WO2008063113A1 - Cell -penetrating peptides and constructs containing them consisting 15-25 amino acids of tumor supressor protein p14arf or p19arf - Google Patents

Abstract

A cell-penetrating peptide or a construct comprising such a peptide of 15 to 25 amino acids comprising at least amino acids 1 - 14 of a mature mammalian tumor suppressor protein ARF or a scrambled or partially inverted sequence thereof, such as the human p14ARF protein SEQ ID NO:2, and the partially inverted sequence thereof SEQ ID NO: 1 optionally linked to one or more of a cargo, such as an anti-cancer drug, a detection marker, a cell-homing peptide, such as a tumor-homing peptide, a receptor ligand, a linker, and a peptide spacer comprising a selective protease cleavage site coupled to an inactivating peptide, is disclosed. The cell-penetrating peptides of the invention derived from ARF can either be biologically inert or have biological activity in themselves. Further, a method of delivering a cargo into a target cell in vivo or in vitro, is described.

Description

Cell-penetrating peptides and constructs containing them consisting 15-25 amino acids of tumor supressor protein pl4ARF or pl9ARF.

The present invention relates to new cell-penetrating peptides and constructs containing them. More precisely, the invention relates to a cell-penetrating peptide of 15 to 25 amino acids comprising at least amino acids 1 - 14 of a mammalian tumor suppressor protein ARF or a partially inverted sequence thereof, and a construct consisting of such a cell-penetrating peptide, linked to one or more members selected from a cargo, a detection marker, a cell-homing peptide, a receptor ligand, a linker, and a peptide spacer comprising a selective protease cleavage site coupled to an inactivating peptide. Further, the invention relates to a method of delivering a cargo into a target cell in vivo or in vitro.

Background of the invention

In the last ten years, a new class of peptides has gained increasing attention. These peptides are usually less than 30 amino acids (aa) in length with a cationic and/or amphipathic nature, and are generally referred to as cell-penetrating peptides (CPPs). These peptides have the ability to rapidly translocate into most mammalian cells carrying various cargo molecules such as oligonucleotides (ONs), peptides, proteins, plasmids, liposomes, and nanoparticles both in vitro and in vivo [1]. Moreover, these peptides have been successfully utilized in delivery of antisense peptide nucleic acids (PNAs) in numerous studies [2]. Thus, CPPs have opened a new avenue in ON-delivery and therapy allowing, otherwise impermeable, therapeutic agents to enter cells and induce biological responses. Still it is unclear what mechanism is responsible for uptake of CPPs and their cargos. It was generally believed that those peptides were internalized via a rapid receptor and energy independent pathway other than endocytosis. However, an emerging number of studies are now emphasizing the role of endocytosis in translocation of CPPs [3-5], and in particular macropinocytosis. Although several CPPs have been developed since the initial discovery of penetratin in 1994 and Tat peptide in 1997, it is still of great importance to find new peptides that are efficient and less toxic at higher concentrations.

Description of the invention

The present invention provides new cell-penetrating peptides that are efficient and less toxic at higher concentrations compared to prior art cell-penetrating peptides.

The invention also describes new cell-penetrating peptides with biological activity in themselves, decreasing proliferation and inducing apoptosis. In particular, the invention provides a cell-penetrating peptide or construct consisting of a cell-penetrating peptide of 15 to 25 amino acids comprising at least amino acids 1 - 14 of a mature mammalian tumor suppressor protein ARF or a scrambled or partially inverted sequence thereof, optionally linked to one or more members of the group consisting of a detection marker, e.g. a fluorescent marker, a cell-homing peptide, a receptor ligand, a linker, and a peptide spacer comprising a selective protease cleavage site coupled to an inactivating peptide. A scrambled or partially inverted sequence of ARF is in the present specification and claims intended to define a sequence wherein the same amino acids in the ARF sequence are included but one or several amino acids are in different positions so that a part of the sequence is inverted or the whole sequence is scrambled

In an embodiment of the invention the sequence of the mammalian tumor suppressor protein ARF is a sequence derived from the human p14ARF protein or a partially inverted sequence thereof, such as the human p14ARF protein with the sequence SEQ ID NO:2, or the partially inverted sequence thereof with the sequence SEQ ID NO: 1.

The amino acids 1-25 of the mature human p14ARF protein are shown in SEQ ID NO: 20. In another embodiment of the invention the sequence of the mammalian tumor suppressor protein ARF is a sequence of a mouse p19ARF protein or a partially inverted sequence thereof, such as the mouse p19ARF protein with the sequence SEQ ID NO: 17, or the partially inverted thereof with the sequence SEQ ID NO: 18.

The amino acids 1-25 of the mature mouse p19ARF protein are shown in SEQ ID NO: 21.

In yet another embodiment of the invention the cell-homing peptide is a tumor- homing peptide.

In still another embodiment of the invention the selective protease cleavage site is a cathepsin cleavage site. In a further embodiment of the invention the cargo is selected from the group consisting of oligonucleotides, such as morpholinos, peptide nucleic acids, locked nucleic acids, small interfering RNAs, antisense oligonucleotides, and plasmid DNAs, peptide ligands, anticancer drugs, such as one selected from an alkylating agent, an antimetabolite and a cytotoxic antibiotic, cytotoxic peptides, bioactive peptides, diagnostic agents and proteins.

Examples of the alkylating agent include 4-[4-Bis(2-chloroethyl)amino)- phenyl]butyric acid (chlorambucil) and 3-[4-(Bis(2-chloroethyl)amino)phenyl]-L-alanine (Melphalan); examples of the antimetabolite include Λ/-[4-(Λ/-(2,4-Diamino-6-pteridinyl- methyl)methylamino)-benzoyl]-L-glutamic acid (Methotrexate); and examples of the cytotoxic antibiotic include (8S,10S)-10-[(3-Amino-2,3,6-trideoxy-α-L-/yxo-hexopyranosyl)oxy]-8- glycoloyl-7,8,9, 10-tetrahydro-6,8, 1 1 -trihydroxy-1 -methoxy-5, 12-naphthacenedione (Doxorubicin). In a further embodiment of the invention, the construct comprises a peptide spacer comprising a selective protease cleavage site coupled to an inactivating peptide and the construct has a sequence selected from the group consisting of SEQ ID NOs: 3 - 6.

In still a further embodiment of the invention the construct comprises a peptide ligand and has a sequence selected from SEQ ID NO: 7 and SEQ ID NO: 8.

The invention provides also a method of delivering a cargo into a cell comprising contacting a construct according to the invention with a target cell in vivo or in vitro.

Thus, the invention provides CPPs, totally or partially derived from mammalian ARF-proteins. The native ARF sequence, M917 induces apoptosis as seen by increased annexin binding to MCF-7 and MDA MB 231 cells. The control peptides in this case, M872 which is a scramble version and M918 where amino acids 3-8 are inverted do not display any increased annexin binding and hence no induction of apoptosis. The invention is illustrated by description of the presently preferred CPP, M918, which efficiently translocates cells per se or as a vector for e.g. RNA, DNA, PNA, or streptavidin in a non-toxic fashion. Strikingly, when comparing the cellular uptake with established CPPs, such as penetratin (SEQ ID NO: 9) and TP10 (SEQ ID NO: 10), the cellular internalization of M918 is several-fold higher in different cell-lines and less cytotoxic (Fig 1 ): Since high cellular uptake of cell-penetrating peptides and cargos does not necessarily correlate with a high biological response, mostly due to material being retained in endosomes, a positive read-out assay was used to assess the delivery efficiency of M918. This assay is based on a splicing reporter system where a plasmid carrying the luciferase coding sequence is interrupted by an insertion of intron 2 from the thalasemic β-globin pre-mRNA carrying a cryptic splice site [6]. Unless the aberrant splice site is masked by antisense PNA, the pre-mRNA of luciferase will be improperly processed and no luciferase produced. By using HeLa pLuc 705 cells, stably transfected with this plasmid, various vectors can be evaluated by measuring the luciferase activity in cells. In contrary to classical antisense experiments, this assay generates a positive read-out rather than a negative read-out that may be more influenced by toxic side-effects on cells. Additionally, since splicing occurs in the nucleolus, the splice correction assay provides localization information about the delivered ONs inside the cell. Data from the splice correction assay clearly demonstrate that M918 induces splice correction in a dose- dependent manner (Fig 9). Also, this peptide increases the quantitative uptake of ONs, and even larger proteins such as streptavidin, in a concentration-dependent manner (Fig 6). Interestingly, addition of a lysosomotrophic agent, chloroquine, enhanced the splice correction of M918-PNA conjugate several-fold, emphasizing that conjugates are internalized via endocytosis (Fig 9c). Involvement of endocytosis in the uptake of M918 was further highlighted when performing experiments at 4°C and with endocytosis inhibitors where a decreased internalization was observed (Fig 7, 9b). This was further confirmed with co- localization studies, showing that the M918 peptide co-localizes with the endocytosis markers; dextran and transferrin (Fig 8). The presented data indicate that the CPP M918 has excellent delivery properties and may be applied at high micromolar concentrations (up to 50μM concentrations) without affecting cell-viability (Fig 1c, d).

The invention is further exemplified by the peptide M917 that is an efficient CPP, comparable to the established peptide TP10 (Fig 10a). In addition, M917 conjugated to a splice correcting PNA as described above restores correct splicing and luciferase expression, hence verifying that M917 has the ability to transport cargoes into the nucleus (Fig 10b). M917 enters and localizes preferentially in vesicles inside cells and partially co- localizes with the marker for clathrin mediated endocytosis, transferrin (Fig 11). The involvement of endocytosis in the uptake of M917 is further exemplified by the decrease in uptake seen with endocytosis inhibitors (Fig 12a) as well as an increase in splice correction after co-incubation of splice correcting M917-PNA conjugate with chloroquine (Fig 10b). The intact M917 peptide was also found intact inside cells for up to 3 hr after extracellular application (Fig 12b). The CPP M917 displays its biological activity by dose dependently decreasing proliferation in MCF-7 and MDA MB 231 cells after extracellular treatments (Fig 13). Importantly, M917 displays low membrane disturbance compared to the control scrambled sequence M872, thereby verifying the specificity of M917 (Fig 14). Since the parent protein to the M917 peptide, p14ARF is a tumor suppressor protein with the capability to induce apoptosis, it is highly important to verify the specificity of M917 in induction of apoptosis. Induction of apoptosis by M917 was established after extracellular peptide treatment of MCF-7 and MDA MB 231 cells after staining for Hoechst and screening for apoptotic nuclei (Fig 15). Additionally, to further assign the mechanism of M917 to apoptosis, flow cytometry of M917 treated cells was investigated after staining for Annexin V and Pl and found to induce apoptosis (Fig 16). Finally, the invention thus embodies that the M917 peptide combines cell-penetrating properties with the functions of the full-length protein p14ARF.

The M918 peptide, as well as other CPPs of the invention, is useful in vivo for its potential to transport e.g. oligonucleotides, other peptides, and anti-cancer agents.

Selective targeting of tumors is achieved using two different strategies. First, the pro-drug strategy where the CPP is prolonged to include a protease cleavable peptide spacer and an inactivating peptide, allowing the CPP only to be activated upon cleavage with proteases in tissues that over-express these proteases, i.e. tumors. The selective cleavage sites will be substrates for the proteases cathepsins. The second strategy is based on selective targeting through receptor ligands or homing peptides conjugated to M918 with or without cargo. This will allow targeting of breast cancers and brain tumors in vivo. Finally, by adding cytotoxic drugs to both of these delivery systems selective tumor killing will be achieved.

The invention will now be further illustrated by the following description of embodiments, including short description of the drawings, materials and methods, experimental results including figures and figure legends as well as sequence listing, but it should be understood that the scope of the claims of the invention is not limited to any specifically mentioned embodiments or details.

Short description of the drawings Fig. 1 shows diagrams regarding internalization of peptides M918, Pen and

TP10 into HeLa cells (a), internalization of M918 into Hifko, CHO, VEGF+ and MCF-7 cells (b), membrane leakage measured by LDH release (c) and long term toxicity recorded by WST-1 assay (d).

Fig. 2 shows a diagram where the annexin binding to MCF-7 and MDA MB 231 cells of control, M917, M918, M872 and staurosporine (STS) is depicted.

Fig. 3 shows confocal microscopy images of Hifko, MCF-7 and HeLa cells after treatment with fluoresceinyl labeled M918.

Fig. 4 shows a diagram where intracellular amounts in pmol/mg of intact and degraded M918 peptide in MCF-7 cells are plotted against time in minutes. Fig. 5 shows diagrams where the relative fluorescence quenching mediated by

Pen, TP10 and M918 are plotted against the molar ratio (peptide/siRNA) (a), and the uptake mediated by complexes with different molar ratios between Pen (b), TP10 (c) and M918 (d) and fluoresceinyl labeled dsDNA or siRNA, respectively.

Fig. 6 shows a diagram of protein transduction in pmol/mg into HeLa cells by co-incubation with CPPs at different concentrations in μM and protein transduction mediated by direct coupling of CPPs to the protein.

Fig. 7 shows diagrams illustrating the effects of endocytosis inhibitors and glycosaminoglycans on quantitative uptake of M918.

Fig. 8 shows confocal microscopy images of the endocytosis markers transferrin and dextran co-localized with M918 into MCF-7 cells.

Fig. 9 shows diagrams of luciferase expression in HeLa pLuc 705 cells after treatment with different M918-PNA conjugates and effects of agents promoting as well as inhibiting endocytosis.

Fig. 10 shows diagrams regarding (a), internalization of peptides M917, M872 and TP10 into MCF-7 cells and (b), luciferase expression in HeLa pLuc 705 cells after treatment with M917-PNA conjugates and effects of a lysosomotrophic agent. Fig. 11 shows confocal microscopy images of the endocytosis marker transferrin co-localized with M917 in MCF-7 cells.

Fig. 12 shows diagrams illustrating involvement of endocytosis inhibitors in uptake of ARF derived peptides and stability of M917 peptide inside cells. Fig. 13 shows decrease in proliferation after treatment with M917 in MCF-7 and

MDA MB 231 cells.

Fig. 14 shows that M917 does not affect membrane integrity, measured by LDH and deoxyglucose leakage.

Fig. 15 shows that M917 induces nuclear morphology changes, indicative of apoptosis as visualized by Hoechst staining of MCF-7 cells.

Fig. 16 shows flow cytometry detection of apoptosis by Annexin and Pl binding induced by M917 while M918 induced cells show no signs of apoptosis.

Material and Methods

Synthesis of Peptides, PNA, conjugates and hybridization of ONs

The peptides penetratin (SEQ ID NO: 9), TP10 (SEQ ID NO: 10), M918 (SEQ ID NO: 1 ), M917 (SEQ ID NO: 2), M872 (SEQ ID NO: 19), 705 PNA (SEQ ID NO: 11 ), and inverted 705 PNA (SEQ ID NO: 12) were synthesized on Applied Biosystems stepwise synthesizer model 433A. Both peptides and PNAs were assembled by f-Boc chemistry using a 4-methylbenzhydrylamine-polysterene resin (MBHA) to generate amidated C-terminus.

Amino acids were purchased from Neosystem, France and coupled as hydro xybenzotriazole (HOBt) esters while PNA was coupled with 2-(7-Aza-1 H-benzotriazole-1-yl)-1 , 1 ,3,3- tetramethyluronium hexafluorophosphate (HATU). PNA used for uptake studies were labelled at the ε-amino group of the C-terminal lysine with carboxyfluorescein using 5 eq. 5,6- carboxyflorescein, 5 eq, DIC, 5 eq. HoBt and 20 eq. DIEA dissolved in DMSO:DMF 1 :1 over night [7].

After cleavage of peptides and PNAs from the resin using HF, synthesis products were purified by reverse phase HPLC Iomega C₁₈ column and analyzed using Perkin Elmer prOTOF™ 2000 MALDI O-TOF Mass Spectrometer. Conjugation of peptides with 3-nitro-2-pyridinesulphenyl (Npys) to cysteine coupled PNA via a disulphide bridge were performed over night in 20% acetonitril/water containing 0.1% TFA and separated by reverse-phase HPLC. Masses of peptides, PNAs and conjugates correlated well with theoretical values. The sequences of the peptides and PNAs are presented in Table 1. RNA sequences were purchased from Dharmacom (USA) and the sequences were taken from Ui-Tei et al [3]. Identical DNA ONs were purchased from Cybergene (Sweden). Sequences are presented in table 1. Single stranded ONs, fluorescein labeled in the 3^'end of the antisense strand were hybridized to the sense strand in MQ water containing 0,9% NaCI for 1 h at 37⁰C to achieve double stranded RNA or DNA with two nucleotides overhang at the 3^'and 5^'end.

Cell culture

HeLa pLuc 705 cells, kindly provided by R. KoIe and B. Leblue, were grown in Dulbecco's Modified Eagle's Media (DMEM) with glutamax supplemented with 0.1 mM nonessential amino acids, 1.0 mM sodium pyruvate, 10 % FBS, 100 U/ml penicillin, 100 mg/ml streptomycin and 200μg/ml hygromycin. MCF-7 cells and MDA-MB-321 cells were grown in RPMI media with glutamax supplemented with 0.1 mM non-essential amino acids, 1.0 mM sodium pyruvate, 10 % FBS, 100 U/ml penicillin, and 100 mg/ml streptomycin. Astrocytoma cell lines (Hifko, VEGFwt, and VEGF+ cells) were grown in minimal essential media (MEM) with glutamax supplemented with 1mM sodium pyruvate, 1OnM Hepes, 5% FBS, 100 U/ml penicillin, and 100 mg/ml streptomycin. Cells were grown at 37 ⁰C in 5% CO₂ atmosphere. All media and chemicals were purchased from Invitrogen (Sweden).

Quantitative Uptake

Cells were treated with 5 μM fluoresceinyl labeled peptides, peptides in various concentrations co-incubated for 30 min with fluoresceinyl labeled RNA, DNA, and streptavidin in 0.9% NaCI solution, or with peptide-PNA conjugates at different concentrations. After treatment, cells were washed twice in Hepes Krebs Ringer (HKR) before trypsination. Cells were centrifuged at 1000 g for 5 min at 4 ⁰C and cell pellets were lysed with 300 μl 0.1 % Triton in HKR for 30 min after which 250μl lysate was transferred to a black 96-well plate. Fluorescence was measured at 494/518 nm on a Spectra Max Gemeni XS fluorometer (Molecular Devices, USA) and recalculated to amount of internalized compound by using the linearity of fluorescein and normalizing to amount of protein (Lowry BioRad, USA). Fluorescein labeled peptides, DNA, RNA, PNA and streptavidin, displayed linear correlation between concentration and fluorescence.

In endocytosis inhibition experiments, the same protocol was used but cells were pretreated for 30 min with 75 μM chloroquine, 0.2 M sucrose or cells and solutions were pre-maintained at 4 ⁰C for 30 min prior exposure of peptides or conjugates for 1 h at 4 ⁰C. All inhibitors were purchased from Sigma-Aldrich, Sweden.

Co-localization study by confocal microscopy

HeLa pLuc 705 cells or Hifko cells were seeded in NUNC 8 chambers to reach a 50 % confluence one day post seeding. Cells were washed twice with HKR, after which fluoresceinyl labelled peptides and rhodamine B labelled dextran were added simultaneously at a final concentration of 1 μM and 1 mg/ml, respectively. After 30 min incubation in serum free DMEM at 37⁰C, cells were washed three times with HKR and examined using UltraView ERS confocal live cell imager (PerkinElmer Ltd, Upplands Va^'sby, Sweden) connected to an Axiovert 200 (Zeiss, Gόttingen, Germany).

Determination of peptide degradation by HPLC

MCF-7 and MDA MB 231 cells were seeded in 24 well plates at a density of 100.000 cells per well two days before peptide treatment. Peptides were added to cells in HKR medium for 30, 60 or 180 min and washed two times with ice-cold PBS, incubated with 200 μl of ice-cold PBS added with diazotized 2-nitroaniline as described previously by Oehlke et al [9] to modify any surface bound peptides. Then cells were lysed with 0.2 ml 0.1% Triton X-100 containing 10 mmol/l trifluoroacetic acid for 2 h at 0 ⁰C. The lysates were used for HPLC analysis on Gyncotek-HPLC-gradient system (Dionex, USA) consisting of a 5 μm (150 mm * 4.6 mm) C18 column, a precolumn containing the same adsorbent and a fluorescence detector and protein levels were determined by Detergent compatible protein assay (BioRad). Quantification of internalized peptide were performed by fluorescence measurement at 524 nm after excitation at 440 nm using calibration values obtained with 10 pmol of the parent peptide analyzed under identical conditions. The areas under the peptide peaks were all proportional to the molar amount of injected peptide. Degraded peptide was identified by shift in retention time from the parent intact peptide.

Ethidium bromide exclusion assay

RNA condensation was measured by quenching of ethidium bromide (EtBr) fluorescence essentially as described in [10]. Briefly, quadruplicates of 0.5 μg of siRNA were complexed with increasing amounts of penetratin or M918 in HKR to a final volume of 25 μl. After 30 min incubation time, 160 μl HKR was added to each tube and transferred to a black 96-well plate after which 15 μl EtBr solution (0.1 mg/ml) was added to each well. The fluorescence was measured after 10 min on a Spectra Max Gemini XS fluorometer (Molecular Devices, USA) at λ_ex = 518 nm and λ_em = 605 nm. Results are given as relative fluorescence and the value of 100% is attributed to the fluorescence of RNA with ethidium bromide (rel. F =

solution)-

Splice correction assay

60 000 HeLa pLuc 705 cells were seeded 48 h prior experiments in 24-well plates in all experiments. Cells were treated for 4 h in serum free media followed by replacement to serum medium for additionally 16 h. Thereafter, the cells were washed twice with HKR buffer and lysed using 100μl 0.1 % triton X100 in HKR for 15 min at room temperature. Luciferase activity was measured on Flexstation Il (Molecular devices, USA) using Promega luciferase assay system. RLU values were normalized to protein content and results are displayed as RLU/mg.

In endocytosis inhibition experiments, cells and solutions were pre-maintained at 4 ⁰C for 30 min prior exposure to conjugates for 1 h at 4 ⁰C. Thereafter, cells were transferred back to 37 ⁰C for additionally 16 h.

In experiments with the agent promoting endosomal escape, chloroquine, conjugates were co-incubated for 4 h with 75 μM chloroquine in serum free DMEM and subsequently the cells were grown for 16 h in complete DMEM.

LDH leakage

Membrane integrity was measured using the Promega Cytox-ONE™ assay. In brief, 10⁴ cells were seeded in 96-well plates two days before treatment with peptides in serum free or complete DMEM. After 30 min, media was transferred to a black fluorescence plate and incubated for 10 min with CytoTox-ONE™ reagent followed by stop solution. Fluorescence was measured at 560/590 nm. Untreated cells were defined as zero and LDH released by lysating in 0.18% triton in HKR as 100% leakage.

Deoxyglucose leakage

10⁵ MCF-7 cells were plated in 12-well plates two days before exposure to 1 , 10, 25 and 50 μM peptides in 500 μl serum free RPMI media. Cells were treated as described in [38]. Untreated cells were defined as 0% and total radioactivity in each well to 100%. Samples were diluted in 5 ml Emulsifier Safe scintillation liquid (Perkin Elmer, USA) and measured on a β-counter (2500TR Packard, Australia).

Nuclear morphology

To assess nuclear morphology in MCF-7 and MDA MB 231 , cells were seeded on coverslips one day before treatment with 10 μM peptides for 2 h in media without addition of serum, followed by addition of full growth media. After 24 h incubation, cells were washed, fixed with 4 % paraformaldehyde and stained with 5 μg/ml Hoechst. Fluorescent nuclei were screened for normal, fragmented and condensed morphology using fluorescence microscopy. 500 cells were counted for each sample and apoptotic nuclei were expressed as percentage of total cells. Detection of apoptosis with Annexin V and Pl

Cells, 60 000 seeded in 24-well plates two days before, were treated with 25 μM peptides or 0.1 μM STS for 4 h in media without serum followed by addition to full growth media with peptide or STS for an additional 14 h. The assay was performed according to the instructions of the manufacture (BD Pharmingen™, BD Biosciences, Sweden). Briefly, cells were washed twice with PBS, and resuspended in binding buffer. Annexin V - FITC and Pl were added to each sample, and the mixture was incubated at room temperature for 15 min. Cells were then subjected to fluorescence-activated cell sorter (FACS) analysis using a Becton-Dickenson FACScalibur flow cytometer and Cell Quest Pro software (Becton- Dickenson, Mountain View, CA, USA) to generate dot plots for calculation of apoptosis. Viable and apoptotic populations were identified from untreated and staurosporine treated cells, respectively.

Wst-1 assay

HeLa pLuc 705 cells were seeded onto 96-well plates, 10⁴ cells/well, two days before treatment. Cells were treated with conjugates in 100 μl serum or serum free DMEM for 24 h. Cells were then exposed to Wst-1 according to manufacturers protocol (Sigma,

Sweden). Absorbance (450-690nm) was measured on absorbance reader Digiscan

(Labvision, Sweden). Untreated cells are defined as 100% viable.

Results

M918 internalizes into several different cell-lines in a non-toxic fashion

M918 is a synthetic peptide partly derived from the tumor suppressor protein p14ARF, i.e. amino acids 1-22 with positions 3-8 inverted. This peptide was developed as a control peptide in a study where the parent peptide M917 was utilized to mimic the activity of the ARF protein. To our surprise, M918 showed excellent cell-penetrating properties, even higher than the parent peptide M917, although it did not induce apoptosis as the corresponding wild type peptide (Fig 2). Therefore, the peptide M918 was further characterized and evaluated regarding its delivery properties and potential to be used as an ON-delivery vector. First, the cellular uptake of M918 was assessed in a quantitative uptake assay.

The results clearly show that M918 is readily internalized into several cell-lines (Fig. 1b) and to a higher extent than the commonly used CPPs; penetratin and TP10 in HeIa cells (Fig. 1a) and CHO cells (data not shown). Furthermore, in order to exclude the possibility that the observed uptake was a result of plasma membrane disturbance, an LDH-leakage assay was performed. According to Fig. 1c, none of the tested peptide concentrations (5-25 μM) had any significant effect on membrane integrity in any tested cell-line. At a 50 μM concentration, a slight cytotoxic effect could be observed (data not shown). When comparing the data obtained here with recently published data from our group [11], measuring release of LDH after TP10- or penetratin treatment, it is apparent that M918 is less cytotoxic, and in particular compared to TP10. However, even though the peptide does not affect the membrane integrity, it might still exhibit other toxic side-effects on cells independent from membrane leakage. Analyzing long-term toxic effects of peptides is an important step towards future in vivo experiments; therefore a wst-1 assay was carried out to determine whether the peptide influences cellular proliferation. In accordance with the LDH-leakage data, no significant effect on proliferation after 72 h was observed with M918 in any cell-line using the same concentration range (Fig 1d). The high quantitative uptake observed for

M918 in different cell lines could be explained by peptide-aggregation in cellular membranes rather than internalized peptides. Therefore, the uptake was analyzed qualitatively using confocal microscopy. Clearly, neither in human breast cancer cells, MCF-7, nor in astrocytoma cells, Hifko, any clusters of peptides could be detected in the plasma membrane (Fig. 3). Most internalized material appears to reside in vesicles throughout the cell with low or negligible amounts in the cell nucleus (Fig. 3). It is, however, very important to point out that peptides may be located in the nucleus although at levels below the detection limit of the microscope.

The high observed uptake of M918 as compared to TP10 and penetratin may be a result of improved peptide stability (Fig. 4). According to Fig 4, at least half of the peptide population appears to be intact after 3h inside cells. This indicates superior peptide stability compared to earlier published data on penetratin stability (Caroline Palm BBA).

M918 readily form complexes with and mediates cellular uptake of siRNA and dsDNA As mentioned earlier, it is of great importance to find new, efficient delivery vectors for various ONs. First, the vehicle must promote complex formation with negatively charged ONs in order to condense and, more important, shield the negative charges to avoid repulsion with the negatively charged proteoglycans and lipids on the cellular membrane. An ethidium bromide exclusion assay was utilized to evaluate the capability of M918, penetratin, and TP10 to form complexes with siRNA (Fig. 5) and dsDNA (data not shown) at various molar ratios. At all tested molar ratios, M918 was proven better than both penetratin and TP10. Already at molar ratio 5 a 60% siRNA condensation was achieved with M918 while the other two peptides only confers 20% complex formation at the same concentration. These results are rather surprising when taking into consideration that M918 and penetratin share the same number (8) of positively charged amino acids within their peptide sequence. Also, unexpected, TP10 was equally competent as penetratin to form complexes with siRNA albeit its primary sequence only contains five positively charged amino acids. Hence it is apparent that complex formation not only depends on the cationic nature of peptides but also other factors as length and hydrophobicity. Similar results were obtained when using dsDNA (data not shown).

Another important property of a delivery vector is its ability to translocate cargo molecules, as ONs, over the plasma membrane to reach the interior of cells. We therefore evaluated the capability of M918, penetratin, and TP10 to transport dsDNA and siRNA inside cells using the same quantitative uptake assay described before. Overall, the tested peptides have higher potential to convey dsDNA than siRNA, with identical oligonucleotide sequence, over the plasma membrane (Fig. 5b-d). All peptides increase the cellular uptake of both ONs in a dose-dependent manner, with the exception for M918 that reduce internalization of dsDNA at 5 μM (molar ratio 50) compared to 2μM peptide. In accordance with uptake data on free fluoresceinyl labeled peptides, M918 present itself as the most competent delivery vector, at least for dsDNA. Interestingly, the uptake data obtained here correlates well with the ability of peptides to form complexes with ONs.

M918 promotes protein transduction

To further assess the potential of M918 in delivery of bioactive agents, proteins were utilized as cargos for penetratin, TP10, and M918. The 66 kDa protein streptavidin (SA) and 60 kDa protein avidin were chosen as model proteins of several reasons. First, both are cheap and can be ordered FITC-labeled. Second, they are average sized proteins but having significantly different isoelectric points (Ip), SA 5.5 and avidin 10.5. Third, they have the ability to bind biotin very strongly generating a covalent-like bond.

In the first set of experiments, peptides were simply mixed with SA(Fig.4a) or avidin (Fig 4c), as in the experiments with ONs, in order to evaluate whether CPPs translocate proteins as non-covalent complexes. According to Fig. 6a, all three peptides confer cellular uptake of FITC-labeled SA at 5 μM peptide concentrations (25x molar excess), M918 being the most efficient vector. When using 1 μM peptides, however, no improvement in delivery of SA was observed with any peptide (data not shown). In contrast, when using 1μM biotinylated peptides that form four covalent-like bonds with SA, the uptake levels were increased significantly (Fig. 6b) as compared to co-incubation with TP10 and M918. Interestingly, penetratin promotes equally low uptakes independent of conjugation strategy (Fig. 6a, b). However, when assessing the quantitative uptake of avidin, negligible levels of internalized protein were observed with any peptide using the co-incubation strategy (Fig. 6c) while a significant increase in cellular uptake was seen when using 1 μM biotinylated peptides. As with SA, using biotinylated peptide, M918 present itself as the most potent delivery vector, closely followed by TP10. Overall, the cellular uptake of avidin (Fig. 6 d) is higher than streptavidin, most likely because of the cationic nature of avidin. The lack of uptake when co- incubating peptides with avidin may also be assigned to the cationic nature of avidin that prohibits electrostatic interactions with cationic CPPs

Internalization of peptide is mediated by endocytosis The high uptake observed for M918 as free peptide or in complex with ONs together with the data received from the confocal microscope showing punctual distribution of peptide inside cells encouraged us to elucidate the internalization mechanism of the peptide. Originally, it was believed that CPPs translocated cells in a receptor-and energy independent manner. However, recent studies suggest that the uptake for most, if not all, CPPs is an energy dependent process with initial binding of peptides to proteoglycans on cell surfaces and concomitant endocytosis of peptides. Two main experimental strategies have been used to investigate the role of endocytosis in uptake of CPPs; use of endocytosis inhibitors that block certain pathways followed by measurements of the cellular uptake of CPPs or co-localization studies utilizing labeled endocytosis markers as dextran or transferrin that are co-added with labeled peptides and analyzed by microscopy.

Here we show that by treating and incubating cells with M918 at 4⁰C, the cellular uptake is decreased significantly in HeLa cells (Fig. 7). Similar results were also obtained in CHO cells and with penetratin and TP10 in both cell lines (data not shown). These results suggest that the peptide is internalized via endocytosis since negligible levels of vesicles are formed at 4⁰C. Also, by co-adding 0.2M sucrose, which blocks receptor- mediated endocytosis, with M918 internalization of fluoresceinyl labeled peptide was decreased significantly which further emphasize endocytic uptake (Fig 7). To further elucidate what endosomal route peptides are utilizing, a co-localization assay was performed between rhodamine labeled dextran (a marker for fluid phase endocytosis) or transferrin (a marker for clathrin-mediated endocytosis) and fluoresceinyl labeled M918. As seen in Fig. 8 the peptide co-localizes with both markers in MCF-7 cells, suggesting that several different mechanisms are involved in the cellular uptake of the peptide. Similar results were also obtained in several other cell lines such as HeIa, CHO, and Hifko cells (data not shown). These data are consistent with recent findings in our group with TP10 [12], however somewhat divergent from results on penetratin showing higher propensity to co-localize with dextran [13].

Cell-penetrating antisense PNA induce splice correction dose-dependently

Although results from uptake experiments with CPPs provide valuable information regarding the efficacy by which peptide translocates cargo molecules, it does not necessarily correlate with the bioavailability of the cargo. Therefore, it is of outmost importance to study a biological response in order to draw accurate conclusions concerning the suitability of a peptide used as a delivery vehicle. Hence, we employed the splice- correction system developed by KoIe et al that generates a positive read-out in form of luciferase [6]. The major strength with this assay, except the fact that it generates production of a protein and is less susceptible to toxic side-effects, in addition, it gives valuable localization information of internalized material since splicing occurs in the nucleolus. By treating cells for one hour with M918 covalently conjugated via a disulphide bridge to fluoresceinyl labeled PNA (M918-PNA), and measuring luciferase expression 16 h later, a dose-dependent increase in uptake and splice-correction was observed (Fig. 9a). Interestingly, there is a linear correlation between uptake and splicing when measuring in the very same cell lysate. In comparison with recently published data from our group using penetratin and Tat peptide in splice correction experiments [14], M918 present itself as a superior vector for antisense PNA, especially compared to Tat. Neither does an inverted PNA (invPNA) sequence conjugated to M918 nor free PNA induce significant splicing at 10 μM concentrations (Fig. 9b). These results clearly indicate that the system is specific and that M918 is responsible for the cellular uptake of PNA, since naked PNA is unable to promote significant splicing of the luciferase pre-mRNA.

Lysosomotrophic agent improves splicing several-fold

After realizing, what possibly could be involvement of endocytosis in uptake of peptide, interpreted by data with endocytosis inhibitors and co-localization studies, we further wanted to corroborate these data by performing the splice correction assay at 4°C. Cells were pre-maintained at 4⁰C for 30 min prior addition of 5 μM M918-PNA and incubated for 1 h after which media was replaced and cells relocated to 37 ⁰C for additionally 16h. As seen in Fig. 9b, splicing is decreased significantly compared to cells treated at 37 ⁰C further emphasizing the uptake to be, at least partially, endocytic.

The major drawback with endocytosis of CPPs and other compounds is the concomitant entrapment of molecules inside endosomes. It is believed that most material remains caught in endosomes and finally degraded in lysosomes, which decrease the bioavailabilty of compounds that are not allowed to reach the cytosplasm (or the nucleoplasm). Several recent reports suggest that addition of lysosomotrophic agents such as bafilomycin and chloroquine promote release of endocytosed material ("endosomal escape"), hence allowing the internalized agent to exert its biological effects [15]. Based on this knowledge, a splice correction experiment was performed where M918-PNA was added in combination with 75 μM chloroquine to pLuc 705 HeIa cells and incubated as previously described. According to Fig. 9c, splicing is increased 8-fold after treatment with chloroquine as compared to using 2 μM M918-PNA alone. The effect of chloroquine is, however, not significant at 1 μM M918-PNA. Again, treatment with M918-invPNA and chloroquine do not confer any splice correction. These data confirm the hypothesis that CPPs are internalized via endocytosis and that treatment with chloroquine promotes endosomal escape, thereby increasing the bioavailability of delivered cargos.

Peptides from the N-terminal part of p14ARF, M917 and M872 translocate into cells

M917 is derived from the 22 first amino acids of the p14ARF protein. To investigate if M917 could translocate into cells, MCF-7 cells were exposed to 5 μM fluoresceinyl labeled peptides for 1 h (Fig. 10a). Uptake levels of M917 and the scrambled version M872 were comparable to TP10. M917 is a CPP with capability to deliver a biologically active cargo. M 197 was conjugated to a splice correcting peptide nucleic acid (PNA) (Table 1 ). HeLa pLuc 705 cells incubated with 5 μM M917-PNA conferred a five-fold increase in splice correction, while the control conjugate, TP10-PNA gave a ten-fold increase (Fig. 10b). Co-treatment of M197-PNA with the lysosomotrophic agent chloroquine that prohibits endosome maturation, dramatically increased splicing efficiency fifteen-fold, suggesting involvement of endocytosis in the uptake mechanism. None of the control conjugates carrying an inverted PNA sequence had any effect on splice correction.

ARF-derived peptides are mainly internalized into cells via endocytosis The uptake and the intracellular distribution of M917 and M872 were investigated by confocal microscopy. Both peptides showed a preferentially punctual distribution inside cells, suggesting internalization via endocytosis (Fig. 11 ). The involvement of endocytosis was further examined using endocytosis inhibitors, which all moderately decreased peptide uptake by approximately 20-40 % (Fig. 12a). Furthermore, partial co-localization with transferrin was observed with both peptides, demonstrating participation of clathrin-mediated endocytosis in uptake of ARF-derived peptides. Importantly, as an evaluation of the method, FITC-dextran and rhodamine- transferrin localized in distinct separate intracellular compartments, which exclude over- bleeding between the fluorophores (Fig. 11 ). These results collectively illustrate the importance of endocytosis in uptake of these peptides, albeit the exact pathway remains elusive.

Intact ARF-derived peptides are identified inside cells

The presence of intact M917 peptides inside cells were demonstrated by, HPLC analysis on cell lysates. Prior to cell lysis, cells were treated with the membrane- impermeable reagent diazotized 2-nitroaniline, which modifies primary amino groups. Treatment with diazotized 2-nitroaniline resulted in increased retention time for the modified surface bound peptides. The main product was identified as fluoresceinyl labeled peptide and a number of smaller peaks corresponded to degraded peptide. Roughly, half of the internalized M917 and M872 peptides were intact after 30, 60, and 180 min exposure (Fig. 12b, c).

M917 decreases proliferation in breast cancer cells

The cell-penetrating ARF peptide M917 decreased proliferation when added exogenously. MCF-7 and MDA MB 231 breast cancer cells were treated twice during 48 h. M917 dose dependently decreased proliferation in both cell lines, resulting in a 45 % decrease in proliferation at 25 μM concentration in both MCF-7 and MDA MB 231 cells (Fig. 13a, b). The M872 peptide had a significant effect on cell proliferation only at the highest concentration. In MDA MB 231 cells, a statistically significant reduction in proliferation could be observed already at 10 μM concentration using the M917 peptide (Fig. 13b). The anti-proliferative effects of M917 were improved by the lysosomotrophic agent chloroquine, where a significant decrease in proliferation was observed already at 5 μM M917, reducing the proliferation to the same extent as the non-selective kinase inhibitor staurosporine (STS)(Fig. 13c). Neither the peptide alone nor M872 had any effect on proliferation at 5 μM.

Some peptides interact strongly with cellular membranes, which might disrupt and introduce holes. To exclude membrane disturbance from proliferation data, deoxyglucose and lactate dehydrogenase (LDH) leakage assays were performed. The M917 peptide displayed low leakage up to 25 μM, but started to mediate leakage at 50 μM (Fig. 14a, c). M872 induced substantial leakage at 25 μM compared to M917 and provoked a small increase in leakage over time at 1 and 10 μM observed in the deoxyglucose leakage assay (Fig. 14b, d).

Induction of apoptosis by the cell-penetrating M917 peptide

To elucidate whether M917 peptide induces apoptosis or necrosis, cells were stained with Hoechst or Annexin V /propidium iodide (Pl). After treatment with 10 μM M917 and stained with Hoechst, MCF-7 cells displayed shrunken nuclei and to a minor extent apoptotic bodies while MDA MB 231 had a similar pattern but less pronounced (Fig. 15a, b). Co-treatment with chloroquine increased the percentage of apoptotic cells further supporting an endocytosis uptake mechanism (Fig. 15a). Control cells, treated with M872 and M918 did not display any change in nuclear morphology. STS mediated the characteristic condensed nuclei's and apoptotic bodies (data not shown).

The effect of M917 on apoptosis and necrosis was further determined by Annexin V and Pl staining after incubation with 25 μM peptide. Representative dot plots of Annexin V/PI staining are shown in Fig. 16a. Flow cytometric analysis indicate that M917 mainly induced apoptosis. Additionally, the viable population was slightly shifted towards annexin and Pl binding, suggesting that there were small apoptotic and necrotic effects on these cells. Importantly, the control peptide, M918 did not induce any significant Annexin V binding (Fig. 16a, b). Also, STS significantly induced Annexin V binding as earlier reported [20].

Table 1

Sequences of cell-penetrating peptides (CPPs) peptide nucleic acids (PNAs), DNA and RNA.

Pen and M918 were fluoresceinyl labelled on the N-terminus. ^* carboxyfluorescein were coupled to the ε-amino group of lysine⁷ in TP10. For building of conjugates, M918 were synthesized with Cys(Npys) on the N-terminal. Bold letters represent nucleobases. The four underlined bases indicate the sequence within PNA, which is inverted to generate the inverted PNA. ⁺ Attachment of carboxyfluorescein on ε-amino group of lysine on PNA for uptake experiments. Conjugates are illustrated as follows: M918(Cys)-s-s-CysPNA. n DNA and RNA were labelled at the 3^'end with fluorescein.

Name AbbreSequence SEQ ID NO viation

M918 M918 MVTVLFRRLRIRRACGPPRVRV-NH₂ SEQ ID NO: 1

M917 M917 MVRRFLVTLRIRRACGPPRVRV-NH₂ SEQ ID NO: 2

M872 M872 FVTRGCPRRLVARLIRVMVPRR-NH₂ SEQ ID NO: 19

Penetratin Pen RQIKIWFQNRRMKWKK-NH₂ SEQ ID NO: 9

Transportan 10 TP10 AGYLLGK*INLKALAALAKKIL-NH₂ SEQ ID NO: 10

M705 PNA PNA CysKKCCTCTTACCTCAGTTACAKK⁺-NH₂ SEQ ID NO: 11

M705 invPNA CysKKCCTCTTACACTCGTTACAKK⁺-NH₂ SEQ ID NO: 12 inverted PNA ssDNA ssDNA ACGCCAAAAACATAAAGAAAG SEQ ID NO: 13 asDNA asDNA TTCTTTATGTTTTTGGCGTCTn SEQ ID NO: 14 ssRNA ssRNA ACGCCAAAAACAUAAAGAAAG SEQ ID NO: 15 asRNA asRNA UUCUUUAUGUUUUUGGCGUCUn SEQ ID NO: 16 Figure legends

Fig 1. M918 display high internalization into cells in comparison with the well characterized cell-penetrating peptides Pen and TP10 in HeLa cells after 1 h exposure with 5 μM peptides (a), as well as in a number of other cell lines (b). Membrane leakage measured by LDH release is negligible even at 25 μM concentration and 30 min exposure (c). In addition, there is no long term toxicity as recorded by the WST-1 assay (d), after exposing cells twice with new peptide after 24 and 48 h and measuring after 72 h.

Fig 2. Annexin binding to MCF-7 and MDA MB 231 cells indicate apoptosis induced by M917 but no significant induction of apoptosis by M918 or M872, ARF scramble,. However, treatment with staurosporine (STS)₁ induce significant binding of annexin. Cells were incubated over night with 25 μM peptide or 1 μM STS followed by detachment by trypsin, wash and incubation with annexin-FITC followed by wash again. Fluorescence was measured and normalized to amount of protein. The values represent the mean of at least three independent experiments done in duplicate (mean ±SEM, n=2). ^** P< 0.01

Fig 3. Confocal microscopy on a) MCF-7, (b) Hifko, and (c) HeLa cells after 30 min treatment with 1 μM fluoresceinyl labeled M918 peptide reveals primarily vesicular uptake.

Fig 4. Intracellular amounts of intact and degraded M918 peptide in MCF-7 cells. Cells were exposed to 5 μM peptides for 30, 60 or 180 minutes. Cell surface associated peptide were modified by diazotized 2-nitroaniline and intact and degraded peptides were identified and quantified by HPLC. Compared to other CPPs M918 is very stable, possibly explaining the observed high uptakes.

Fig 5. M918 forms complexes with siRNA more readily than Pen and TP10 with increasing molar ratio measured by quenching of ethidium bromide (a). Complexes between Pen (b), TP10 (c) and M918 (d) and fluoresceinyl labeled dsDNA or siRNA are translocated into HeLa cells in conjunction with their efficiency to form complexes.

Fig 6. Protein transduction by co-incubation with CPPs and coupling via biotin. FITC labeled streptavidin (a, b) and avidin (c, d) was co-incubated for 30 min in 0,9% NaCI with different concentrations of peptide (a, c) and added to HeLa cells for 90 min or incubated with 4 times molar excess of biotin labeled peptide (b, d) enabling binding to all pockets. The data clearly illustrate that M918 is a potent delivery vector, conferring protein transduction both by co- incubation and by coupling via biotin. Fig 7. M918 internalizes to cells via endocytosis independent of GAGs on the cell surface, (a) 200.000 HeLa cells/well seeded in 12-well plates 24 h prior experiment, were incubated with 5 μM fluoresceinyl labeled M918 or 25 μg/ml rhodamine-transferrin for 1 h in 500 μl serum free DMEM and pre-treated for 30 min with 75 μM chloroquine (cq), 0.2 M sucrose, or low temperature, (b) CHO and CHO2242 cells treated as in (a). Uptake of the three peptides in CHO cells were set to 100%. The values represent the mean of at least three independent experiments done in triplicate (mean ±SEM, π=3). ^* P< 0.05, *^* P< 0.01 , ^*** P< 0.001 (a)(ANOVA Dunnett^'s), (b)(T-test).

Fig 8. Co-localization study between M918 and the endocytosis markers transferrin and dextran reveal partial co-localization with both markers, (a) brightfield image, (b) M918, (c) endocytosis markers, dextran or transferrin, (d) composite. MCF-7 cells were treated with 1 μM fluoresceinyl labelled peptide for 30 min together with 1 mg/ml of the fluid-phase endocytosis marker rhodamine-dextran or 25 μg/ml of the clathrin mediated endocytosis marker rhodamine-transferrin. Cells were washed three times with HKR prior analysis.

Fig 9. M918-PNA enters cells via macropinocytosis and dose-dependently induces splice-correction. 60 000 HeLa pLuc 705 cells/well were seeded 48 h prior experiment in 24-well plates after which cells were treated with peptide-PNA conjugates for 1 h in serum free DMEM media followed by 16 h in serum media, (a) Splice correction after treatment with 5 μM peptide-PNA conjugates presented as fold-increase in splicing over untreated cells. 10 μM inverted PNA (invPNA) and PNA represent controls for specificity and delivery efficacy of M918, respectively, (b) Uptake and splicing after treatment with different M918-PNA concentrations presented as RLU/mg. (c) Cells were pre-treated with 4 μM cytochalasin D, 50 nM wortmannin, or pre-maintained at 4⁰C for 30 min and treated at for 1 h with 5 μM M918-PNA. RLU/mg values obtained with only conjugate were set to 100 % splice correction, (d) Cells were treated as in (a) with M918-PNA at indicated concentrations together with 75 μM chloroquine (cq). The values represent the mean of three independent experiments each done in triplicate (mean ±SEM, n=3). ^* P< 0.05, ^** P< 0.01 , ^*** P< 0.001 (c) (ANOVA Dunnett^'s), (d) (ANOVA Bonferroni).

FIG. 10. Peptides derived from the N-terminal part of p14ARF M917 and M 872 translocate into cells, (a) Quantitative uptake of peptides in MCF-7 cells. Cells, 100 000 seeded in 12-well plates two days before, were incubated with 5 μM fluoresceinyl labeled peptides for 1 h in HKR followed by washing 2 times with HKR. Cells were detached by trypsination, centrifuged and lysed with 0.1 M NaOH. Fluorescence was measured at 494/518 nm. TP10 serve as positive controls and YDEGE as a negative control to set the level of peptide endocytosis. The values represent the mean of at least three independent experiments done in triplicate (mean ±SEM, n=3). (b) Splice correction assay in HeLa pLuc 705 cells. Cells were treated with 5 μM conjugates for 2 h in serum free media followed by replacement to serum medium for additionally 16 h. Cells were washed and lysed in 0.1 % Triton X-100. Luminescence was normalized to amount of protein and set to fold increase compared to untreated cells. 75 μM chloroquine (cq) was optionally co-added with conjugates. Values represent the mean of at least three independent experiments done in triplicate (mean ±SEM, n=3). * P< 0.05, ^** P< 0.01 (ANOVA Dunnett's).

FIG. 11. Confocal microscopy of M917, M872 and TP10 verify translocation into cells.

(a) Dextran, (b) M917, (c) M872, (d) TP10. MCF-7 cells seeded in NUNC 8 chambers two days prior experiment to reach 50% confluence were exposed to 1 μM fluoresceinyl labeled peptide or 1 μM FITC-dextran for 30 min together with 25 μg/ml rhodamine-transferrin in HKR. Cells were washed three times with HKR before confocal microscopy.

FIG. 12. Endocytosis is involved in uptake of ARF derived peptides and M917 and M 872 is relatively stable inside cells.

(a) MCF-7 cells were pre-treated with indicated endocytosis inhibitor for 30 min after which media was replaced with new HKR containing 5 μM peptide and inhibitor. Following procedure were performed as in (Fig. 1a). The values represent the mean of at least three independent experiments done in duplicate (mean ±SEM, n=3). ^* P< 0.05, ^** P< 0.01 (ANOVA Dunnett^'s)

Intracellular amounts of intact and degraded M917 (b), and M872 (c) peptides in MCF-7 cells. Cells, 100 000 seeded in 24-well plates the day before, were exposed to 5 μM peptides for 30, 60 or 180 min. Cell surface associated peptide was modified by diazotized 2- nitroaniline and intact and degraded peptides were quantified by HPLC. The values represent the mean of three independent experiments done in duplicate (mean ±SEM, n=3).

FIG. 13. Effects of M917 and M872 on cell proliferation. MCF-7 cells (a), and MDA MB 231 cells (b). Cells were treated 24 and 48 h after seeding of 10 000 cells in 96 well plates, with indicated peptide concentration and proliferation was assayed after 72 h by WST-1. (c) MCF-7 cells were treated as in (a) and (b) with 5 μM peptides but with addition of 75 μM chloroquine (cq) or 0.2 μM staurosporine (STS). Values represent the mean of at least three independent experiments done in quadruplicate (mean ±SEM, n=3). ^** P< 0.01 (a, b, c) (ANOVA Dunnett^'s), (c) (ANOVA Bonferroni). FIG. 14. Assessment of cell membrane integrity. Membrane disturbance was measured for M917 peptide (a, c) and M872 (b, d) by LDH leakage (a, b) and deoxyglucose leakage (c, d) in MCF-7 cells. LDH leakage from MCF-7 cells, 25 000 cells seeded in 96 wells the day before, was measured by CytoTox-ONE assay after exposure to 1 , 10, 25 or 50 μM peptide for 30 and 60 min. Deoxyglucose leakage was measured after loading cells, 100 000 seeded in 24-well plates the day before, with ³H-deoxyglucose and treating with indicated peptide concentration for 5, 15, 30 or 60 min. The values represent the mean of at least three independent experiments done in triplicate (mean ±SEM, n=3).

FIG. 15. Detection of nuclear morphology changes after M917 treatment (a) Nuclear morphology changes of MCF-7 cells were visualized and scored after Hoechst staining. Cells were treated with 10 μM peptides with and without 75 μM chloroquine in RPMI 1640 media for 2 h, and changed to full growth media for 24 h. Fluorescent nuclei were screened for normal and apoptotic nuclei (fragmented and condensed nuclei), (b) Representative pictures of nuclear morphology in MCF-7 cells, as scored in (a). The values represent the mean of 2 independent experiments done in triplicate (mean ±SEM, n=2). ^** P< 0.01 (ANOVA Dunnett^'s) (ANOVA Bonferroni).

FIG. 16. Flow cytometric detection of apoptosis by annexin V and Pl. M917 induces apoptosis while M918 dispay no signs of apoptosis.

(a) Representative dot plots of propidium iodide (Pl) vs annexin V - FITC. MCF-7 cells were treated with 25 μM peptide or 0.1 μM STS for 4 h in media without serum followed by addition of full growth media with peptide for an additional 14 h. The lower left contains the vital (double negative) population. The lower right contains the apoptotic (Annexin V7PI^~) population. The upper right contains the late apoptotic/necrotic (Annexin V⁺ZPI⁺) population. In addition, the upper left quadrant contains the pre-necrotic (Annexin V7PI⁺) population, (b) Data showing the percentage of apoptotic and late apoptotic/necrotic cells counted from dot plots in one representative experiment performed in triplicate. References

I . El-Andaloussi S, Holm T and Langel U. Cell-penetrating peptides: mechanisms and applications. Curr Pharm Des 2005; 11: 3597-3611. 2. Pooga M, Soomets U, Hallbrink M, et al. Cell-penetratingg PNA constructs regulate galanin receptor levels and modify pain transmission in vivo. Nat Biotechnol 1998; 16: 857-861. 3. Richard JP, Melikov K, Vives E, et al. Cell-penetrating peptides. A reevaluation of the mechanism of cellular uptake. J Biol Chem 2003; 278: 585-590. 4. Fittipaldi A, Ferrari A, Zoppe M, et al. Cell membrane lipid rafts mediate caveolar endocytosis of HIV-1 Tat fusion proteins. J Biol Chem 2003; 278: 34141-34149. 5. Lundberg M and Johansson M. Positively charged DNA-binding proteins cause apparent cell membrane translocation. Biochem Biophys Res Commun 2002; 291: 367-371. 6. Kang SH, Cho MJ and KoIe R. Up-regulation of luciferase gene expression with antisense oligonucleotides: implications and applications in functional assay development. Biochemistry 1998; 37: 6235-6239. 7. Fischer R, Mader O, Jung G, et al. Extending the applicability of carboxyfluorescein in solid-phase synthesis. Bioconjug Chem 2003; 14: 653-660. 8. Ui-Tei K, Naito Y, Takahashi F, et al. Guidelines for the selection of highly effective siRNA sequences for mammalian and chick RNA interference. Nucleic Acids Res 2004; 32: 936-948.

9. Oehlke J, Scheller A, Wiesner B, et al. Cellular uptake of an alpha-helical amphipathic model peptide with the potential to deliver polar compounds into the cell interior non-endocytically. Biochim Biophys Acta 1998; 1414: 127-139.

10. Petersen H, Kunath K, Martin AL, et al. Star-shaped poly(ethylene glycol)-block- polyethylenimine copolymers enhance DNA condensation of low molecular weight polyethylenimines. Biomacromolecules 2002; 3: 926-936.

I 1. Saar K, Lindgren M, Hansen M, et al. Cell-penetrating peptides: a comparative membrane toxicity study. Anal Biochem 2005; 345: 55-65.

12. El-Andaloussi S, Johansson H, Magnusdottir A, et al. TP10, a delivery vector for decoy oligonucleotides targeting the Myc protein. J Control Release 2005; 110: 189- 201.

13. Lundberg P and Langel U. Uptake Mechanisms of Cell-Penetrating Peptides Derived from the Alzheimer^'s Disease Associated Gamma-Secretase Complex. International

Journal of Peptide Research and Therapeutics 2006; 12: 105-114. 14. El-Andaloussi S₁ Johansson HJ, Lundberg P, et al. Induction of splice correction by cell-penetrating peptide nucleic acids. J Gene Med 2006; 8: 1262-1273.

15. Abes S, Williams D, Prevot P, et al. Endosome trapping limits the efficiency of splicing correction by PNA-oligolysine conjugates. J Control Release 2006; 110: 595-604.

Claims

1. A cell-penetrating peptide or construct consisting of a cell-penetrating peptide of 15 to 25 amino acids comprising at least amino acids 1 - 14 of a mature mammalian tumor suppressor protein ARF or a scrambled or partially inverted sequence thereof, optionally linked to one or more members of the group consisting of a cargo, a detection marker, a cell- homing peptide, a receptor ligand, a linker, and a peptide spacer comprising a selective protease cleavage site coupled to an inactivating peptide.

2. The construct according to claim 1 , wherein the sequence of the mammalian tumor suppressor protein ARF is a sequence of the human p14ARF protein or a scrambled of partially inverted sequence thereof.

3. The construct according to claim 1 , wherein the sequence of a mammalian tumor suppressor protein ARF is a sequence of a mouse p19ARF protein or a scrabled or partially inverted sequence thereof.

4. The construct according to claim 2, wherein the sequence of the human p14ARF protein is SEQ ID NO:2, and the partially inverted sequence thereof is SEQ ID NO: 1.

5. The construct according to claim 3, wherein the sequence of the mouseARF protein is SEQ ID NO: 17, and the partially inverted sequence thereof is SEQ ID NO: 18.

6. The construct according to claim 1 , wherein the cell-homing peptide is a tumor-homing peptide.

7. The construct according to claim 1 , wherein the selective protease cleavage site is a cathepsin cleavage site.

8. The construct according to any one of claims 1 , wherein the cargo is selected from the group consisting of oligonucleotides, peptide nucleic acids, locked nucleic acids, small interfering RNAs₁ antisense oligonucleotides, plasmid DNAs, peptide ligands, anticancer drugs, cytotoxic peptides, bioactive peptides, diagnostic agents and proteins.

9. The construct according to claim 8, wherein the anticancer drugs are selected from an alkylating agent, an antimetabolite and a cytotoxic antibiotic.

10. The construct according to claim 9, wherein the alkylating agent is 4-[4-Bis(2- chloroethyl)amino)phenyl]butyric acid (chlorambucil) or 3-[4-(Bis(2-chloroethyl)amino)- phenyl]-L-alanine (Melphalan), the antimetabolite is Λ/-[4-(Λ/-(2,4-Diamino-6-pteridinyl- methyl)methylamino)-benzoyl]-L-glutamic acid (Methotrexate) and the cytotoxic antibiotic is (8S,10S)-10-[(3-Amino-2, 3, 6-trideoxy-α-L-/yxo-hexopyranosyl)oxy]-8-glycoloyl-7, 8,9,10- tetrahydro-6,8,11-trihydroxy-1-methoxy-5,12-naphthacenedione (Doxorubicin).

11. The construct according to claim 1 , wherein the construct comprises a peptide spacer comprising a selective protease cleavage site coupled to an inactivating peptide and the construct has a sequence selected from the group consisting of SEQ ID NOs: 3 - 6.

12. The construct according to claim 1 , wherein the construct comprises a peptide ligand and has a sequence selected from SEQ ID NO: 7 and SEQ ID NO: 8.

13. A method of delivering a cargo into a cell comprising contacting a construct according to any one of claims 1 -12 with a target cell in vivo or in vitro.