Bridged Analogues for p53-Dependent Cancer Therapy Obtained by S-Alkylation

Abstract

A small library of anticancer, cell-permeating, stapled peptides based on potent dual-specific antagonist of p53–MDM2/MDMX interactions, PMI-N8A, was synthesized, characterized and screened for anticancer activity against human colorectal cancer cell line, HCT-116. Employed synthetic modifications included: S-alkylation-based stapling, point mutations increasing hydrophobicity in key residues as well as improvement of cell-permeability by introduction of polycationic sequence(s) that were woven into the sequence of parental peptide. Selected analogue, ArB14Co, was also tested in vivo and exhibited potent anticancer bioactivity at the low dose (3.0 mg/kg). Collectively, our findings suggest that application of stapling in combination with rational design of polycationic short analogues may be a suitable approach in the development of physiologically active p53–MDM2/MDMX peptide inhibitors.

Introduction

In recent years lack of funds has made cutting edge research in peptide chemistry significantly more challenging. A case in point is “stapling” of peptides that emerged as a leading strategy in the development of novel peptide-based drug candidates since a pivotal Blackwell & Grubbs paper (Blackwell and Grubbs, 1998). Numerous studies used this approach to develop potent bioactive compounds even resulting in the foundation of a specialized company, Aileron Therapeutics, Inc. However, reagents necessary to carry out stapling related research are rather expensive, especially olefin functionalized non-natural amino acids which form “staple” (α-helix-stabilizing olefin bridge) during a ring-closing metathesis (RCM) reaction. This makes related SAR peptide studies particularly expensive and risky, especially if one works on difficult or long sequences where yield becomes an issue. In this “classical” approach the peptide staple is efficiently created in a two-step process between strategically positioned olefin functionalized non-natural amino acid side chains. The first step, catalyzed by Grubbs catalyst, results in olefin containing bridge that is subsequently catalytically reduced to saturated hydrocarbon (alkane), effectively locking the peptide into a stable α-helix conformation (Bernal et al., 2007; Blackwell et al., 2001; Blackwell and Grubbs, 1998; Walensky et al., 2004). Such helix stabilization had been shown to dramatically increase the helicity, potency, resistance to proteolytic degradation and cell permeability of α-helical peptides (Bautista et al., 2010; Bernal et al., 2007; Bird et al., 2010; Kim et al., 2011; Kim and Verdine, 2009; Long et al., 2013; Schafmeister et al., 2000)

Over the years novel strategies for peptide stapling have emerged as alternatives to the RCM approach (Lau et al., 2015a). These include hydrazone bridge (Cabezas and Satterthwait, 1999), oxime bridge (Haney et al., 2011), 1,4-disubstituted-[1,2,3]-triazole linkage (Holland-Nell and Meldal, 2011; Ingale and Dawson, 2011; Kawamoto et al., 2012; Lau et al., 2014c; Lau et al., 2014b; Lau et al., 2014a; Lau et al., 2015b; Scrima et al., 2010), metal chelation (Ghadiri and Choi, 1990; Ruan et al., 1990), disulfide bond formation (Almeida et al., 2012; Jackson et al., 1991; Leduc et al., 2003), lactam ring formation (Fujimoto et al., 2008; Geistlinger and Guy, 2001; Geistlinger and Guy, 2003; Houston, Jr. et al., 1995; Osapay and Taylor, 1992; Phelan et al., 1997) and S-alkylation based staples employing either α-haloacetamide alkylation of single cysteine (Brunel and Dawson, 2005; Cardoso et al., 2007; Galande et al., 2004; Woolley, 2005) or bridging two cysteines with bis-S-alkylating linker(s) (de Araujo et al., 2014; Jo et al., 2012; Muppidi et al., 2011b; Muppidi et al., 2011a; Muppidi et al., 2012; Spokoyny et al., 2013; Szewczuk et al., 1992; Timmerman et al., 2005; Wilkinson et al., 2007; Zhang et al., 2007; Zhang et al., 2008). Among these, the last seems to be most flexible approach as a wide range of inexpensive bis-thiol-reactive linkers is commercially available, including rigid aromatic derivatives (Chua et al., 2015; Jo et al., 2012; Muppidi et al., 2011b; Muppidi et al., 2011a; Muppidi et al., 2012; Timmerman et al., 2005; Zhang et al., 2007) and aliphatic counterparts (Byrne and Stites, 1995; Chua et al., 2015; Lindman et al., 2001; Wilkinson et al., 2007). In addition, the availability of various cysteine homologs: (L)Cys, (D)Cys, (L)homoCys, (D)homoCys, (L)Pen, and (D)Pen provides an additional option for “fine tuning” of pre-selected active derivatives. Moreover, those S-alkylation/stapling reactions can be carried out in water-based solutions without any protecting groups and due to use of standard amino acids (Cys and its homologs) costs are relatively low. Notably, the use of multi-thiol-reactive linkers has a remarkably long tradition, as an application for this purpose of the aromatic derivative, 1,3,5-tris(bromomethyl)benzene, was initially reported by Kemp & McNamara in 1985 (Kemp and McNamara, 1985) and use of its bis-reactive homologs, 1,2-bis(bromomethyl)benzene, and 1,3-bis(bromomethyl)benzene was described only few years later (Szewczuk et al., 1992). This methodology was also successfully applied in peptide drug development (Timmerman et al., 2007), including phage display (Angelini et al., 2012a; Baeriswyl et al., 2012; Baeriswyl et al., 2013; Baeriswyl and Heinis, 2013a; Baeriswyl and Heinis, 2013b; Bellotto et al., 2014; Chen et al., 2012; Chen et al., 2013; Chen et al., 2014b; Chen et al., 2014a; Heinis et al., 2009; Rentero-Rebollo et al., 2014; Timmerman et al., 2007) as well as peptide-albumin (Angelini et al., 2012c; Pollaro et al., 2014) and peptide-antibody drug conjugates (ADCs) (Angelini et al., 2012b).

We decided to apply this “cheap” strategy to the synthesis of p53–MDM2/MDMX bridged peptide inhibitors that are currently of great interest (Bernal et al., 2007; Bernal et al., 2010; Brown et al., 2009; Brown et al., 2013; Chang et al., 2013; Khoo et al., 2014).

The p53 pathway remains major target in anti-cancer research (Brown et al., 2009; Cheok et al., 2011). Generally, its deregulation causes malignant transformation of normal cells and approximately 50% of all cancers have a mutation in TP53 gene (Menendez et al., 2009), with varying mutations’ frequencies, ranging from ~10–70%, dependent on cancer type (Brosh and Rotter, 2009). Moreover, in all p53 pathway related cancers, only half is associated with TP53 mutations, with remaining cases being caused by overexpression of the p53-negative regulators: MDM2 (Momand et al., 1992) and MDM4, (Danovi et al., 2004; Laurie et al., 2006; Riemenschneider et al., 1999) or deletion/inactivation of ARF, the p53-positive regulator and MDM2 inhibitor (Esteller et al., 2001; Lowe and Sherr, 2003; Sherr and Weber, 2000).

The p53 stress response protein is a transcription factor with a crucial role in tumour suppression (Lowe et al., 2004; Vousden and Lane, 2007). It regulates various biological activities (Menendez et al., 2007; Menendez et al., 2009; Vousden and Prives, 2009) including, apoptosis (Bensaad et al., 2006; Vousden, 2006), cell cycle (Kastan et al., 1992; Kuerbitz et al., 1992), senescence (Garbe et al., 2007), immune response (Taura et al., 2008), cell differentiation (Molchadsky et al., 2008), motility (Qin et al., 2009) and migration (Roger et al., 2006; Singh et al., 2007), angiogenesis (Pal et al., 2001; Teodoro et al., 2006; Zhang et al., 2000), DNA (Helton and Chen, 2007) and energy (Green and Chipuk, 2006; Matoba et al., 2006; Vousden and Ryan, 2009) metabolism, microRNA processing (Suzuki et al., 2009), cell–cell communication (Yu et al., 2006) and the DNA damage response and repair (de Souza-Pinto et al., 2004; Sengupta and Harris, 2005; Sommers et al., 2005). Normal levels of p53 protein are low, due to rapid ubiquitin-dependent proteasomal degradation which is controlled largely by the E3 ubiquitin ligase MDM2 (Brooks and Gu, 2006) (known as HDM2 in human), which in turn is also a target of transcriptional regulation by p53 (Murray-Zmijewski et al., 2008). Additionally, MDM2 may regulate p53 function through direct, high affinity binding to p53’s N-terminal transactivation domain which effectively blocks its function (Momand et al., 1992), and by impairing nuclear import of the p53 protein (Haupt et al., 1997; Wade and Wahl, 2009). MDM4 (also known as HDM4, MDMX or HDMX), another negative regulator of p53, also plays an important role in its function. Although MDM4 shows a similar to MDM2 p53-binding properties, it lacks its E3 ubiquitin ligase activity and consequently cannot affect p53s stability (Marine et al., 2007; Marine, 2011; Migliorini et al., 2002a; Migliorini et al., 2002b; Toledo et al., 2006; Toledo and Wahl, 2006; Wade and Wahl, 2009). However, MDM4, a structural homologue of MDM2, can form a MDM2/MDMX heterodimers through their C-terminal RING finger domains, which potentiate the ubiquitination of p53, and subsequently its degradation (de Graaf et al., 2003; Linares et al., 2003; Pan and Chen, 2003). Notably, in contrast to MDM2, MDMX is not under transcriptional control of p53 (Riedinger and McDonnell, 2009). In addition stability of p53 is influenced by ARF (also known as p14ARF) a tumour suppressor that interacts with MDM2 and inhibits p53 degradation(Sherr and Weber, 2000).

Various stressors, including DNA damage and oncogenic stress, increase the amount of p53 by disrupting its degradation resulting in p53’s stabilization and functional activation (Lavin and Gueven, 2006; Murray-Zmijewski et al., 2008; Sherr, 2006). Once activated, p53 can elicit various cellular responses, including growth arrest, senescence and apoptosis.

Since p53 is commonly inactivated in various human cancers, its “re-activation” represents a major therapeutic strategy. Specifically, targeting the interactions of p53 with its major negative regulators, MDM2 and MDM4, recently emerged as an important therapeutic approach, with multiple examples of both, small molecules and peptides being utilized as p53-MDM2/MDM4 inhibitors (Li and Lozano, 2013; Saha et al., 2013; Wang et al., 2012; Zhao et al., 2013). p53-MDM2/MDM4 interactions are mediated primarily by three key hydrophobic residues in p53: ¹⁹Phe, ²³Trp, and ²⁶Leu (Kussie et al., 1996; Pazgier et al., 2009; Popowicz et al., 2008), and similar though not identical, hydrophobic surface grooves in MDM2 and MDM4 (Zhao et al., 2013). Over past few years, mutational studies and phage display has produced multiple peptide-based p53–MDM2/MDMX inhibitors, including all-D, stapled and bridged compounds with increased resistance to proteolysis (Bernal et al., 2007; Bernal et al., 2010; Chang et al., 2013; Guo et al., 2014; Hu et al., 2007; Li et al., 2009; Li et al., 2010b; Li et al., 2010a; Liu et al., 2010b; Liu et al., 2010a; Madden et al., 2011; Madhumalar et al., 2009; Pazgier et al., 2009; Zhan et al., 2012). Available in vitro/in vivo data (Chang et al., 2013; Hu et al., 2007; Liu et al., 2010a) suggest that the limiting factor in the inhibitory activity of these compounds is low cell-permeability since liposomal based formulations appear to be significantly more potent than “naked” peptide (Liu et al., 2010a). Even though, peptide stapling was postulated to increase cell permeability (Bernal et al., 2007) its effects seems to be limited as liposomal formulation (Liu et al., 2010a) of ^DPMI-α (dosage: 3 and 7.5 mg/kg, K_D-MDM2=219 nM(Liu et al., 2010b)) was significantly more potent than MPEGDSPE-based micellar formulation of ATSP-7041(Chang et al., 2013) (dosage: 15–30 mg/kg, K_D-MDM2=0.91 nM), despite a superior binding affinity of the latter, although this assessment may not be totally valid since different cell lines/schedules were used in both studies. Conjugates of similar peptides with cell permeating molecules (Arg₃, (D)Arg₉ and spermine) were also prepared, though respective in vivo studies were not reported (Lau et al., 2014c; Lau et al., 2014b; Liu et al., 2010a; Muppidi et al., 2011a). Other limited studies employing an S-alkylation of a pair of cysteines located at i,i+7 positions of analogous compounds, using 4,4’-bis-bromomethyl-biphenyl (Bph) and 6,6’-bis-bromomethyl-[3,3’]-bipyridine (Bpy) were also reported (Muppidi et al., 2011b; Muppidi et al., 2011a). Again, in vivo anticancer activity was not presented.

Materials and Methods

Peptide Synthesis

All peptides were synthesized as C-terminal amides (except for ArB14Ao) by the solid phase method using CEM Liberty automatic microwave peptide synthesizer (CEM Corporation Inc., Matthews, NC), applying 9-fluorenylmethyloxycarbonyl (Fmoc) chemistry and commercially available amino acid derivatives and reagents (EMD Biosciences, San Diego, CA and Chem-Impex International, Inc., Wood Dale, IL). Rink Amide MBHA Resin (Piperazine-2-Chlorotrityl for ArB14Ao) (EMD Biosciences, San Diego, CA) was used as a solid support. Peptides were cleaved from resin using modified reagent K (TFA 94% (v/v); phenol, 2% (w/v); water, 2% (v/v); TIS, 1% (v/v); EDT, 1% (v/v); 2 hours) and precipitated by addition of ice-cold diethyl ether. Linear peptides were purified by preparative reverse-phase high performance liquid chromatography (RP-HPLC) and their purity evaluated by matrix-assisted laser desorption ionization spectrometry (MALDI-MS) as well as analytical RP-HPLC.

Stapling using the bis(halogenomethyl)aryl linkers

1. 1,2-bis(bromomethyl)benzene,

2. 1,3-bis(bromomethyl)benzene,

3. 1,4-bis(bromomethyl)benzene,

4. 1-(bromomethyl)-3-[3-(bromomethyl)benzyl]benzene,

5. 1-(chloromethyl)-4-[4-(chloromethyl)phenoxy]benzene, may be performed either on the solid support or in solution.

In solution

Peptide (1 eq) containing 2 properly placed cysteines was mixed in the solution of DMF (N,N-dimethylformamide) and DMSO (dimethylsulfoxide) (1:1 ratio, concentration of peptide: 0.5–1 mg/mL) with the bis(halogenomethyl)aryl compound (XCH₂-Ar-CH₂X, 1 eq.) in the presence of cesium carbonate (Cs₂CO₃, 10 eq.) and tetrabutylammonium iodide (TBAI, 1 eq.). Reaction was monitored by HPLC/mass spectrometry. Alternatively stapled compounds were obtained using modified Timmerman protocol (Timmerman et al., 2005). Briefly, linear peptides were solubilized in 50% acetonitrile in water containing 50 mM NH₄HCO₃ (~0.5 mg/mL) and then appropriate bis-functional S-alkylating reagent was added. Reaction(s) were stirred for 18 h and solutions subsequently lyophilized. Obtained crude compounds were purified by preparative RP-HPLC and their purity evaluated by MALDI-MS as well as analytical RP-HPLC

On the solid support

After assembly of the peptide, final Fmoc group was removed and side chains of Cys residues deprotected (2% trifluoroacetic acid (TFA), 2% triisopropyl silane (TIS) in dichloromethane (DCM), 3×10 min). Then peptide was washed with DMF and bis(thioether)arylbridging performed on resin using appropriate bis-functional S-alkylating reagent (1 eq.) in the presence of Cs₂CO₃, (10 eq) and TBAI (1 eq) in DMF (overnight). Subsequently, peptides were cleaved and processed as described above.

Stapling with 2-chloro-N-(2-(2-chloro-acetyloamino)-ethyl)-acetamide linker: was performed using Timmerman protocol (Timmerman et al., 2005) described above.

Stapling with divinylsulfone (DVS): was performed using previously described protocol (Wilkinson et al., 2007) from corresponding purified linear peptides in reaction with 1 eq. (mole:mole ratio) of DVS in solution of 10 mM NH₄HCO₃ in 50% ACN in water (overnight). Subsequently, solution was lyophilized, and obtained crude solid residues purified by preparative RP-HPLC.

Stapling with 1,4-dibromobutane and 1,7-dibromoheptane: was performed in solution using modified Włostowski S-alkylation protocol (Wlostowski et al., 2010) that we adapted to peptides (Chua et al., 2015). Briefly, linear peptides were solubilized in anhydrous methanol (~0.5 mg/mL) containing, if necessary up 25% of DMSO. Subsequently 1 equivalent of either 1,4-dibromobutane and 1,7-dibromoheptane was added followed by 1,1,3,3-tetramethylguanidine (final conc.=0.35%, vol/vol). Reaction mixture was stirred 15 min at room temperature and then 90 min at 50 °C. Subsequently, solution was lyophilized, and obtained crude solid residues purified by preparative RP-HPLC.

Synthesis of analogues ArB14k1÷k4 and ArB18Ek: Peptides ArB14k1÷k4 and ArB18Ek were assembled using SPPS (Fmoc-(D)Lys(Dde)-OH was used). Then side chain of (D)Lys was selectively deprotected using 2% hydrazine in DMF (3×10 min) and subsequently chloroacetylated using chloroacetic anhydride (5 eq./30 min). Then peptides were cleaved from the resin and cyclized in solution of 65 mM NH₄HCO₃ in 70% ACN in water (overnight) giving thioether bridge. Subsequently, solution was lyophilized, and obtained crude solid residues purified by preparative RP-HPLC.

Oxidation of thioethers to corresponding sulfones: (ArB14Dmx and ArB14Emx): Oxidation of thioethers to sulfones was performed using OXONE^® (Duran et al., 2006) (Sigma-Aldrich, KHSO₅×0.5 KHSO₄×0.5 K₂SO₄) in methyl alcohol:water mix (3:1) using 3 eq. of OXONE^® per each thioether group. Reaction was carried out in room temperature for 3 h. Subsequently solutions were freeze-dried and purified using preparative RP-HPLC.

Synthesis of ArB conjugates with

Azidothymidine (AZT): was conjugated using Cu(I)-catalyzed azide-alkyne 1,3-dipolar Huisgen's cycloaddition (CuAAC, “click chemistry”). Alkyne component was synthesized from ArB14Co in one step synthesis: 5-Hexynoic acid in dry DMSO (1 eq., Sigma-Aldrich Corp., St. Louis, MO) was pre-activated with N-(3-dimethylaminopropyl)-N’-ethylcarbodiimide (EDC) in the presence of 3 eq. of N-hydroxysuccinimide (NHS) for 3 h. Subsequently, 1 eq. of ArB14Co in dry DMSO was added followed by 5 eq. of NMM. The solution was agitated overnight and subsequently evaporated under reduced pressure on SpeedVac. Remaining residue was purified on RP-HPLC. Obtained ArB-alkyne derivative was then “clicked” with AZT using published protocol (Lau et al., 2015b) using CuSO₄/sodium ascorbate catalyst in 50% tert-buthanol in water. Subsequently, reaction mixture was freeze-dried and purified using RP-HPLC.

Bleomycin A₅ (BLM): was conjugated with ArB analogue(s) using bis-amine-reactive linker, dithiobis(succinimidyl propionate) (DSP, Pierce Biotechnology, Rockford, lL). Briefly, specific ArB peptide was synthesized with an additional 6-aminohexanoic acid (Ahx) residue at N-terminus. Then chosen stapling procedure was performed and stapled peptide was purified using RP-HPLC. Subsequently, equimolar amounts of Ahx-ArB peptide and BLM (EMD Millipore, Temecula, CA) were mixed in dry DMSO (5 mg/ml). Then 10 eq. of N-methylmorpholine (NMM) was added followed by 1 eq. of DSP. The reaction was mixed for 18 h and subsequently evaporated under reduced pressure on SpeedVac. Remaining residue was purified using RP- HPLC.

Chlorambucil (CLB): was introduced using N-terminal amine group employing N,N’-diisopropylcarbodiimide (DIC, Sigma-Aldrich Corp., St. Louis, MO) as a coupling reagent. Reactions were carried out on resin using N,N-dimethylformamide as a solvent (2h).

Cholesterol (Chol): was introduced using N-terminal amine group employing cholesterol chloroformate (Sigma-Aldrich Corp., St. Louis, MO) resulting in urethane type connectivity. Reactions were carried out on resin in the presence of DIPEA using dichloromethane as a solvent (2h).

Docetaxel (DTX): was conjugated with ArB analogue(s) using the same strategy that was employed for BLM conjugates. Suitable DTX derivative, 2’-O-glycyl-docetaxel, was synthesized using previously described method (Zhigaltsev et al., 2010). Briefly, Trt-Gly-OH (1 eq.) and DTX (1 eq.) were dissolved in dry dichloromethane and cooled in ice bath under the inert-gas blanket. Subsequently, 2 eq. of 2-fluoro-1-ethyl pyridinium tetrafluoroborate (FEP) dissolved in minimal amount of dry dichloromethane were added followed by 3 eq. of DIPEA. The reaction was agitated for 18 h with gradual warming to room temperature and then evaporated. Solid residue was re-dissolved in solution of 2% TFA in dry dichloromethane to remove Trt protecting group (30 min) and evaporated again. Obtained residue was purified using RP-HPLC giving 2’-O-glycyl-docetaxel that was subsequently used for conjugation(s).

Methotrexate (MTX): was introduced using N-terminal amine group employing N,N’-diisopropylcarbodiimide (DIC, Sigma-Aldrich Corp., St. Louis, MO) as a coupling reagent. Reactions were carried out on resin in the presence of DIPEA using N,N-dimethylformamide as a solvent (2h).

Analytical RP-HPLC

Experiments were performed on a Varian ProStar 210 HPLC system equipped with ProStar 325 Dual Wavelength UV-Vis detector with the wavelengths set at 220 nm and 280 nm (Varian Inc., Palo Alto, CA). Mobile phases consisted of solvent A, 0.1% TFA in water, and solvent B, 0.1% TFA in acetonitrile. Analyses of peptides were performed with an analytical reversed-phase C₁₈ SymmetryShield™ column, 4.6×250 mm, 5 µm (Waters, Milford, MA) or (*) an analytical reversed-phase C₁₈ Vydac 218TP54 column, 4.6×250 mm, 5 µm (Grace, Deerfield, IL) applying linear gradient of solvent B from 0 to 100% over 100 min (flow rate: 1 ml/min).

In vitro cells’ growth inhibition assay

Experiments were carried out using PrestoBlue™ Cell Viability Reagent (Invitrogen, Carlsbad, CA) according to manufacturer's protocol. Briefly, HCT-116 cells were plated in a 96-well plate at a density of 5×10³ cells/well in a total volume of 50 µl of culture media and treated with various concentrations of tested peptides (50 µl of 0–200 µM peptides in culture media). The cells' viability was assessed after 48 h by fluorescence measurement (E_x/E_m:560/590, incubation time 30 min) employing the SpectraMAX M2 microplate reader (Molecular Devices, Sunnyvale, CA). All experiments were carried out in triplicate.

Animal experiments

All animal experiments were approved by the UCLA Animal Care and Use Committee and conformed to local and national guidelines. For subcutaneous engraftment model experiments BALB/SCID gnotobiotic mice (8 weeks old) were obtained from the UCLA AALAC-accredited Department of Radiation Oncology Facility and subcutaneously injected with 2.0×10⁶ cells of human colorectal cancer cell line (HCT-116, leg). After 3 weeks, palpable tumours of approximately 5 mm diameter appeared and treatment was initiated. In general, each animal received intraperitoneally a total of 10 doses of ArB14Co at 3.0 mg/kg or 7.5 mg/kg on days 1–5 and 8–12 with 8 mice per group. The peptide was formulated in 2% Cremophor EL (Sigma-Aldrich, St Louis, MO) in phosphate-buffered saline (PBS, vol/vol). Control animals were injected with vehicle alone. Tumour(s) size was assessed with calliper and its volume calculated using formula: V=L*W²/2 (V=tumour volume, L=length, W=width, L>W) and animals sacrificed as necessary according to the UCLA Animal Care guidelines.

Results and Discussion

Based on the analysis of literature data, we decided to synthesize library of p53–MDM2/MDMX inhibitors with enhanced cell permeability afforded by multiple Arg residues “woven” into the sequence of the parent peptide. Such modifications provide desired permeability effect and simultaneously limit the size/costs of making the potential drug candidate. Physiological stability/enzymatic resistance of these peptides was enhanced by the use of unusual- and (D)-amino acids as well as by stapling based on bis-cysteine bridges obtained by means of S-alkylation with selected bis-functional compounds. Employed modifications are schematically represented in Fig. 1.

Fig. 1.

Schematic representation of modifications in ArB analogues

As a starting point for our modifications we decided to use PMI-N8A peptide, the potent dual-specific antagonist of p53–MDM2/MDMX interactions with respective K_d values of 490 pM and 2.4 nM. PMI and its derivative PMI-N8A were originally developed in Lu’s group using phage display (Li et al., 2010a; Pazgier et al., 2009). Analogous all-D-peptide inhibitors called ^DPMI-α, ^DPMI-γ and ^DPMI-δ were also discovered by the same group using mirror-image phage display (Liu et al., 2010b; Liu et al., 2010a; Zhan et al., 2012) and subsequently tested in an animal model. However even proteolysis resistant ^DPMI-α required a liposomal delivery system to be active in vivo.

Based on analysis of p53/MDM2 and PMI-N8A/MDM2 crystal structures we placed various combinations of (L)- and (D)Cys residues in specific positions of the PMI sequence taking into account the importance of particular amino acids for peptide bioactivity (Table 1). These peptides were in turn stapled using selected bis-thiol-reactive compounds (Fig. 2) that fitted into the available spatial arrangements/dimensions, based on computer-aided molecular modeling. Pairs of cysteines were placed in positions: i,i+3; i,i+4; i,i+6; and i,i+7 of the peptides. Since ^DPMI-α presumably lacks cell permeating properties, in addition, we engineered several analogues with multiple (L)/(D)Arg residues to increase cell permeability, in combination with (D)- and non-proteinaceous amino acids to enhance stability and binding properties of the peptides. Selected analogues were also conjugated with cholesterol (Chol), bleomycin A₅ (BLM), chlorambucil (CLB), docetaxel (DTX) and methotrexate (MTX), using the thiol-cleavable linker, dithiobis(succinimidyl propionate) (DSP, Pierce Biotechnology, Rockford, lL).

Table 1.

Analytical data for ArB analogues synthesized for these studies

Name	Sequence	Composition	MW Calc / Found	RT (min)	EC50 [µM] HCT-116
PMI-N8A	Thr-Ser-Phe-Ala-Glu-Tyr-Trp-Ala-Leu-Leu-Ser-Pro	C67H94N14O18	1383.56 / 1383.59	38.17*	178.8
(15–29)p53	Ser-Gln-Glu-Thr-Phe-Ser-Asp-Leu-Trp-Lys-Leu-Leu-Pro-Glu-Asn	C82H124N20O26	1806.00 / 1806.88	37.32*	NA
ArB14o	Thr-Ser-Phe-Ala-Cys-Tyr-Trp-Cys-Ala-Leu-Ser-Pro	C70H92N14O16S2	1449.71 / 1449.86	39.38*	248.6
ArB14Ao	Thr-Ser-PF5-Ala-Cys-Tyr-Na1-Cys-Ala-Cha-Ser-Pipz	C74H92N13O15F5S2	1562.74 / 1562.46	45.71*	83.3
ArB14Bo	Arg-Arg-Na2-Arg-Cys-Tyr-Na1-Cys-Arg-Cha-Arg-DISC	C97H132N28O13S2	1962.41 / 1962.36	37.05	10.7±0.7
ArB14Co	Arg-Arg-PF5-Arg-Cys-Tyr-Na1-Cys-Arg-Cha-Arg-DISC	C93H125N28O13F5S2	2002.30 / 2003.02	36.94	11.2±1.2
ArB14Cs	Arg-Arg-PF5-Arg-Cys-Tyr-Na1-Cys-Arg-Cha-Arg-DISC	C89H125N28O15F5S3	2018.32 / 2018.63	35.61	5.6±2.0
ArB14C2s	Arg-Arg-PF3-Arg-Cys-Tyr-Na1-Cys-Arg-Cha-Arg-DISC	C90H129N28O15F3S3	1996.36 / 1997.12	37.13	14.3± 2.5
ArB14Ck1	Ac-Aib-Arg-PF3-Arg-Lys-Tyr-Trp-Cys-Arg-Cha-Arg	C80H119N26O14F3S	1758.04 / 1758.35	36.21	82.14
ArB14Ck2	Ac-Arg-Arg-PF3-Arg-Lys-Tyr-Trp-Cys-Arg-Cha-Arg	C82H124N29O14F3S	1829.12 / 1829.45	34.16	66.32
ArB14Ck3	Ac-Aib-Arg-PF5-Arg-Lys-Tyr-Trp-Cys-Arg-Cha-Arg	C79H115N26O14F5S	1779.99 / 1780.75	34.79	>250
ArB14Ck4	Ac-Arg-Arg-PF5-Arg-Lys-Tyr-Trp-Cys-Arg-Cha-Arg	C81H120N29O14F5S	1851.08 / 1851.56	32.92	69.59
ArB14Co-AZT	AZT-Arg-Arg-PF5-Arg-Cys-Tyr-Na1-Cys-Arg-Cha-Arg-DISC	C110H145N32O18F5S2	2362.65 / 2363.66	37.96	30.2±2.5
ArB14Co-CLB	CLB-Ahx-Arg-Arg-PF5-Arg-Cys-Tyr-Na1-Cys-Arg-Cha-Arg-DISC	C113H153N30O15Cl2F5S2	2401.66 / 2402.78	56.41	7.5±1.1
ArB14Co-Ch	Chol-Ahx-Lys-Ahx-Arg-Arg-PF5-Arg-Cys-Tyr-Na1-Cys-Arg-Cha-Arg-DISC	C139H203N32O18F5S2	2769.45 / 2769.14	57.99	NT
ArB14Co-Ch/CLB	Chol-Ahx-LysCLB-Ahx-Arg-Arg-PF5-Arg-Cys-Tyr-Na1-Cys-Arg-Cha-Arg-DISC	C153H220N33O19Cl2F5S2	3055.65 / 3057.04	58.30	7.9±1.3
ArB14Co-MTX	MTX-Ahx-Arg-Arg-PF5-Arg-Cys-Tyr-Na1-Cys-Arg-Cha-Arg-DISC	C119H156N37O18F5S2	2551.89 / 2553.86	46.51	6.9±0.4
ArB14Co-TFAo	Tfa-Arg-Arg-PF5-Arg-Cys-Tyr-Na1-Cys-Arg-Cha-Arg-DISC	C95H124N28O14F8S2	2098.31 / 2098.91	38.00	18.7±3.4
ArB14Co-BLM	BLM-SS-Ahx-Arg-Arg-PF5-Arg-Cys-Tyr-Na1-Cys-Arg-Cha-Arg-DISC	C162H231N48O37F5S6	3730.26 / 3731.46	38.96	5.5±1.1
ArB14Co-DTX	DTX-SS-Ahx-Arg-Arg-PF5-Arg-Cys-Tyr-Na1-Cys-Arg-Cha-Arg-DISC	C150H198N31O31F5S4	3154.64 / 3154.61	46.89	21.6±1,4
ArB14Cs-BLM	BLM-SS-Ahx-Arg-Arg-PF5-Arg-Cys-Tyr-Na1-Cys-Arg-Cha-Arg-DISC	C158H231N48O39F5S7	3746.28 / 3746.15	34.19	1.3±0.5
ArB14Cs-DTX	DTX-SS-Ahx-Arg-Arg-PF5-Arg-Cys-Tyr-Na1-Cys-Arg-Cha-Arg-DISC	C146H198N31O33F5S5	3170.65 / 3171.19	45.90	8.4±1.5
ArB14Dm	Thr-Ser-Phe-Arg-Arg-Tyr-Cys-Arg-Arg-Cys-Ser-Pro	C71H109N25O16S2	1632.92 / 1632.89	28.59*	118.3
ArB14Em	Thr-Ser-Phe-Arg-Arg-Tyr-Cys-Arg-Arg-Cys-Ser-Pro	C71H109N25O16S2	1632.92 / 1633.52	28.17*	113.9
ArB14Dmx	Thr-Ser-Phe-Arg-Arg-Tyr-Cys-Arg-Arg-Cys-Ser-Pro	C71H109N25O20S2	1696.92 / 1696.72	24.00*	141.5
ArB14Emx	Thr-Ser-Phe-Arg-Arg-Tyr-Cys-Arg-Arg-Cys-Ser-Pro	C71H109N25O20S2	1696.92 / 1697.21	24.79*	NA
ArB15Am	Thr-Ser-Phe-Ala-Cys-Tyr-Trp-Gly-Cys-Leu-Ser-Pro	C69H90N14O16S2	1435.68 / 1436.82	40.10	NA
ArB15Bm	Thr-Ser-Phe-Ala-Cys-Tyr-Trp-Gly-Cys-Leu-Ser-Pro	C69H90N14O16S2	1435.68 / 1436.59	41.01	NA
ArB15Cp	Thr-Ser-Phe-Ala-Cys-Tyr-Trp-Gly-Cys-Leu-Ser-Pro	C69H90N14O16S2	1435.68 / 1436.04	40.07	>250
ArB15Dp	Thr-Ser-Phe-Ala-Cys-Tyr-Trp-Gly-Cys-Leu-Ser-Pro	C69H90N14O16S2	1435.68 / 1436.67	39.99	NA
ArB15Fo	Arg-Phe-Cys-Arg-Tyr-Trp-Cys-Arg-Leu-Arg-Pro	C78H113N25O12S2	1657.03 / 1657.86	32.64	49.64
ArB15Gm	Arg-Phe-Cys-Arg-Tyr-Trp-Cys-Arg-Leu-Arg-Pro	C78H113N25O12S2	1657.03 / 1657.46	31.92	52.19
ArB15FNico	Nic-Arg-Phe-Cys-Arg-Tyr-Trp-Cys-Arg-Leu-Arg-Pro	C84H116N26O13S2	1762.13 / 1763.63	35.97	189.0
ArB15GNicm	Nic-Arg-Phe-Cys-Arg-Tyr-Trp-Cys-Arg-Leu-Arg-Pro	C84H116N26O13S2	1762.13 / 1762.51	35.36	126.4
ArB15F/Gr	Arg-Phe-Cys-Arg-Tyr-Trp-Cys-Arg-Leu-Arg-Pro	C70H107N25O12S2	1554.89 / 1555.39	25.06*	>250
ArB15F/GNicr	Nic-Arg-Phe-Cys-Arg-Tyr-Trp-Cys-Arg-Leu-Arg-Pro	C76H110N26O13S2	1659.99 / 1660.56	28.24*	64.5
ArB15F/Gs	Arg-Phe-Cys-Arg-Tyr-Trp-Cys-Arg-Leu-Arg-Pro	C74H113N25O14S3	1673.05 / 1674.05	27.06	49.9
ArB15F/Gs-BLM	BLM-SS-Arg-Phe-Cys-Arg-Tyr-Trp-Cys-Arg-Leu-Arg-Pro	C137H208N44O37S7	3287.85 / 3288.42	29.49	1.1±0.7
ArB15F/Gs-DTX	DTX-SS-Arg-Phe-Cys-Arg-Tyr-Trp-Cys-Arg-Leu-Arg-Pro	C125H175N27O31S5	2712.22 / 2712.60	43.77	62.2
ArB15F/GNics	Nic-Arg-Phe-Cys-Arg-Tyr-Trp-Cys-Arg-Leu-Arg-Pro	C80H116N26O15S3	1778.14 / 1779.20	30.63	34.2±5.9
ArB15F/Gt	[Arg-Phe-Cys-Arg-Tyr-Trp-Cys-Arg-Leu-Arg-Pro]2Ph(CH2)4	C150H220N50O24S4	3235.95 / 3235.80	30.72	NT
ArB15F/GNict	[Nic-Arg-Phe-Cys-Arg-Tyr-Trp-Cys-Arg-Leu-Arg-Pro]2Ph(CH2)4	C162H226N52O26S4	3446.14 / 3447.36	33.62	NT
ArB15F/G0/Nic	[Nic-Arg-Phe-Cys-Arg-Tyr-Trp-Cys-Arg-Leu-Arg-Pro]Ph(CH2)4	C156H223N51O25S4	3341.05 / 3341.51	34.42*	NT
Mix	[Arg-Phe-Cys-Arg-Tyr-Trp-Cys-Arg-Leu-Arg-Pro]			34.86*
ArB15Eb	Thr-Ser-Phe-Ala-Cys-Tyr-Trp-Gly-Cys-Leu-Ser-Pro	C76H96N14O16S2	1525.81 / 1526.03	46.95	162.9
ArB15Hr	Ac-Thr-Ser-Phe-Cys-Glu-Tyr-Trp-Cys-Leu-Leu-Ser	C64H89N13O18S2	1392.60 / 1393.25	46.46	NA
ArB15Ho	Ac-Thr-Ser-Phe-Cys-Glu-Tyr-Trp-Cys-Leu-Leu-Ser	C72H95N13O18S2	1494.74 / 1494.80	49.23	NA
ArB15Hm	Ac-Thr-Ser-Phe-Cys-Glu-Tyr-Trp-Cys-Leu-Leu-Ser	C72H95N13O18S2	1494.74 / 1494.83	48.12	NA
ArB15Hg	Ac-Thr-Ser-Phe-Cys-Glu-Tyr-Trp-Cys-Leu-Leu-Ser	C68H95N13O18S2	1446.69 / 1446.79	46.31	NA
ArB15Hs	Ac-Thr-Ser-Phe-Cys-Glu-Tyr-Trp-Cys-Leu-Leu-Ser	C68H95N13O20S3	1510.75 / 1510.90	39.17	NA
ArB17Ac	Thr-Cys-Phe-Ala-Gly-Tyr-Trp-Cys-Ala-Leu-Ser-Pro	C75H94N14O16S2	1511.78 / 1511.01	44.24	NA
ArB17Bc	Thr-Cys-Phe-Ala-Gly-Tyr-Trp-Cys-Ala-Leu-Ser-Pro	C75H94N14O16S2	1511.78 / 1511.26	45.53	NA
ArB18Ac	Thr-Cys-Phe-Ala-Gly-Tyr-Trp-Gly-Cys-Leu-Ser-Pro	C74H92N14O16S2	1497.75 / 1499.28	42.66	NA
ArB18Bc	Thr-Cys-Phe-Ala-Gly-Tyr-Trp-Gly-Cys-Leu-Ser-Pro	C74H92N14O16S2	1497.75 / 1498.49	45.85	NA
ArB18Cr	Arg-Arg-Phe-Cys-Arg-Tyr-Trp-Arg-Arg-Leu-Cys	C71H112N28O12S2	1613.97 / 1614.23	28.47*	>250
ArB18Dr	Aib-Arg-Phe-Cys-Arg-Tyr-Trp-Arg-Arg-Leu-Cys	C69H107N25O12S2	1542.88 / 1543.26	29.75*	>250
ArB18Cd	Arg-Arg-Phe-Cys-Arg-Tyr-Trp-Arg-Arg-Leu-Cys	C77H120N30O14S2	1754.11 / 1754.28	24.89	NA
ArB18Dd	Aib-Arg-Phe-Cys-Arg-Tyr-Trp-Arg-Arg-Leu-Cys	C75H115N27O14S2	1683.03 / 1683.35	26.30	NA
ArB18Ch	Arg-Arg-Phe-Cys-Arg-Tyr-Trp-Arg-Arg-Leu-Cys	C78H124N28O12S2	1710.14 / 1710.72	23.74	NA
ArB18Dh	Aib-Arg-Phe-Cys-Arg-Tyr-Trp-Arg-Arg-Leu-Cys	C76H119N25O12S2	1639.06 / 1639.25	25.31	NA
ArB18Er	Ac-Thr-Ser-Phe-Cys-Glu-Tyr-Trp-Ala-Leu-Leu-Cys	C64H89N13O17S2	1376.60 / 1377.82	49.03	NA
ArB18Ed	Ac-Thr-Ser-Phe-Cys-Glu-Tyr-Trp-Ala-Leu-Leu-Cys	C70H97N15O19S2	1516.74 / 1516.20	42.72	>250
ArB18Ec	Ac-Thr-Ser-Phe-Cys-Glu-Tyr-Trp-Ala-Leu-Leu-Cys	C78H99N13O18S2	1570.84 / 1571.21	55.34	NA
ArB18Es	Ac-Thr-Ser-Phe-Cys-Glu-Tyr-Trp-Ala-Leu-Leu-Cys	C68H95N13O19S3	1494.75 / 1494.67	47.06	>250
ArB18Ek	Ac-Thr-Ser-Phe-Cys-Glu-Tyr-Trp-Ala-Leu-Leu-Lys	C69H96N14O18S	1441.66 / 1442.86	43.50	74.8
ArB18Eh	Ac-Thr-Ser-Phe-Cys-Glu-Tyr-Trp-Ala-Leu-Leu-Cys	C71H101N13O17S2	1472.77 / 1472.99	53.28	NA
ArBMim	[-Bmpa-bhPro-Arg-Ala-Trp-Arg-Lys-Inp-CH2CH2CH2CH2-S-]2	C114H172N32O16S2	2310.94 / 2311.61	33.70*	23.4±2.1
BLM					0.3±0.6
DTX					74.5±3.7

Abbreviations: Aib-Aminoisobutyric acid; AZT-azidothymidine; BLM-Bleomycin A5; Bmpa-4-(Bromomethyl)phenylacetic acid; Cha-Cyclohexyl-(L)-alanine; CLB-Chlorambucil; Chol-Cholesterol; DISC -(R,S)-1,3-dihydro-2H-isoindole carboxylic acid; DTX-Docetaxel; Inp-isonipecotic acid; MTX-Methotrexate; Nic-Nicotinic acid; Na1-1-Naphthyl-(L)-alanine; Na2-2-Naphthyl-(L)-alanine; PF5-Pentafluoro-(L)-phenylalanine; bhPro –β-homo-(L)-proline; Pipz-Piperazine monoamide; PF3-p-Trifluoromethyl-(L)-phenylalanine; Tfa-Tfrifluoroacetic acid. D-amino acids are underlined cursive e.g. Cys.

All peptides were synthesized as C-terminal amides except for ArB14Ao – that was synthesized as C-terminal piperazine monoamide (Pipz). An analytical RP-HPLC was performed with reversed-phase C₁₈ SymmetryShield™ column, 4.6×250 mm, 5 µm (Waters, Milford, MA) or (*) an reversed-phase C₁₈ Vydac 218TP54 column, 4.6×250 mm, 5 µm (Grace, Deerfield, IL). Small letter at the end of peptide’s name indicates that peptide was bridged using: (b) 1-(bromomethyl)-3-[3-(bromomethyl)benzyl]benzene; (c) 1-(chloromethyl)-4-[4-(chloromethyl)phenoxy]benzene; (d) 2-chloro-N-(2-(2-chloro-acetylamino)-ethyl)-acetamide; (g) 1,4-dibromobutane; (h) 1,7-dibromoheptane; (m) 1,3-bis(bromomethyl)benzene; (o) 1,2-bis(bromomethyl)benzene; (p) 1,4-bis(bromomethyl)benzene; (r) reduced/not stapled peptide; (s) divinylsulfone (DVS); (t) peptide was stapled and dimerized using 1,2,4,5-tetrakis(bromomethyl)benzene; (x) thioether groups were oxidized to sulfones with OXONE^®. For peptides ArB14k1÷k4 and ArB18Ek intramolecular bridge was achieved via side chain of Lys that was chloroacetylated and subsequently reacted with side chain of Cys forming thioether linker.

Fig. 2.

Cysteine(s) based staples used in the synthesis of ArB-analogues

Generally, all analogues were synthesized as C-terminal amides (except ArB14Ao) by the solid phase method using CEM Liberty automatic microwave peptide synthesizer (CEM Corporation Inc., Matthews, NC), applying 9-fluorenylmethyl-oxycarbonyl (Fmoc) chemistry (Fields and Noble, 1990) and standard, commercially available amino acid derivatives and reagents (EMD Biosciences, San Diego, CA and Chem-Impex International, Inc., Wood Dale, IL). Non-stapled peptides were purified by preparative reverse-phase high performance liquid chromatography (RP-HPLC) and their purity evaluated by matrix-assisted laser desorption ionization spectrometry (MALDI-MS) as well as analytical RP-HPLC (see Table 1 and Fig. 3).

Fig. 3.

The representative analytical RP-HPLC profile and corresponding MALDI-MS spectra obtained for analogue ArB14Cs

The stapling/S-alkylation reactions were carried out in solution using various protocols (for details see Experimental Section) using both, water-based solutions and “exclusively” organic solvents. Generally, for S-alkylation of synthetic linear peptides using bis-halogenomethyl-aryl derivatives and 2-chloro-N-(2-(2-chloro-acetylamino)-ethyl)-acetamide, we employed the protocol described by Timmerman and co-workers (Timmerman et al., 2005). Those reactions were carried out at ambient temperature in a solution of 50 mM ammonium bicarbonate (NH₄HCO₃) dissolved in a mixture of acetonitrile (ACN) and water. The synthesis of mono-S-alkylated compounds (ArB14Ck1-k4) was achieved using the same protocol. The stapling with divinylsulfone (DVS) was performed using a previously described protocol (Wilkinson et al., 2007) with conditions very similar to NH₄HCO₃/ACN method, with the only difference being the NH₄HCO₃ concentration (10 mM). The highest efficiency of stapling reaction(s) was observed for divinylsulfone (DVS), 2-chloro-N-(2-(2-chloro-acetylamino)-ethyl)-acetamide and 1,2- and 1,3-bis(bromomethyl)benzene bridged analogues. Simultaneously, 1-(bromomethyl)-3-[3-(bromomethyl)benzyl]benzene and 1-(chloromethyl)-4-[4-(chloromethyl)phenoxy]benzene staples were created with particularly low efficiency (yield <3%). Therefore in these cases (ArB15Eb, ArB17Ac and ArB18Bc) we also attempted to synthesize desired analogues using DMF/DMSO/Cs₂CO₃/TBAI system that we recently described (Micewicz et al., 2014). Unfortunately, this method did not improve synthetic efficiency. Notably, stapling with ortho- and meta-bis(bromomethyl)benzene can be also performed on resin using analogous protocol (see Experimental Section for details). The poor synthetic yields were also attributed to the stapling performed with 1,4-dibromobutane and 1,7-dibromoheptane that was carried out using a previously described 1,1,3,3-tetramethylguanidine (TMG) driven alkylation of thiol(s) in organic solvents that we adapted to peptides (Chua et al., 2015). Synthesis of oxidized analogues ArB14Dmx and ArB14Emx was performed using OXONE^® (Duran et al., 2006) (Sigma-Aldrich, KHSO₅×0.5 KHSO₄×0.5 K₂SO₄) in a 75% solution of methyl alcohol in water employing 3 eq. of OXONE^® per each thioether group. The reaction was carried out in room temperature for 3 h giving desired sulfones‘ containing products. For the analogues ArB14Co and ArB14Cs we additionally synthesized conjugates with known anticancer compounds, namely azidothymidine (AZT), bleomycin A₅ (BLM), chlorambucil (CLB), docetaxel (DTX) and methotrexate (MTX) (see Table 1 and Fig. 4) to ascertain possible synergy effects. AZT conjugate was synthesized using Cu(I)-catalyzed azide-alkyne 1,3-dipolar Huisgen's cycloaddition (CuAAC, “click chemistry”). CLB and MTX were attached as N-terminal moieties directly on resin during SPPS. In the case of BLM and DTX conjugates, we employed direct conjugation of proper unprotected ArB peptides with the desired anticancer entity using bis-amine reactive crosslinking reagent, dithiobis(succinimidyl propionate) (DSP, Pierce Biotechnology, Rockford, lL) in DMSO solution (5 mg/ml, 18 h).

Fig. 4.

Structures of selected conjugates of ArBs prepared for this study

An initial evaluation of bioactivity of our compounds was carried out in vitro using growth inhibition assay (PrestoBlue™, Invitrogen, Carlsbad, CA) and HCT-116 colorectal cancer cell line that is sensitive to MDM2 inhibitor(s) treatment (Hu et al., 2007; Tovar et al., 2006; Vassilev et al., 2004). In our view, this simple approach provides more reliable data which take into account many factors like the compounds’ cell permeability, stability, binding potency, etc., and in this particular case is better suited for the screening than pure biophysical method(s) e.g. measurements of binding affinity, especially in the light of reported discrepancies between physicochemical/in vitro data and in vivo activity (Chang et al., 2013; Liu et al., 2010b; Liu et al., 2010a). Notably, we had previously employed this approach in the synthesis of drug candidates targeting other intracellular targets (Smac) with positive results (Micewicz et al., 2014). Examples of cell growth curves are presented in Fig. 5 and the in vitro screening data are summarized in Table 1. The most active compounds were found to be i,i+3 stapled analogues (ArB14Cs, ArB14Co, and ArB14Bo) containing a polyarginine motif (Fig. 6) with staple created either by 1,2-bis(bromomethyl)benzene (ArB14Co and ArB14Bo) or DVS (ArB14Cs) between (D)Cys in position 5 and (L)Cys in position 8. Notably, ArB14Cs seems to be ~2× more active than ArB14Co/ArB14Bo counterparts suggesting that the DVS-staple produces a sterically less constrained compound than the ortho-substituted-thio-methyl-benzene-linker. Moreover, the DVS staple is structurally/electrostatically similar to the Glu side chain that is present in this particular position in the parental compound (PMI-N8A), which has previously been reported important for overall activity (Li et al., 2010a). Position 3 seems to be also quite permissive as the activity observed for the pentafluoro-phenylalanine-containing analogue (ArB14Co) is almost identical to that with a 2-naphthyl-alanine residue in the same position (ArB14Bo). The polyArg motif consists of 5 arginine residues in positions 1, 2, 4, 9 and 11, with positions 1 and 9 being occupied by (D)Arg residues. Introduction of (D)Arg enantiomers in these particular positions allowed for formation of a polycationic face for the molecule, which was reported as necessary feature for efficient cell entry (Futaki et al., 2002; Wender et al., 2008), while simultaneously providing additional resistance to proteolytic cleavage afforded by unnatural/(D) amino acids. Notably, all ArB analogues lacking the polyArg motif showed poor anticancer properties. Modifications in the crucial hydrophobic triad: ¹⁹Phe→Phe^F5/2-Nal, ²³Trp→1-Nal, and ²⁶Leu→Cha were prompted by reports that certain modifications enhancing hydrophobicity may result in analogues with enhanced properties (Liu et al., 2010b; Zhan et al., 2012), with an additional benefit being enhanced resistance to oxidative degradation afforded by elimination from the molecule of the Trp residue which is prone to such process (Simat and Steinhart, 1998). Interestingly, in the case of ¹⁹Phe→Phe^CF3 substitution effects seem to be less pronounced than for ¹⁹Phe→Phe^F5 analogue, suggesting that the additional para-trifluoromethyl-group may be associated with limited steric hindrance, which is contrary to previously reported results (Zhan et al., 2012), although such conclusion may be partially incorrect since compared peptide sequences are not identical. The observed in vitro anticancer effects were sequence specific and seem to be independent on polArg toxicity as various other polycationic sequences synthesized for this study showed limited potency (e.g. ArB14Dm, ArB15Em, ArB15Fo, ArB15FNico, etc.). In certain cases, conjugation with known anticancer entities seems to be beneficial as well. Specifically, BLM-conjugates had improved anticancer activity (ArB14Cs-BLM versus ArB14Cs; ArB14Co-BLM versus ArB14Co). A similar, but less pronounced, trend was also observed for MTX- and CLB-conjugated counterparts.

Fig. 5.

Cell viability data for selected compounds. Compounds possess following EC₅₀ values (µM): ArB14Co:11.2±1.2; ArB14Bo:10.7±0.7; ArB14Co-CLB:7.5±1.1; ArB14Co-MTX:6.9±0.4; ArB14Co-AZT:30.2±2.5

Fig. 6.

The most active ArB analogues

To ascertain whether our ArB compounds possess therapeutic potential in vivo, we performed studies in a murine subcutaneous tumour engraftment model (HCT-116/SCID). Based on the in vitro results and ease of synthesis/availability, we selected stapled peptide ArB14Co as the drug candidate for in vivo experiments. The treatment of cancer-bearing animals with this analogue resulted in potent anticancer effects (Fig. 7). Animals treated with 10 doses of the compound at a concentration of 3.0 mg/kg showed ~14.6 days delay in tumour growth. Interestingly, 2.5× higher dose (7.5 mg/kg) showed similar efficacy. Since all reported to date in vivo data for p53–MDM2/MDMX peptide inhibitors utilized different murine models, direct comparison between them should be rather cautious due to inherent differences in tumour growth rates, physiology, etc. Nonetheless, comparison with recently reported results for ATSP-7041 (Chang et al., 2013) (Aileron Therapeutics, Inc.) seems to be very favorable as therapeutic dosage for ATSP-7041 was in much higher range (15–30 mg/kg). Interestingly, in an MDM2-amplified osteosarcoma xenograft model (SJSA-1) anticancer activity of ATSP-7041, like ArB14Co, showed no dose response as the in vivo activitiy for 15 mg/kg and 30 mg/kg doses was basically the same, providing roughly ~4–5 days of tumour growth delay, which is ~3× less than for ArB14Co administered at significantly lower dose (3.0 mg/kg). Activity of ATSP-7041 in the MDMX overexpressing MCF-7 human breast cancer xenograft model is more promising, however the dosage utilized is still high (20–30 mg/kg). Comparison of ArB14Co with targeted liposomal formulation of ^DPMI-α (Liu et al., 2010a) (human glioma cell line U87/nude BALB/c model) is somewhat difficult as presented tumour volumes are reported in relative scale and tumour growth delay values are not reported, Nonetheless, presented results suggest significant activity, at least at 7.5 mg/kg dose. However, in this case ArB14Co may possess an advantage as well since it does not require liposomal delivery system. Notably, naked ^DPMI-α was virtually inactive in reported in vitro experimets in concentrations as high as 100 µM.

Fig. 7.

In vivo results from subcutaneous engraftment mouse model (HCT-116/SCID). Administration of ArB14Co results in ~14.6 days of tumour growth delay

Conclusions

A new family of anticancer, cell-permeating, stapled peptides was synthesized, characterized and screened for anticancer activity against human colorectal cancer cell line, HCT-116. Selected analogue, ArB14Co was tested in vivo showing potent anticancer activity at the low dose (3.0 mg/kg). Utilized stapling/modification strategy yielding short, cell permeating analogues, containing polycationic motif woven into parental sequence of peptide(s) seems to be a suitable approach in the development of p53–MDM2/MDMX bridged peptide inhibitors.