chem. 1990, 55, 5867. (39) Semenyuk, A.; Foldesi, A.; Johansson, T.; Estmer-Nilsson, C.; Blomgren, P.; Brannvall, M.; Kirsebom, L. A.; Kwiatkowski, M. J. Am. Chem. Soc. 2006, 128, 12356. (40) Shan, D.; Nicolaou, M. G.; Borchardt, R. T.; Wang, B. J. Pharm. Sci. 1997, 86, 765. (41) Levine, M. N.; Raines, R. T. Chem. Sci. 2012, 3, 2412. (42) Jung, M. E.; Piizzi, G. Chem. Rev. 2005, 105, 1735. (43) Li, T.; Gao, W.; Liang, J.; Zha, M.; Chen, Y.; Zhao, Y.; Wu, C. Anal. Chem. 2017, 89, 8501.

E

DOI: 10.1021/acs.orglett.8b03501 Org. Lett. XXXX, XXX, XXX−XXX