Improved Cellular Activity of Antisense Peptide Nucleic Acids by

Jul 23, 2008 - We find that simply conjugating a lipid domain (fatty acid) to the cationic peptide (a CatLip conjugate) increases the biological effec...
0 downloads 0 Views 672KB Size
Download PDF
1526

Bioconjugate Chem. 2008, 19, 1526–1534

Improved Cellular Activity of Antisense Peptide Nucleic Acids by Conjugation to a Cationic Peptide-Lipid (CatLip) Domain Uffe Koppelhus,#,§ Takehiko Shiraishi,§ Vladimir Zachar,# Stanislava Pankratova, and Peter E. Nielsen* Department of Cellular and Molecular Medicine, The Panum Institute, Faculty of Health Sciences, University of Copenhagen, Blegdamsvej 3c, DK-2200 Copenhagen N, Denmark. Received February 22, 2008; Revised Manuscript Received June 20, 2008

Conjugation to cationic cell penetrating peptides (such as Tat, Penetratin, or oligo arginines) efficiently improves the cellular uptake of large hydrophilic molecules such as oligonucleotides and peptide nucleic acids, but the cellular uptake is predominantly via an unproductive endosomal pathway and therefore mechanisms that promote endosomal escape (or avoid the endosomal route) are required for improving bioavailability. A variety of auxiliary agents (chloroquine, calcium ions, or lipophilic photosensitizers) has this effect, but improved, unaided delivery would be highly advantageous in particular for future in ViVo applications. We find that simply conjugating a lipid domain (fatty acid) to the cationic peptide (a CatLip conjugate) increases the biological effect of the corresponding PNA (CatLip) conjugates in a luciferase cellular antisense assay up to 2 orders of magnitude. The effect increases with increasing length of the fatty acid (C8-C16) but in parallel also results in increased cellular toxicity, with decanoic acid being optimal. Furthermore, the relative enhancement is significantly higher for Tat peptide compared to oligoarginine. Confocal microscopy and chloroquine enhancement indicates that the lipophilic domain increases the endosomal uptake as well as promoting significantly endosomal escape. These results provide a novel route for improving the (cellular) bioavailability of larger hydrophilic molecules.

INTRODUCTION Bioavailability in general and effective cellular delivery in particular are major challenges for nucleic acid derived gene targeting agents and thus for drug discovery and development of such agents. Delivery of large DNA vectors for cellular transfection and gene therapy and of smaller oligonucleotides for antisense and RNA interference approaches is effectively accomplished by lipofection using a variety of cationic lipids (1–4). Unfortunately, this method is not without complications for in ViVo use mainly due to toxicity. Neutral (or positively charged) antisense agents such as peptide nucleic acids (PNAs) (5) or morpholino oligomers (6) cannot effectively be delivered via lipofection, but codelivery with a partially cDNA oligonucleotide is possible (7–9) as is delivery of PNA oligomers conjugated to a variety of lipophilic ligands such a fatty acids (10) or polyaromates (11). Several years ago it was discovered that a group of cationic peptides of biological origin as exemplified by the HIV derived Tat-peptide (12, 13) and the Antennapedia peptide, penetratin (14) (which is contained in a transcription factor form Drosophila), are taken up by eukaryotic cells and are able to deliver a conjugated molecular cargo to such cells. These so-called cell penetrating peptides (CPP) (15) typically contain multiple arginines, and indeed simple oligoarginines exhibit good cellular delivery activity (16–18). Although the first reports claimed that the cellular uptake of CCPs was exclusively a physical phenomenon involving binding to negatively charged sites on the membrane followed by internalization by transient membrane destabilization or by pore formation by the peptides, a range of recent studies have reached the consensus that the major * Corresponding author. Telephone: 0045 3532 7762. Fax: 0045 35396042. E-mail: [email protected]. # Laboratory for Stem Cell Research, Institute for Health Technologies, Aalborg University, Fredrik Bajers Vej 3B, DK-9220 Aalborg Ø, Denmark. § These authors contributed equally.

cellular port of entrance for the cationic CPPs is via endocytosis (19–22), maybe predominantly macropinocytosis (23–25). Consequently, delivery is to endosomes (which eventually turn into lysosomes), and therefore the biological effect is very limited, as the compounds are trapped in these compartments and cannot reach their biological target in the cytoplasm or in the nucleus. Thus one of the major challenges for exploiting CPPs for drug delivery is the discovery of methods or mechanisms that allow efficient endosomal escape of the drugs. This may be accomplished by using auxiliary agents such as chloroquine (26, 27), calcium ions (26), or lipophilic photosensitizers (28, 29), but delivery may also be improved by chemical modification of the peptide. Most obviously, addition of a (viral) endosome disruption domain to the peptide should increase efficiency as already demonstrated (25), and more complex “chimeric” peptides such as arginine modified penetratin (30) and especially transportan also appear more effective (at least in a PNA context) than simple cationic peptides (17), although the mechanism(s) of uptake is still very poorly understood. The galanin domain of transportan is rather hydrophobic while the mastoparan domain contains the only cationic moieties of this peptide. It would therefore appear that the cellular uptake of transportan could rely more on hydrophobic interactions with the membrane than on electrostatic interactions. It is of course reasonable that both membrane binding as well as membrane penetration should be facilitated by lipophilic ligands. On the other hand a cationic domain would contribute both electrostatic cellular adhesion as well as increased aqueous solubility of the ligand. The efficiency of cationic lipids for cellular delivery of (anionic) oligonucleotides and DNA is well established (31), and recently it was shown that fatty acid conjugates of oligoarginines may also be successfully exploited as delivery vehicles for DNA and RNA (32). In this case, however, the main function of the cationic peptide most likely is condensation of the anionic nucleic acids. We set out to study the cellular delivery properties of cationic peptides containing a simple lipophilic domain, in the form of a fatty acid. We have

10.1021/bc800068h CCC: $40.75  2008 American Chemical Society Published on Web 07/23/2008

Improved Cellular Activity of Antisense PNAs

Bioconjugate Chem., Vol. 19, No. 8, 2008 1527

Table 1. List of PNAs no.

sequencea

name

2395 2456 2534

Tat-PNA Deca-PNA Tat-Deca-PNA

2821

Deca-Tat-Deca-PNA

2785 2801 2786 2787 2802 2788 2752

(D-Arg)7-PNA (D-Arg)7-Deca-PNA Deca-(D-Arg)7-PNA (D-Arg)8-PNA (D-Arg)8-Deca-PNA Deca-(D-Arg)8-PNA mismatchb

2591 3100 3101 2592 2739 2740 2646 2647 2648 2822 2823 2824 2825 2826 2827

(Arg)6-Deca-PNA (Arg)7-Deca-PNA (Arg)8-Deca-PNA (Arg)9-Deca-PNA (D-Arg)6-Deca-PNA (D-Arg)9-Deca-PNA (Arg)6-Octa-PNA (Arg)6-Dodeca-PNA (Arg)6-Hexadeca-PNA (RK)3-PNA Deca-(RK)3-PNA (RK)4-PNA Deca-(RK)4-PNA (RRA)3-PNA Deca-(RRA)3-PNA

2828 2829

(RRA)4-PNA Deca-(RRA)4-PNA

2830 2831

R4K4-PNA Deca-R4K4-PNA

2832 2833

K2R4K2-PNA Deca-K2R4K2-PNA

2738 3016

(D-Arg)9-PNA Chol-(Arg)7-PNA

3266 3265

Fl-(Arg)8-PNA Fl-(Arg)8-Deca-PNA

H-GRKKRRQRRRPPQ-eg1-CCT CTT ACC TCA GTT ACA-NH2 Deca-eg1-CCT CTT ACC TCA GTT ACA-NH2 H-GRKKRRQRRRPPQ-Lys(Deca)-Gly-CCT CTT ACC TCA GTT ACA -NH2 H-Lys(Deca)-GRKKRRQRRRPPQ-Lys(Deca)-Gly-CCT CTT ACC TCA GTT ACA-NH2 H-(D-Arg)7-Gly-CCT CTT ACC TCA GTT ACA-NH2 H-(D-Arg)7-Lys(Deca)-Gly-CCT CTT ACC TCA GTT ACA-NH2 Deca-(D-Arg)7-Gly-CCT CTT ACC TCA GTT ACA-NH2 H-(D-Arg)8-Gly-CCT CTT ACC TCA GTT ACA-NH2 H-(D-Arg)8-Lys(Deca)-Gly-CCT CTT ACC TCA GTT ACA-NH2 Deca-(D-Arg)8-Gly-CCT CTT ACC TCA GTT ACA-NH2 H-GRKKRRQRRRPPQ-Lys(Deca)-Gly-CCT CTG ACC TCA TTT ACA-NH2 H-(Arg)6-Lys(Deca)-Gly-CCT CTT ACC TCA GTT ACA -NH2 H-(Arg)7-Lys(Deca)-Gly-CCT CTT ACC TCA GTT ACA -NH2 H-(Arg)8-Lys(Deca)-Gly-CCT CTT ACC TCA GTT ACA -NH2 H-(Arg)9-Lys(Deca)-Gly-CCT CTT CTT ACC TCA GTT ACA -NH2 H-(D-Arg)6-Lys(Deca)-Gly-CCT CTT ACC TCA GTT ACA-NH2 H-(D-Arg)9-Lys(Deca)-Gly-CCT CTT ACC TCA GTT ACA-NH2 H-(Arg)6-Lys(Octa)-CCT CTT ACC TCA GTT ACA-NH2 H-(Arg)6-Lys(Dodeca)-CCT CTT ACC TCA GTT ACA-NH2 H-(Arg)6-Lys(Hexadeca)-CCT CTT ACC TCA GTT ACA-NH2 H-(D-Arg D-Lys)3-Gly-CCT CTT ACC TCA GTT ACA-NH2 H-Lys(Deca)-(D-Arg D-Lys)3-Gly-CCT CTT ACC TCA GTT ACA-NH2 H-(D-Arg D-Lys)4-Gly-CCT CTT ACC TCA GTT ACA-NH2 H-Lys(Deca)-(D-Arg D-Lys)4-Gly-CCT CTT ACC TCA GTT ACA-NH2 H-(D-Arg D-Arg Ala)3-Gly-CCT CTT ACC TCA GTT ACA-NH2 H-Lys(Deca) (D-Arg D-Arg Ala)3 Gly-CCT CTT ACC TCA GTT ACA-NH2 H-(D-Arg D-Arg Ala)4-Gly-CCT CTT ACC TCA GTT ACA-NH2 H-Lys(Deca) (D-Arg D-Arg Ala)4-Gly-CCT CTT ACC TCA GTT ACA-NH2 H-(D-Arg)4(D-Lys)4-Gly-CCT CTT ACC TCA GTT ACA-NH2 H-Lys(Deca) (D-Arg)4(D-Lys)4-Gly-CCT CTT ACC TCA GTT ACA-NH2 H-(D-Lys)2(D-Arg)4(D-Lys)2-Gly-CCT CTT ACC TCA GTT ACA-NH2 H-Lys(Deca) (D-Lys)2(D-Arg)4(D-Lys)2-Gly-CCT CTT ACC TCA GTT ACA-NH2 H-(D-Arg)9-Gly-CCT CTT ACC TCA GTT ACA-NH2 Cholesteryl hemisuccinate-(Arg)7-eg1-CCT CTT ACC TCA GTT ACA-NH2 H-Flk-(D-Arg)8-AAT CTC ACC TGA TAG T-NH2 H-Deca-Flk-(D-Arg)8-AAT CTC ACC TGA TAG T-NH2

a The sequences of the PNAs are written from N-terminal to C-terminal end. Cell penetrating peptides, decanoic acid (Deca) or Lys(Deca) moietiy was covalently linked to PNA at the N-terminal through an ethyleneglycol linker (eg1: 8-amino-3,6-dioxaoctanoic acid) or glycine (Gly) via continuous synthesis. b Mismatch PNA for the Tat-Deca-PNA (PNA2534) whereas two mismatches are indicated in bold. Flk, fluorescein (Fl) attached to the ε-amino group of a lysine residue (Lohse, J., Nielsen, P. E., Harrit, N., Dahl, O. (1997) Bioconjugate Chem. 8, 503-509).

characterized such two-domain “CatLip” peptides in a PNA antisense context (Table 1) using the quantitative and very sensitive and accurate mRNA splicing correction luciferase pLuc HeLa cell system (33, 34), and we find that adding a fatty acid moiety increases the biological activity of Tat and oligoarginine CPP-PNA conjugates by up to 2 orders of magnitude.

EXPERIMENTAL PROCEDURES PNA Synthesis. The sequences of the PNAs used are listed in Table 1. PNA synthesis was carried out by the standard Boc strategy as reported previously (35). Peptides were linked to the PNA at the N-terminal through an ethylene glycol linker (eg1: 8-amino-3,6-dioxaoctanoic acid) or glycine via continuous solid phase synthesis. The fatty acid was conjugated to the ε-amino group of a lysine moiety during synthesis using orthogonal Fmoc protection. The PNA conjugates were purified by HPLC and characterized by MALDI-TOF mass spectrometry (see Supporting Information). The PNAs were lyophilized and stored at 4 °C until use. Cell Culture. HeLa pLuc705 cells were purchased from Gene Tools (USA). Cells were grown in RPMI1640 medium (Sigma) supplemented with 10% fetal bovine serum (FBS, Sigma), 1%

glutamax (Gibco), 100 U/mL penicillin (Gibco), and 100 µg/ mL streptomycin (Gibco) at 37 °C in humidified air with 5% CO2. For studies of HeLa pLuc705 cells in 96 well plate format, cells were trypsinized and seeded (1.2 × 104 cells/well) 16 h before treatment. For studies in 24-well plate, 7.2 × 104 cells/ well were used. PNA Treatment. The cells, replated in 96 well (Nunc) or 24 well plates (Nunc) 16-24 h earlier, were treated with 100 µL/well (for 96 well plate) or 0.3 mL/well (for 24 well plate) of OPTI-MEM (Gibco) containing the PNA at the desired concentration for 4 h. After PNA treatment, cells were supplemented with the 100 µL/well (for 96 well plate) or 0.3 mL/ well (for 24 well plate) of RPMI1640 medium (containing 20% FBS and 1% Glutamax), incubated further for 24 h, and subjected to the desired analysis unless otherwise stated. For the endosome disruption treatments, cells were treated as reported previously (36). Briefly, for photochemical internalization (PCI) treatment, cells were plated the day before PNA treatment with the antibiotics free growth medium containing 5 µg/mL aluminum phthalocyanine (AlPcS2a, Frontier Scientific, Logan, USA). After the 4 h PNA treatment, the medium was replaced with fresh growth medium (RPMI1640 containing 10%

1528 Bioconjugate Chem., Vol. 19, No. 8, 2008

Figure 1. Relative cellular luciferase antisense activity and cellular toxicity in HeLa pLuc705 cells of PNAs conjugated to Tat-peptide and/ or decanoic acid (attached to the ε-amino group of a lysine). Cells plated in the 96 well plates were treated with PNAs (Tat-PNA, Deca-PNA, Tat-Deca-PNA, and Deca-Tat-Deca-PNA) for 4 h in OPTI-MEM medium at the designated concentrations (0-8 µM) and incubated further for 24 h after addition of serum-containing growth medium (without removing the PNA solution). The cell samples were then subjected to luciferase analysis and cellular viability test. All tests were performed in triplicate and the results are given as the average values ( standard deviations (SD). (A) Luciferase activities were analyzed by Blight-Gro reagent (Promega) and given as relative light units (RLU/ well). (B) Cellular viabilities were analyzed by MTS assay (Promega) and obtained values were normalized to the average value of non-PNAtreated sample.

Figure 2. Relative cellular luciferase antisense activity in HeLa pLuc705 cells of PNAs conjugated to a cell penetrating peptide ((D-Arg)7-, (DArg)8-, or Tat-) and/or a decanoic acid (attached to the ε-amino group of a lysine). Cells plated in 96 well plate were treated with PNAs (0-8 µM) for 4 h in OPTI-MEM medium and incubated further for 24 h after addition of serum-containing growth medium without removing the transfection solution. Then the cell samples were subjected to luciferase analysis. Luciferase activities were analyzed by Blight-Gro reagent (Promega) and given as relative light units (RLU/well). All tests were performed in triplicate and the results are given as the average values ( standard deviations (SD).

FBS and 1% glutamax) and incubated for 4 h before irradiation with a fluorescence light tube (Alcadia, FO18) at 7.77 µW/cm2 at 660 nm (15 nm bandwidth). Following irradiation, cells were incubated for additional 24 h before subjected to further analysis. For the chloroquine (CQ) treatment and Ca2+ treatment, cells were treated with PNA in OPTIMEM containing 120 µM CQ or 6 mM CaCl2, respectively.

Koppelhus et al.

Figure 3. (A) Confocal fluoresence microscopy analysis of the cellular uptake of PNAs in HeLa pLuc705 cells. The fluorescein (Fl)-labeled PNAs (Fl-(Arg)8-PNA (PNA3266) and Fl-(Arg)8-Deca-PNA (PNA3265) were used. The cells were treated with 0.5 µM of PNA together with 20 µg/mL of tranferrin (Tf) for 24 h and analyzed by confocal fluoresence microscopy. Merged images (merged) and pseudocolor (ps) images were created by using the Lasersharp 2000 software package (BioRad). (B) Flow cytometric analysis of HeLa pLuc705 cells treated with PNAs (Fl-(Arg)8-PNA (CPP) or Fl-(Arg)8-Deca-PNA (CL)) at three different concentrations (0.3, 1, and 3 µM) for 24 h and collected by pipetting using 1% BSA solution (in PBS) for the flow cytometric analysis. Cells (10 000 events for each sample) were analyzed using BD FACSCalibur cytometer and the data were analyzed by the CellQuest software. The histogram data of the cells treated with the PNAs at three different concentrations were superimposed. (C) Histogram representation of mean fluorescence of the cells.

Luciferase Assay. At 24 h after the PNA treatment, the cells were subjected to luciferase activity analysis by using the BrightGlo Luciferase assay system (Promega) or the Luciferase Assay System (Promega) for the cells in 96 well plates or the cells in 24 well plates, respectively. Measurements were performed according to the manufacturer’s instructions. Luminescence readings obtained with the Bright-Glo-Luciferase assay system were background-subtracted and are presented as relative light units (RLU/well). For the luciferase assay with the Luciferase Assay System, the cells were lysed with 0.1 mL/well of Passive Lysis Buffer (Promega) and used for both luciferase analysis and RNA extraction as mentioned below. Luminescence readings obtained with the Luciferase Assay System were backgroundsubtracted and normalized by protein concentrations and were presented as RLU/µg protein. Protein concentrations were determined by the BCA protein assay (Pierce) according to the manufacturer’s instruction.

Improved Cellular Activity of Antisense PNAs

Bioconjugate Chem., Vol. 19, No. 8, 2008 1529

Figure 4. Relative cellular luciferase antisense activity in HeLa pLuc705 cells of PNAs conjugated to a cell penetrating peptide ((D-Arg)7-, (DArg)8-, or Tat-) and/or a decanoic acid (attached to the ε-amino group of a lysine). The cells were treated as described in Figure 1. Then cells were subjected to the further analysis. (A and B) PNAs (Deca-Tat-Deca-PNA (PNA2821), Tat-Deca-PNA (PNA2534), (D-Arg)7-Deca-PNA (PNA2801), (D-Arg)8-Deca-PNA (PNA2802), and (D-Arg)8-PNA (PNA2787)) were used at the indicated concentrations (0-6 µM). (A) Luciferase activity was measured by Luciferase assay system (Promega), normalized to protein concentration and given as RLU/ µg of protein. (A) RT-PCR analysis of the splicing correction of pre-mRNAs by PNAs. Total RNA was extracted from the cells after the PNA treatment and subjected to RT-PCR analysis. Uncorrected indicates the 268 bp fragment without mis-splicing correction, and corrected indicates the 142 bp correctly spliced fragment. Numbers under each lane indicate the amount of the corrected form relative to the sum of corrected form and uncorrected form. (C and D) Comparison of Tat-Deca-PNA (PNA2534) with its mismatch PNA (mismatch, PNA2752). PNAs were used at the indicated concentrations (0-8 µM). Analysis was performed as described in (A and B). (C) Luciferase activity was measured and normalized to protein concentration. (D) RT-PCR analysis of mis-splicing correction by PNAs.

Cytotoxicity Test. Cells in 96 well plates were subjected to the MTS assay for cell viability test by using the CellTiter 96 Aqueous Non-Radioactive Cell Proliferation Assay (Promega) according to the manufacturer’s instructions. The absorbance is presented as relative cellular viability (absorbance from nonPNA treated cells was set as 100%). RT-PCR. Total RNA was extracted from the cellular lysate (see the luciferase assay section) by using RNeasy Mini kit (Qiagen) and subjected to RT-PCR analysis. A total of 3 ng RNA was used for each RT-PCR reaction (20 µL). RT-PCR was performed by using OneStep RT-PCR kit (Qiagen) following the manufacture’s instructions. Primers for the RT-PCR were as follows: forward primer, 5′-TTGATATGTGGATTTCGAGTCGTC-3′; reverse primer, 5′-TGTCAATCAGAGTGCTTTTGG-CG-3′. The RT-PCR program was as follows: [(55 °C, 35 min) × 1 cycle, (95 °C, 15 min) × 1 cycle, (94 °C, 0.5 min; 55 °C, 0.5 min; 72 °C, 0.5 min) × 29 cycles]. RT-PCR products were analyzed on 2% agarose gel with 1× TBE buffer and visualized by ethidium bromide staining. Gel images were captured by ImageMaster (Pharmacia Biotech) and analyzed by UN-SCAN-IT software (Silk Scientific corporation). Confocal Microscopy. Exponentially growing HeLa pLuc705 cells were plated in 8 well Laboratory-Tek Chambered Coverglass (Nunc) at a cell density of 5 × 104 cells/well the day

before transfection. Following 16-24 h incubation, the cells were treated with PNA by incubation in OPTI-MEM medium (0.25 mL/well) containing the fluorescein-labeled PNA at the desired concentration and 20 µg/mL Alexa Fluor 633-conjugated human transferrin (80 kDa, Molecular Probes) for 4 h. Then the cells were incubated further for 20 h after supplement of RPMI 1640 growth medium containing 20% FBS (0.25 mL/ well). Subsequently, the cells were washed with Hank’s solution and subjected to microscopy analysis in this solution using a MultiProbe 2001 laser scanning confocal system equipped with an argon laser and a red laser diode (excitation wavelength 488 and 638 nm, Radiance2000, BioRad) connected to a Nicon eclipse TE200 microscope (oil immersion 60× 1.4 NA objective, Nikon, Tokyo, Japan). The Lasersharp 2000 software package (BioRad) was used for image acquisition and processing. Flow Cytometry. Exponentially growing HeLa pLuc705 cells were plated in 24 well plates at 1.0 × 105 cells/well the day before PNA treatment. After 16-24 h incubation, the cells were treated with fluorescein-labeled PNA as described above. After 24 h, the cells were washed twice with phosphate buffered saline (PBS) containing 1% BSA and collected by pipetting from the BSA/PBS solution. Then the cells were filtered with the cell strainer (70 µM, BD falcon) and subjected to flow cytometric analysis (10 000 events for each samples) using a BD FACS-

1530 Bioconjugate Chem., Vol. 19, No. 8, 2008

Figure 5. Relative cellular luciferase antisense activity in HeLa pLuc705 cells of PNAs conjugated to decanoic acid and different L-arginine or oligo D-arginine peptides: PNAs 2591 (R6, (Arg)6-Deca-PNA), 3100 (R7, (Arg)7-Deca-PNA); 3101(R8, (Arg)8-Deca-PNA); 2592 (R9, (Arg)9-Deca-PNA); 2739 (r6, (D-Arg)6-Deca-PNA); 2801 (r7, (D-Arg)7Deca-PNA); 2802 (r8, (D-Arg)8-Deca-PNA), and 2740 (r9, (D-Arg)9Deca-PNA) were used. Experimental details as in Figure 1. All tests were performed in triplicate and the results are given as the average values ( standard deviations (SD). (A) Luciferase activities were analyzed by Blight-Gro reagent (Promega) and given as relative light units (RLU/well). (B) Cellular viability was analyzed by the MTSassay (Promega) and the data were normalized to the average value of non-PNA-treated sample.

Calibur cytometer, and the mean fluorescence signal was calculated by CellQuest software.

RESULTS The Tat peptide belongs to the group of cell penetrating peptides that was originally introduced for cellular delivery (13) and for which it has subsequently been shown that, although cellular uptake is quite efficient, this is almost exclusively via an endosomal pathway (19, 21). Thus antisense PNA-Tat conjugates show only very limited activity in cell culture (17). However, the activity is dramatically enhanced by endosome disruption agents such as chloroquine (26) or lipophilic photosensitizers (29), as well as by calcium ions (26). Therefore we chose to prepare antisense PNA-Tat-fatty acid conjugates targeted to the aberrant splice site in the engineered luciferase gene in pLucHeLa 705 cells. This antisense assay yields a positive readout (luciferase activation) that is readily quantified with high sensitivity (33, 34), and therefore allows reliable and quantitative comparison of different delivery methods (for antisense agents) (9, 11). The results presented in Figure 1 clearly demonstrate that while the PNA-Tat conjugate shows only very weak antisense activity in this assay, the analogous conjugate containing a decanoic acid attached to the ε-amino group of a lysine (TatDeca-PNA) is dramatically more active. Thus, at 2, 4, and 8 µM, the PNA-Tat conjugate induced luciferase activities of 80, 165, and 2832 RLU; while the corresponding values for the Tat-Deca-PNA were 1275, 11 696, and 22 747, respectively. The activity is further enhanced by having two decanoic acids in the conjugate (Deca-Tat-Deca-PNA) resulting in a luciferase

Koppelhus et al.

Figure 6. Relative cellular luciferase antisense activity and cellular toxicity in HeLa pLuc705 cells of hexaarginine-PNAs conjugated to different fatty acids (octanoic acid, decanoic acid, dodecanoic acid, or hexadecanoic acid) for the cellular uptake and the cellular toxicity in the HeLa pLuc705 ells. Experimental details as in Figure 1 using PNAs 2646 ((Arg)6-Octa-PNA), 2591 ((Arg)6-Deca-PNA), 2647 ((Arg)6Dodeca-PNA), and 2648 ((Arg)6-Hexadeca-PNA). All tests were performed in triplicate and the results are given as the average values ( standard deviations (SD). (A) Luciferase activity was analyzed by Blight-Gro reagent (Promega) and is presented as relative light units (RLU/well). (B) Cellular viability was analyzed by MTS assay (Promega) and the data normalized to the average value of non-PNAtreated sample.

activity of 14 310 at 2 µMsmore than 150 times higher than that found for the pure Tat-PNA conjugate at this concentration. However, increased cellular toxicity is observed for the Tat-fatty acid conjugates and this is especially pronounced for the difatty acid compound. Thus, at higher concentrations (4 and 8 µM) the luciferase activation by this conjugate is markedly reduced, due to cellular toxicity. The control fatty acid-PNA conjugate (Deca-PNA) showed no activity. In order to address whether conjugation of a lipophilic moiety (a fatty acid) to a cationic cell penetrating peptide domain could be a general principle to obtain increased cellular bioavailability we made a series of PNA conjugates based on oligo-arginines. The results presented in Figure 2 show that although hepta- and octa-arginine PNA conjugates as reported previously (17) show significantly higher antisense activity than the PNA-Tat conjugate, the activity is further enhanced several fold by fatty acid conjugation, and this enhancement is more pronounced at lower concentrations (1 and 2 µM), but is not seen at 8 µM due to increased toxicity of the fatty acid conjugates (Figures 2 and 5A,B). We also note that although the Tat-Deca-PNA conjugate is slightly less active than the Arg-Deca-PNA conjugates at lower concentrations, it exhibits the highest activity at 8 µM because of lower toxicity. The results also show that the position of the fatty acid relative to the peptide (amino terminal (e.g., Deca-(D-Arg)7-PNA) or C-terminal between the peptide and the PNA (e.g., (D-Arg)7-Deca-PNA) does not seem of major importance for activity. In order to trace the intracellular path of the CatLip PNA a series of experiments were performed using fluorescein labeled

Improved Cellular Activity of Antisense PNAs

Bioconjugate Chem., Vol. 19, No. 8, 2008 1531

Figure 7. Relative cellular luciferase antisense activity in HeLa pLuc705 cells of PNAs conjugated to a cell penetrating peptide (CPP) or conjugated to both CPP and a decanoic acid. Experimental details as in Figure 1 using PNAs (RK)3-PNA, Deca-(RK)3-PNA, (RK)4-PNA, Deca-(RK)4-PNA, (RRA)3-PNA, Deca-(RRA)3-PNA, (RRA)4-PNA, Deca(RRA)4-PNA, R4K4-PNA, Deca-R4K4-PNA, K2R4K2-PNA, Deca-K2R4K2-PNA, TatPNA, Tat-Deca-PNA, (D-Arg)9-PNA, (D-Arg)9-Deca-PNA). Luciferase activity was analyzed by Blight-Gro reagent (Promega) and are presented as relative light units (RLU/well). All tests were performed in triplicate and the results are given as the average values ( standard deviations (SD).

octa-arginine PNA conjugates in combination with FACS and fluorescence microscopy analyses. These experiments first of all demonstrate that the lipophilic domain significantly increases the cellular attachment and possibly the uptake of, in this case, the PNA as indicated from the increased intracellular fluorescence observed by confocal fluorescence microscopy (Figure 3A), and quantitatively confirmed by FACS analysis (Figure 3B,C). However, it is quite clear from the fluorescence micrographs that the major part of the fluorescence still appears associated with endosomes, although some increase in diffuse cytoplasmatic and nuclear fluorescence may be present. Therefore, these results would indicate that the enhancing effect of the lipophilic domain on the nuclear antisense activity has contributions from both an increase in cellular (endosomal) uptake as well as a more favorable intracellular distribution with (slightly) reduced endosomal confinement, although even for these improved conjugates, most of the uptake is still within endosomal entrapment (Vide infra). The luciferase activity reading is a secondary effect from the molecular event of splicing redirection by the PNA, and we therefore found it important to verify the antisense effect directly at the mRNA level. The results show a very good correlation between the luciferase activity (Figure 4A) and the relative amount of correctly spliced luciferase mRNA (Figure 4B) at PNA concentrations up to 4 µM. However, at 6 µM where toxicity becomes a limiting factor and where the luciferase readings are reduced for two of the conjugates (Deca-Tat-DecaPNA and (D-Arg)8-Deca-PNA), the RNA data shows increased relative correction activity also for these two PNAs, indicating an expected dose response on splice correction in spite of a decreased amount of total mRNA due to general toxicity. The RNA data also confirm the strong enhancement exerted by the lipid moiety. Thus at 1 µM a relatively efficient peptide such as (D-Arg)8 only promotes 3% splicing correction when conjugated to the PNA, while (D-Arg)8-Deca-PNA under similar conditions promotes 44% splicing correction, corresponding to more than 10-fold enhancement. At higher concentrations the relative enhancement decreases gradually to less than 2-fold at 6 µM (Figure 4B). The CatLip conjugates are amphiphiles and thus most probably shall form micelles at a well defined cmc (critical micelle concentration). If micelle formation was instrumental for delivery efficiency, a clear concentration

threshold (cooperative) effect would be expected. However, the data (e.g., Figure 5) do not specifically support such an effect, although some experiments may hint at nonlinearity of the dose dependence (e.g., Figures 1and 4). Thus more elaborate dose response studies are warranted to specifically address the issue of possible micelle formation by the conjugates and the eventual importance of such putative micelles on cellular uptake and also on toxicity. In this context it is also worth noting that toxicity of the conjugates does appear to be correlated to disruption of the cell membrane integrity as assayed using propidium iodide staining (Supporting Information) (which very much parallel the MTS toxicity assay), thereby indicating a direct physical membrane disturbance by the amphiphilic CatLip conjugates. As a control for the sequence specificity of the antisense effect we synthesized a double base pair mismatch derivative of identical base composition (two base interchange) of the TatDeca-PNA and tested this. As expected, this PNA exhibited dramatically reduced activity both in terms of luciferase activity as well as in terms of mRNA correction activity (Figure 4C,D). A comparison between D-Arg and L-Arg conjugates (Figure 5A) shows slightly higher activity of the non-natural D-form, clearly arguing against the involvement of specific receptor interactions of the peptide. These data also stress the very fine balance between uptake and toxicity. At 2 µM the activity increases with increasing number of arginines, but as the concentration increases, this order changes as the conjugates of the longer oligo-arginines are relatively more toxic than those of the shorter ones. Finally, the D-Arg conjugates show higher antisense activity but in parallel are also more toxic to the cells (Figure 5B). This delicate activity/toxicity balance is further illustrated when analyzing the effect of the length of the fatty acid. For this experiment we chose the less active and less toxic hexaarginine conjugate (Figure 6). Clearly, at lower PNA concentrations (up to 2 µM) the conjugates with longer fatty acid tails are the more active, but as the dodeca- and hexadeca- conjugates are clearly more toxic than the deca- conjugate, the latter shows the highest activity at 4 µM. Thus increased fatty acid length, and thus lipophilicity, produces more potent antisense agents but also compounds of higher cellular toxicity, and decanoic acid appears as a good compromise. A small series of argininerich peptides were tested for enhancement by fatty acid acylation

1532 Bioconjugate Chem., Vol. 19, No. 8, 2008

Koppelhus et al.

Figure 8. Effects of the endosome disruption agents on the cellular luciferase antisense activity in HeLa pLuc705 cells of PNA conjugates. HeLa pLuc705 cells (plated in the 96 well plate) were treated with the PNA in combination with one of the endosome disruption treatments (chloroquine (CQ), photochemical treatment (PCI) or Ca2+ treatment) and subjected to luciferase analysis. Luciferase activity was analyzed by Blight-Gro reagent (Promega) and are presented as relative light units (RLU/well). All tests were performed in triplicate and the results are given as the average values ( standard deviations (SD). (A) Cells were treated with the PNA (Tat-PNA, Tat-Deca-PNA) in combination with the Ca2+ treatment. For the Ca2+ treatment, cells were incubated with the PNA for 4 h in OPTI-MEM medium containing 6 mM Ca2+ and incubated further for 24 h after addition of the serum-containing growth medium without removing the OPTI-MEM medium. (B) Cells were treated with the PNA (2395, Tat-PNA; 2534, Tat-Deca-PNA) in combination with PCI treatment. For the administration of the photosensitizer, cells were incubated with 5 µg/mL AlPcS2a in the growth medium for 16-24 h before being subjected to the PNA treatment. Cells were then treated with the PNA for 4 h in OPTI-MEM medium and incubated further for 4 h after the replacement of the PNA solution with fresh serum-containing growth medium. Then cells were subjected to 10 min red light irradiation and incubated further for 24 h before being subjected to the luciferase analysis.

as illustrated by the data presented in Figure 7. It is clear that for all conjugates the acylated form exhibits superior activity relative the nonacylated form, but it is also noteworthy that the relative enhancement differs significantly for the various peptides as already demonstrated by the difference between the Tat and oligo-arginine conjugates. In particular, the very dramatic enhancement (up to 2 orders of magnitude) of the (Arg-ArgAla)4 conjugate upon fatty acid conjugation is interesting in terms of the molecular mechanism(s) of the cellular uptake as this conjugate resembles the behavior of the Tat peptide more than that of the oligo-arginines. These data indicate a clear possibility for optimizing the peptide-fatty acid (CatLip) moiety in terms of delivery/toxicity ratio. We have previously found that the antisense activity of CPP-PNA conjugates is significantly augmented in the presence of submillimolar concentrations of chloroquine or millimolar concentrations of calcium ions (26) or by photochemical internalization (PCI) using photosensitizers (29), agents that all promote endosome disruption. In order to address whether these agents could also enhance the effect of the fatty acid conjugates, we studied the effect of chloroquine, Ca2+, and PCI treatment on the antisense activity of the Tat-Deca-PNA conjugate. In all cases a significant enhancement was observed especially at lower concentrations (Figure 8). However, the relative enhancement is much more pronounced for the Tat-PNA (10-25-fold) than

for the Tat-Deca-PNA (1-15-fold) (Figure 8A), possibly indicating that the fatty acid moiety is facilitating endosomal escape or is supporting a nonendosomal pathway for uptake. Interestingly, we note that a CatLip construct based on a cholesterol ester (PNA 3016) which exhibits considerably higher antisense activity than the Tat-decanoic acid conjugate (PNA 2534) (but also higher cellular toxicity) is not enhanced by chloroquine (Figure 8C). Further studies are needed to determine whether this very interesting behavior is due to a nonendosomal route of entry or to efficient endosomal escape by this type of CatLip conjugates.

CONCLUSIONS The present results clearly demonstrate that the nuclear antisense activity of PNA oligomers can be dramatically enhanced by conjugation to a CatLip domain consisting of a cationic peptide and a fatty acid, and that this enhancement (dependent on the specific peptide) is up to 2 orders of magnitude higher than that obtained by using the peptide alone. The antisense effect of the CatLip PNAs is further enhanced by endosome disruptive agents but the enhancing effect is significantly less than that seen for the corresponding PNApeptide conjugates, thereby giving a clear indication that part, but not all, of the effect of the lipophilic domain is due to facilitation of endosomal escape. Obviously the lipophilic

Improved Cellular Activity of Antisense PNAs

domain supports the interaction of the PNA conjugates with the cellular (and endosomal) membranes and most likely contributes to membrane destabilization (as also indicated by the increased cellular toxicity of these conjugates), which could also facilitate nonendosomal, direct passage through the membrane. Increased cellular association was indeed observed through FACS and fluorescence microscopy analyses. These experiments also illustrate that even for the most effective CatLip conjugates a major part of the antisense agent was still trapped in endosomes and increased activity could be obtained using endosome releasing agents such as chloroquine. The cholesteryl conjugate (which also exhibit high toxicity) could be an example of a conjugate for which the endosomal route is not the more significant one. Clearly, the present results open the door to a new family of drug delivery conjugates, and further studies will reveal the interplay between different (peptidic as well as nonpeptidic) cationic domains and lipophilic domains in terms of cellular delivery potency, mechanism(s) of cellular entry and toxicity in various cell types. Furthermore, the behavior of such conjugates for increasing the in ViVo bioavailability of potential drugs could be very interesting. Finally, there is little doubt that the delivery properties of CatLip domains may be optimized far beyond the limited examples studied so far.

ACKNOWLEDGMENT This work was supported by the Lundbeck Foundation, The Danish Cancer Foundation and the European Commission (6th FP EMIL and SNIPER contracts LSHC-CT-2004-503569 & LSHB-CT-2004-005204). The authors would like to acknowledge Dr. Anders E. Pedersen for help with flow cytometry, Dr. Darya Kiryushko for help with confocal laser microscopy, and Ms. Jolanta Ludvigsen and MSc. Nadia Bendifallah for synthesizing the PNA conjugates. Supporting Information Available: Additional information as described in the text. This material is available free of charge via the Internet at http://pubs.acs.org. Note Added after ASAP Publication: This paper was modified on August 6, 2008, to include the following additional information. The authors note that a paper reporting analogous delivery strategy has been published. See: Hu, J., and Corey, D. R. (2007) Inhibiting Gene Expression with Peptide Nucleic Acid (PNA)-Peptide Conjugates That Target Chromosomal DNA. Biochemistry 46, 7581-7589.

LITERATURE CITED (1) Dass, C. R. (2004) Lipoplex-mediated delivery of nucleic acids: factors affecting in vivo transfection. J. Mol. Med. 82, 579–91. (2) Kumar, V. V., Singh, R. S., and Chaudhuri, A. (2003) Cationic transfection lipids in gene therapy: successes, set-backs, challenges and promises. Curr. Med. Chem. 10, 1297–306. (3) Liu, D., Ren, T., and Gao, X. (2003) Cationic transfection lipids. Curr. Med. Chem. 10, 1307–15. (4) Nicolazzi, C., Garinot, M., Mignet, N., Scherman, D., and Bessodes, M. (2003) Cationic lipids for transfection. Curr. Med. Chem. 10, 1263–77. (5) Nielsen, P. E., Egholm, M., Berg, R. H., and Buchardt, O. (1991) Sequence-selective recognition of DNA by strand displacement with a thymine-substituted polyamide. Science 254, 1497–500. (6) Summerton, J., and Weller, D. (1997) Morpholino antisense oligomers: design, preparation, and properties. Antisense Nucleic Acid Drug DeV. 7, 187–95. (7) Hamilton, S. E., Simmons, C. G., Kathiriya, I. S., and Corey, D. R. (1999) Cellular delivery of peptide nucleic acids and inhibition of human telomerase. Chem. Biol. 6, 343–51.

Bioconjugate Chem., Vol. 19, No. 8, 2008 1533 (8) Morcos, P. A. (2001) Achieving efficient delivery of morpholino oligos in cultured cells. Genesis 30, 94–102. (9) Rasmussen, F. W., Bendifallah, N., Zachar, V., Shiraishi, T., Fink, T., Ebbesen, P., Nielsen, P. E., and Koppelhus, U. (2006) Evaluation of transfection protocols for unmodified and modified peptide nucleic acid (PNA) oligomers. Oligonucleotides 16, 43– 57. (10) Bendifallah, N., Kristensen, E., Dahl, O., Koppelhus, U., and Nielsen, P. E. (2003) Synthesis and properties of ester-linked peptide nucleic acid prodrug conjugates. Bioconjugate Chem. 14, 588–92. (11) Shiraishi, T., Bendifallah, N., and Nielsen, P. E. (2006) Cellular delivery of polyheteroaromate-peptide nucleic acid conjugates mediated by cationic lipids. Bioconjugate Chem. 17, 189–94. (12) Silhol, M., Tyagi, M., Giacca, M., Lebleu, B., and Vives, E. (2002) Different mechanisms for cellular internalization of the HIV-1 Tat-derived cell penetrating peptide and recombinant proteins fused to Tat. Eur. J. Biochem. 269, 494–501. (13) Vives, E., Brodin, P., and Lebleu, B. (1997) A truncated HIV-1 Tat protein basic domain rapidly translocates through the plasma membrane and accumulates in the cell nucleus. J. Biol. Chem. 272, 16010–7. (14) Derossi, D., Joliot, A. H., Chassaing, G., and Prochiantz, A. (1994) The third helix of the Antennapedia homeodomain translocates through biological membranes. J. Biol. Chem. 269, 10444–50. (15) Mae, M., and Langel, U. (2006) Cell-penetrating peptides as vectors for peptide, protein and oligonucleotide delivery. Curr. Opin. Pharmacol. 6, 509–14. (16) Moulton, H. M., Nelson, M. H., Hatlevig, S. A., Reddy, M. T., and Iversen, P. L. (2004) Cellular uptake of antisense morpholino oligomers conjugated to arginine-rich peptides. Bioconjugate Chem. 15, 290–9. (17) Bendifallah, N., Rasmussen, F. W., Zachar, V., Ebbesen, P., Nielsen, P. E., and Koppelhus, U. (2006) Evaluation of cellpenetrating peptides (CPPs) as vehicles for intracellular delivery of antisense peptide nucleic acid (PNA). Bioconjugate Chem. 17, 750–8. (18) Wender, P. A., Mitchell, D. J., Pattabiraman, K., Pelkey, E. T., Steinman, L., and Rothbard, J. B. (2000) The design, synthesis, and evaluation of molecules that enable or enhance cellular uptake: peptoid molecular transporters. Proc. Natl. Acad. Sci. U.S.A. 97, 13003–8. (19) Koppelhus, U., Awasthi, S. K., Zachar, V., Holst, H. U., Ebbesen, P., and Nielsen, P. E. (2002) Cell-dependent differential cellular uptake of PNA, peptides, and PNA-peptide conjugates. Antisense Nucleic Acid Drug DeV. 12, 51–63. (20) Drin, G., Cottin, S., Blanc, E., Rees, A. R., and Temsamani, J. (2003) Studies on the internalization mechanism of cationic cell-penetrating peptides. J. Biol. Chem. 278, 31192–201. (21) Richard, J. P., Melikov, K., Vives, E., Ramos, C., Verbeure, B., Gait, M. J., Chernomordik, L. V., and Lebleu, B. (2003) Cellpenetrating peptides. A reevaluation of the mechanism of cellular uptake. J. Biol. Chem. 278, 585–90. (22) Fotin-Mleczek, M., Fischer, R., and Brock, R. (2005) Endocytosis and cationic cell-penetrating peptides-a merger of concepts and methods. Curr. Pharm. Des. 11, 3613–28. (23) Nakase, I., Niwa, M., Takeuchi, T., Sonomura, K., Kawabata, N., Koike, Y., Takehashi, M., Tanaka, S., Ueda, K., Simpson, J. C., Jones, A. T., Sugiura, Y., and Futaki, S. (2004) Cellular uptake of arginine-rich peptides: roles for macropinocytosis and Actin rearrangement. Mol. Ther. 10, 1011–22. (24) El-Andaloussi, S., Johansson, H. J., Lundberg, P., and Langel, U. (2006) Induction of splice correction by cell-penetrating peptide nucleic acids. J Gene Med. (25) Wadia, J. S., Stan, R. V., and Dowdy, S. F. (2004) Transducible TAT-HA fusogenic peptide enhances escape of TATfusion proteins after lipid raft macropinocytosis. Nat. Med. 10, 310–5.

1534 Bioconjugate Chem., Vol. 19, No. 8, 2008 (26) Shiraishi, T., Pankratova, S., and Nielsen, P. E. (2005) Calcium ions effectively enhance the effect of antisense peptide nucleic acids conjugated to cationic tat and oligoarginine peptides. Chem. Biol. 12, 923–9. (27) Turner, J. J., Ivanova, G. D., Verbeure, B., Williams, D., Arzumanov, A. A., Abes, S., Lebleu, B., and Gait, M. J. (2005) Cell-penetrating peptide conjugates of peptide nucleic acids (PNA) as inhibitors of HIV-1 Tat-dependent trans-activation in cells. Nucleic Acids Res. 33, 6837–49. (28) Boe, S., Longva, A. S., and Hovig, E. (2007) Photochemically induced gene silencing using small interfering RNA molecules in combination with lipid carriers. Oligonucleotides 17, 166– 73. (29) Shiraishi, T., and Nielsen, P. E. (2006) Photochemically enhanced cellular delivery of cell penetrating peptide-PNA conjugates. FEBS Lett. 580, 1451–6. (30) Abes, S., Turner, J. J., Ivanova, G. D., Owen, D., Williams, D., Arzumanov, A., Clair, P., Gait, M. J., and Lebleu, B. (2007) Efficient splicing correction by PNA conjugation to an R6Penetratin delivery peptide. Nucleic Acids Res. 35, 4495–502. (31) Azzam, T., and Domb, A. J. (2004) Current developments in gene transfection agents. Curr. Drug DeliVery 1, 165–93.

Koppelhus et al. (32) Futaki, S., Ohashi, W., Suzuki, T., Niwa, M., Tanaka, S., Ueda, K., Harashima, H., and Sugiura, Y. (2001) Stearylated argininerich peptides: a new class of transfection systems. Bioconjugate Chem. 12, 1005–11. (33) Kang, S. H., Cho, M. J., and Kole, R. (1998) Up-regulation of luciferase gene expression with antisense oligonucleotides: implications and applications in functional assay development. Biochemistry 37, 6235–9. (34) Kang, S. H., Zirbes, E. L., and Kole, R. (1999) Delivery of antisense oligonucleotides and plasmid DNA with various carrier agents. Antisense Nucleic Acid Drug DeV. 9, 497–505. (35) Christensen, L., Fitzpatrick, R., Gildea, B., Petersen, K. H., Hansen, H. F., Koch, T., Egholm, M., Buchardt, O., Nielsen, P. E., Coull, J., and Berg, R. H. (1995) Solid-phase synthesis of peptide nucleic acids. J. Pept. Sci. 1, 175–83. (36) Shiraishi, T., and Nielsen, P. E. (2006) Enhanced delivery of cell-penetrating peptide-peptide nucleic acid conjugates by endosomal disruption. Nat. Protocols 1, 633–36. BC800068H