Optimal Stapling of a Helical Peptide‐Foldamer Hybrid Using a C‐Terminal 4‐Mercaptoproline Enhances Protein Surface Recognition and Cellular Activity

Abstract

Structural analysis of a co‐crystal of a helically‐folded peptide‐foldamer hybrid in complex with hDM2 E3 ubiquitin ligase, revealed a unique orientation for the C‐terminal proline with the pyrrolidine ring pointing backwards in the sequence, and suggested new opportunities for macrocyclization. In particular, we found that the C‐terminal prolyl residue could be replaced by its (2S,4S)‐4‐mercaptoprolyl analogue for optimal bisthioether crosslinking with a cysteine residue installed at position 4 in the sequence. The resulting i,i+7 stapled peptide‐foldamer is a high‐affinity binder to hDM2, is cell permeable and restores the p53 signalling pathway in p53wt cancer cells. The co‐crystal structure of hDM2 and the stapled peptide‐foldamer hybrid was determined at 1.84 Å, fully validating the original design and further highlighting the potential of cis‐4‐mercaptoproline in the context of peptide and foldamer stapling.

Reorientation of the C‐terminal proline in a peptide‐oligourea hybrid foldamer was exploited to generate macrocyclic helix mimetics through Pro→cis‐mercapto‐proline (4MP) replacement and subsequent bis‐thioether crosslinking. Applied to hDM2‐p53 interaction inhibition, this approach led to cell permeable, high‐affinity binders and was validated by a crystal structure of the complex.

Introduction

Peptide macrocyclization and backbone modification (e. g. foldamers) have emerged as valuable techniques for reducing the size of protein domains and designing peptide analogues endowed with enhanced properties, such as increased resistance to proteolytic degradation.[ ¹ , ² , ³ , ⁴ , ⁵ , ⁶ ] Surprisingly, the combination of the two approaches to amplify molecular diversity and properties has been reported only in a few cases. Noteworthy examples include cell‐permeable stapled α/β‐peptide analogues of the Bim BH3 domain from the Gellman group, ^[7] stapled β‐peptide inhibitors of the p53‐hDM2 interaction from the Schepartz group, ^[8] and stapled D‐sulfono‐γ‐AApeptides mimicking ubiquitin‐conjugating enzyme E2 from the Cai group. ^[9] An intriguing and recent case of such a combination was reported by the Suga and Huc groups. ^[10] It involves a macrocyclic foldamer‐peptide system, in which an aromatic oligoamide foldamer acts as a giant staple, stabilizing the helical conformation of the α‐peptide segment. Our recent effort to integrate foldamer chemistry and macrocyclization has focused on oligourea‐peptide foldamers as p53 helix mimics to inhibit the interaction between hDM2 (and hDMX) and p53 (Figure 1A). These oligomers which contain a three‐urea motif at the C‐terminus of a peptide segment (e. g. 1 a) ^[11] were cyclized by introducing a bisthioether crosslink ^[12] between a cysteine residue in the peptide segment and the thiol side chain of the cognate monomer (Cys^u) in the oligourea segment. ^[13] Of note, compounds in this series with a neutral overall charge, such as 2, were found to enter the cytoplasm and to reactivate the p53 signalling pathway. Moreover, despite the polarity likely introduced by the urea moieties, compound 2 nonetheless exhibited a concentration‐dependent effect on the viability of p53 wild‐type (p53wt) cancer cells (e. g. HCT116, human colon cancer cell line). In the present work, while evaluating sequences with different patterns of α‐amino acid and urea residues, we identified new opportunities in terms of macrocyclization for optimal binding to the surface of the protein of interest. More specifically, by looking at the co‐crystal structure of hDM2 in complex with 3, a peptide‐based ligand containing two urea units followed by a C‐terminal prolyl residue, we identified a particular orientation of the prolyl residue induced by the diurea insert as compared to the native p53 peptide sequence (Figure 1B and 1C), making it ideally positioned for further macrocyclization (Figure 1D). Herein, we considered the replacement of the C‐terminal proline by the corresponding (2S,4S)‐4‐mercaptoproline (cis‐4MP) residue and its use in macrocyclization strategies for bisthioether crosslinking to further stabilize the bound helical conformation (Figure 1E). To our knowledge, only a very limited number of studies have considered the use of 4MP as a constrained residue for the cyclization of peptides through either disulfide bond formation or bisthioether crosslinking.[ ¹⁴ , ¹⁵ , ¹⁶ ] It is noteworthy that the (2S,4R)‐4‐mercaptoproline (trans‐4MP) was recently introduced by the Kritzer group for the formation of stapled β‐Hairpins. ^[16] In the present study, the resulting peptide‐foldamer macrocycles incorporating 4MP were investigated for their ability to mimic the α‐helical conformation of p53 peptide and inhibit the hDM2‐p53 interaction, enter the cytoplasm, and reactivate the p53 signalling pathway. We also used X‐ray crystallography to gain precise information on the conformation of the stapled peptide‐foldamer inhibitor bound to hDM2 and validate the design strategy.

Figure 1.

A. Formulae of previously described linear and macrocyclic hDM2 binders 1 a, 1 b and 2 incorporating a oligourea backbone at the C‐terminus.[ ¹¹ , ¹³ ] B. Formula of the peptide‐oligourea foldamer 3 that was used as a starting point in this study. C. Co‐crystal structure of hDM2 with 3 (PDB ID 9FQL) and comparison of the ligand bound conformation to that of the natural p53 sequence bound to hDM2 (PDB ID 1YCR). D. The structure reveals a specific orientation of the proline residue in bound 3 that would be compatible with i,i+3 and i,i+7 macrocyclization patterns (violet dashed lines for visual rendering). E. General approach described in this work whereby the C‐terminal Pro in 3 is replaced by a cis‐mercaptoprolyl residue (cis‐4MP) to allow bisthioether macrocyclization using a series of three different bis‐electrophiles a–c.

Results and Discussion

We have previously shown that N,N’‐linked oligoureas adopt a helical structure akin to the α‐helix and can be interfaced with α‐peptides to create hybrid oligourea‐peptide sequences to target protein surfaces.[ ¹¹ , ¹⁷ , ¹⁸ , ¹⁹ ] High‐affinity binders for hDM2 such as 1 a and 1 b (Figure 1A) were obtained by replacing the C‐terminal residues of the wild‐type p53 sequence or a known hDM2 peptide ligand (PMI) by short tri‐urea inserts.[ ¹¹ , ¹³ ] In an extension of this work, we have prepared and screened linear hybrid sequences containing a shorter di‐urea segment, for their ability to bind hDM2 and its homolog hDMX (or hDM4). We next assessed the ability of the compounds to inhibit the interaction between wild‐type p53 and hDM2 or hDMX using a time‐resolved fluorescence energy transfer (TR‐FRET) assay. Compound 3 which contains a C‐terminal prolyl residue after the diurea insert was found to inhibit these interactions with IC₅₀ values of 54±17 nM and 247±42 nM for hDM2 and hDMX, respectively (Figure 1B, Table 1 and Figure S3). To gain additional insight into the binding mode of this compound, we solved its co‐crystal structure with hDM2 at 2.0 Å resolution (PDB ID 9FQL, Figure 1C and Table S1). The structure shows that the inhibitor binds hDM2 in a helical conformation that matches well the binding mode of the p53wt transcription activation domain (TAD) peptide (PDB ID 1YCR). ^[20] However, a comparison of the two structures reveals differences at the C‐termini, with a distinct orientation of prolyl residues in the two peptides. While the prolyl residue in the native p53 sequence (Pro27) terminates the helix and positions the remaining two residues in an extended backbone conformation, the prolyl residue in 3 remains connected to the helix through a H‐bond involving the urea between Ala^u10 and Pro11 and the carbonyl of Ala8 (Figure 1C). This arrangement causes the pyrrolidine ring to point backward, aligning parallel to the helix axis. This observation led us to consider the potential exploitation of the pyrrolidine ring orientation by introducing a function at the C4 position to create a staple with another residue in the sequence. As an illustration, the distances between the C4 atom of the pyrrolidine ring and the methyl groups of Ala8 (4.7 Å) and Ala4 (8.5 Å) seem compatible with the formation of i,i+3 and i,i+7 staples (Figure 1D). Moreover, the structural analysis indicates that the methyl side chain of the urea residue at position 10 (Ala^u10) points towards the hydrophobic cavity and suggests that a longer hydrophobic side chain would more effectively occupy the cavity. Consequently, we first synthesized a series of four sequences (compounds 4–7) with alternative side chains at position 10 (iBu, iPr, sBu, nBu) prior to macrocyclization exploration. The corresponding activated urea monomers N₃‐Xaa^u‐OSu were prepared and used for the solid‐phase synthesis of the corresponding diurea‐peptide hybrids as previously described.[ ²¹ , ²² ] The four oligomers were recovered in 30–45 % yield after HPLC purification. The compounds were screened for their ability to inhibit the p53‐hDM2 and p53‐hDMX interactions at 100 nM using the TR‐FRET assay. All tested compounds exhibited greater potency than 3 in this assay (Table 1 and Figure S1). Compound 4 with IC₅₀ values of 6.13±0.08 nM (hDM2) and 115±5 nM (hDMX) in the TR‐FRET assay (Table 1 and Figure S4), was selected as the reference compound for subsequent macrocyclization studies. To generate the stapled foldamers, we considered bisthioether crosslinking and the replacement of Pro by the corresponding (2S,4S)‐4MP derivative. The suitably protected cis‐4MP derivative, Fmoc‐4MP(Mmt)‐OH (Mmt=monomethoxy trityl) was prepared in 8 steps and an overall yield of 17 % from Boc‐L‐4‐trans‐Hyp‐OH following a previously reported route (Scheme S1).[ ²³ , ²⁴ ] We next prepared three analogues of compound 4, by replacing Pro11 by cis‐4MP and either Ala4 (8), or Ala8 (9) by a Cys residue for bisthioether crosslinking (Table 1). The linear peptide‐diurea hybrid precursors 8 and 9 were synthesized on solid support and were obtained in 33–34 % yield after HPLC purification. We then performed dithiol alkylation in solution using three different electrophiles (a–c) to generate the corresponding macrocycles 8 a, 8 b, 9 a and 9 c (Figure 2). The cyclic oligomers were obtained in an overall yield ranging from 5–20 % after the final purification by C₁₈ RP‐HPLC (Table 1).

Table 1.

Linear and macrocyclic peptide‐foldamer hybrids described in this work and their properties.

Cmpd	Sequence												Linker	MW (g.mol−1)	t R [a] (min)	Yield (%)	Inhibition at 100 nM (%)		IC50±SEM[b] (nM)
	N‐ter	1	2	3	4	5	6	7	8	9	10	11					hDM2	hDMX	hDM2	hDMX
1 b [13]	Ac	Lys	Thr	Phe	Ala	Glu	Tyr	Trp	Alau	Alau	Valu	–	–						5.7±0.2	23.8±0.9
2 [13]	Ac	Lys	Thr	Phe	Cys*	Glu	Tyr	Trp	Cysu*	Alau	Valu	–	a						9.2±0.3	73±6
3	H	Thr	Ser	Phe	Ala	Glu	Tyr	Trp	Ala	Leuu	Alau	Pro	–	1312.49	5.41	52	43	19	54±17	247±42
4	H	Thr	Ser	Phe	Ala	Glu	Tyr	Trp	Ala	Leuu	Leuu	Pro	–	1354.58	5.12	45	99	46	6.13±0.08	115±5
5	H	Thr	Ser	Phe	Ala	Glu	Tyr	Trp	Ala	Leuu	Ileu	Pro	–	1354.58	4.81	45	99	54	–	–
6	H	Thr	Ser	Phe	Ala	Glu	Tyr	Trp	Ala	Leuu	Valu	Pro	–	1340.55	4.98	30	99	54	–	–
7	H	Thr	Ser	Phe	Ala	Glu	Tyr	Trp	Ala	Leuu	Nleu	Pro	–	1354.58	4.80	45	99	33	–	–
8 a	H	Thr	Ser	Phe	Cys*	Glu	Tyr	Trp	Ala	Leuu	Leuu	4MP*	a	1500.84	6.15	20	99	66	1.91±0.05	75±4
8 b	H	Thr	Ser	Phe	Cys*	Glu	Tyr	Trp	Ala	Leuu	Leuu	4MP*	b	1558.84	4.94	16	99	43	2.7±0.1	119±31
9 a	H	Thr	Ser	Phe	Ala	Glu	Tyr	Trp	Cys*	Leuu	Leuu	4MP*	a	1500.84	6.36	5	70	−12	–	–
9 c	H	Thr	Ser	Phe	Ala	Glu	Tyr	Trp	Cys*	Leuu	Leuu	4MP*	c	1458.76	5.73	8	97	2	–	–
10 a	Ac	Lys	Thr	Phe	Cys*	Glu	Tyr	Trp	Ala	Leuu	Leuu	4MP*	a	1583.98	6.00	17	–	–	3.0±0.1	63±12
11 a	Ct	Lys	Thr	Phe	Cys*	Glu	Tyr	Trp	Ala	Leuu	Leuu	4MP*	a	1847.74	7.01	14	–	–	–	–

[a] HPLC: H₂O+0.1 % TFA, ACN+0.1 % TFA, gradient 10–100 % at 25 °C, 10 min, λ=200 nm. [b] Concentration–response curves (9 points) were performed in duplicates. Xaa*=bis‐thiol crosslinked. ct=chloroalkane tag. 4MP=4‐mercaptoproline.

Figure 2.

A. Dithiol alkylation of 9 to 9 a and the corresponding HPLC traces and MS analysis of crude oligomers following resin cleavage (9) and macrocyclization (9 a). B. Formulae of macrocyclic peptide‐foldamer hybrids 8 a, 8 b and 9 c. Xaa*=bis‐thiol crosslinked.

We next assessed their ability to inhibit p53‐hDM2 and p53‐hDMX interactions at 100 nM in the TR‐FRET assay (Table 1). Macrocycles with i,i+7 crosslinks (8 a and 8 b) were found to be the best hDM2 binders in this series as also confirmed by IC₅₀ measurements (Figure 3, Figures S5 and S6). With an IC₅₀ of 1.91±0.05 nM for inhibiting the p53‐hDM2 interaction, compound 8 a with a n‐hexyl crosslink was found to outperform the activity of cognate linear sequence 4 as well as Nutlin‐3a (IC₅₀=23±1 nM), a potent and well‐characterized hDM2 inhibitor. A similar trend was also observed for the p53‐hDMX interaction (IC₅₀=75±4 nM for 8 a) (Figure 3 and Table 1).

Figure 3.

Dose‐dependent inhibition of p53‐hDM2 and p53‐hDMX interaction in the TR‐FRET assay. Data are presented as mean±SD (n=2).

To gain additional insight into the binding mode of this high‐affinity binder for hDM2 and verify the accuracy of our design, we generated co‐crystals of hDM2 and 8 a and solved the structure of the complex at 1.84 Å (PDB ID 9GFK, Table S2). As shown in Figure 4, 8 a adopts a helical conformation that fully matches the bound conformation of 3 and the canonical binding mode of p53 peptide. Remarkably, the staple ideally connects Cys4 and cis‐4MP11, without introducing any substantial deformation to the helix. The n‐hexyl chain of the staple adopts an extended conformation and sits on the edge of the binding site, thus possibly contributing positively to the tight binding interaction through additional Van der Waals contacts. This aligns with earlier reports on i,i+7 stapled peptide inhibitors of hDM2, where crystal structures have similarly shown interactions between the elongated staple and the protein.[ ²⁵ , ²⁶ , ²⁷ ]

Figure 4.

Co‐crystal structure of 8 a bound to hDM2 (PDB ID 9GFK). A. Top‐view and B. Side view. C‐atoms of amino acid residues are colored in slate blue, C‐atoms of the two ureido‐residues are colored in orange and the C‐atoms of the linker, Ala and 4MP residues are colored in pink.

We next investigated the activity of the stapled diurea‐peptide 8 a in cells and its ability to restore the p53 signalling pathway. We have previously shown that increasing the overall positive charge of macrocyclic hDM2 ligands derived from 1, by using an N‐terminal (Arg)₃Lys extension, although increasing cell permeability led to compounds with intrinsic toxicity (measured by the release of Lactate dehydrogenase (LDH)) unable to reactivate p53 signalling. ^[13] Concurrently, we found that neutral (or negatively charged) compounds, N‐acetylated at the N‐terminus and containing a Lys‐Thr or Leu‐Thr termination, such as 2, were as effective to enter the cytoplasm and restored p53 signalling pathway. ^[13] The observation that some neutral N‐acetylated hDM2 peptide inhibitors may be superior to peptide ligands with a net positive charge has also been made during the development of Sulanemadlin (ALRN‐6924) and other potent stapled peptides targeting both hDM2 and hDMX.[ ²⁸ , ²⁹ ] We thus prepared compound 10 a (Figure 5A), an N‐acetylated analogue of 8 a. Compound 10 a demonstrated a high affinity for hDM2, with an IC₅₀ value of 3.0±0.1 nM, and retained binding to hDMX (IC₅₀=63±12 nM) compared to compound 8 a (see Table 1 for IC₅₀ values and Figure S7). Evidence that this neutral stapled foldamer sequence effectively penetrates the cytoplasm was obtained using the quantitative chloroalkane penetration assay (CAPA) based on the HaloTag system and developed by the Kritzer group.[ ³⁰ , ³¹ ]. A CP₅₀ of 0.5±0.1 μM was measured for 11 a, the analogue of 10 a bearing a chloroalkane tag at its N‐terminus (Figure 5B). This value is comparable to that of the chloroalkane derivative of compound 2 (CP₅₀=0.51±0.07 μM). ^[13] Compound 10 a was then evaluated on p53wt HCT116 cells that overexpress hDM2, for its ability to restore the p53 pathway using western blot analysis in comparison with Nutlin‐3a as a positive control. In this assay, treatment of cells with compound 10 a yielded a dose‐dependent increase of p53 levels (Figure 5C), similar to Nutlin‐3a. Activation of the p53 pathway in HCT116 cells was further supported by the dose‐dependent induction of mRNA of target genes p21 and hDM2 (Figure 5D). Finally, we found that treatment with 10 a was effective in reducing the viability of p53wt HCT116 cells in a concentration‐dependent manner (Figure 5E). In this assay, compound 10 a exhibited an activity similar to that of the linear triurea‐peptide hybrid 1 b but was less potent than the macrocyclic triurea foldamer 2, which we recently reported. ^[13]

Figure 5.

A. Compound 10 a and its chloroalkane tagged version 11 a. B. Compound 11 a tested by CAPA is cell‐permeable. CP₅₀: the concentration at which 50 % cell penetration is observed. Data are presented as mean±SD of biological replicates (n=2) vs. untreated control cells. The CP₅₀ value measured under the same conditions for the highly cell‐permeable control peptide Ct‐Trp‐NH₂ is 0.029±0.006 μM. ^[30] C. HCT116 cells were treated by 10 a during 6 hours and p53wt protein levels were assessed through western blotting. This Western blot is representative of three independent biological replicates (see also Figure S8). D. HCT116 cells were treated by 10 a during 6 hours and the determination of targeted genes p21 and hDM2 was evaluated by RT‐qPCR. Each data point represents a biological replicate, the error bar represents the SD of biological replicates (n=3) and * represents p‐values <0.05. E. Viability measurements using the MTS assay on HCT116 cells after 3 days of incubation with 10 a, 2 and 1 b. Results are expressed as mean±SD based on two independent experiments (with three technical replicates each) vs. untreated control cells.

Conclusions

This study demonstrates key advantages stemming from the introduction of a foldamer backbone into a peptide to generate high‐affinity binders for protein targets. Specifically, our work has delineated how the insertion of a short oligourea foldamer into a helical peptide can induce a favorable conformational rearrangement, thereby uncovering new possibilities for peptide macrocyclization. This was exemplified through a series of oligourea‐peptide hybrids designed to interact with hDM2. Our initial structural study of a co‐crystal of a helically‐folded linear peptide‐diurea foldamer hybrid in complex with hDM2 revealed a distinct orientation for the C‐terminal prolyl residue, compared to the native peptide. The strategic replacement of the C‐terminal urea‐linked prolyl residue by cis‐4MP in these sequences has led to the development of a highly effective i,i+7 bisthioether stapled peptide‐foldamer with high‐affinity for hDM2 and hDMX. The co‐crystal structure of this compound with hDM2 confirmed the original design and the helical bound conformation, with the hydrocarbon staple aligning with the protein binding site. This finding underscores the potential of 4MP in the realm of peptide and foldamer macrocyclization. In this respect, it also resonates with the recent work from the Kritzer group which utilized trans‐4MP for stabilizing β‐hairpin structure. ^[16] The newly cyclic hDM2 inhibitor reported here was found to exhibit cell permeability, and notably to restore the p53 signalling pathway in p53wt cancer cells. These findings open up promising avenues for the development of innovative peptide‐based therapeutic strategies targeting hDM2, which could potentially aid in the treatment of p53‐associated cancers.

Experimental Section

General

All steps were performed under inert atmosphere and microwave irradiation on the Liberty Blue or Discover Bio system (CEM). The temperature was maintained by modulation of power and controlled with a fiber optic sensor. All compounds were obtained with a final purity ≥95 % (see supporting information). The suitably protected cis‐4MP derivative, (2S,4S)‐Fmoc‐4MP(Mmt)‐OH was prepared from Boc‐L‐4‐trans‐Hyp‐OH following a previously reported procedure (Scheme S1). ^[23] The requested succinimidyl (2‐azidoethyl)carbamate monomers for oligourea synthesis (Figure S6) were prepared as previously described.[ ²¹ , ²² ]

General Solid‐Phase Synthesis Methods

Loading of the Resin

The Rink Amide resin (loading 0.51, 0.42 or 0.35 mmol/g) was swollen in DMF and the Fmoc protecting group was removed with 20 % piperidine in DMF (3 mL) under MW irradiation at 70 °C, 30 W for 5 min and the step was repeated once. The coupling of the first Fmoc‐protected amino acid to the resin was performed by adding the amino acid (1.5 equiv.), DIC (1.5 equiv.) and Oxyma (1.5 equiv.) in DMF and the reaction was performed at 75 °C, 30 W for 15 min. This loading step was repeated once.

Synthesis of the Oligourea Segment

The desired succinimidyl (2‐azidoethyl)carbamate monomers (N₃‐Xaa^U‐Osu, 1.5 equiv.) and DIEA (3 equiv.) in 2 mL DMF were reacted with the amino group of the growing chain on the resin under MW irradiation at 75 °C, 30 W for 20 min and the coupling step was repeated once. The resin was filtered, washed with DMF and with a mixture of 1,4‐dioxane/H₂O (7 : 3, v/v). Azide reduction was performed by adding PMe₃ (1 M in THF, 10 equiv.) in dioxane/H₂O (7 : 3, v/v) to the resin and the reaction was performed at 50 °C, 50 W for 15 min. The azide reduction step was repeated once. The resin was filtered, washed with a mixture of 1,4‐dioxane/H₂O (7 : 3, v/v, 1×2 mL) and DMF (4×2 mL). The coupling and reduction efficiencies were monitored with the chloranil test.

Synthesis of the Peptide Segment

Fmoc‐Xaa‐OH (5 equiv.), DIC (5 equiv.) and Oxyma (5 equiv.) in DMF were added to the resin and the reaction mixture was shaken at 75 °C, 70 W for 15 sec and then 90 °C, 35 W for 110 sec. Fmoc deprotection was performed by adding 20 % piperidine in DMF to the resin which was shaken at 75 °C, 155 W for 15 sec and then at 90 °C, 35 W for 50 sec.

Final Acetylation

After the removal of the last Fmoc protecting group, the resin was acetylated using a mixture of acetic anhydride (0.5 mL) and DIEA (10 equiv.) in 2 mL DMF at rt for 1 h. The linear sequence was cleaved from resin and side chains deprotected using TFA/TIS/H₂O/DODT (92.5 : 2.5 : 2.5 : 2.5, v/v/v/v) at rt for 2.5 h. The cleavage solution was evaporated and the residue was precipitated in cold ether. The obtained precipitate was dissolved in H₂O (CH₃CN was added if necessary) and lyophilized to give the crude linear oligourea‐peptide hybrid which was directly used for stapling (vide infra).

Coupling of the Chloroalkane Tag

The chloroalkane tag (ct‐COOH) was coupled to the resin following the removal of the last Fmoc protecting group. Ct‐COOH (2.5 equiv.), PyBOP (2.5 equiv.) and DIEA (5.0 equiv.) in DMF were added to the resin and the mixture was shaken at rt during 2 h.

General Procedure for Bisthioether Crosslinking

The crude linear sequence was dissolved in 1 : 1 (v/v) mixture of 0.1 M NH₄HCO₃ (pH=8) and CH₃CN (0.5 mg/mL) and TCEP (3–5 equiv.) was added. The solution was stirred at rt for 10–30 min, followed by addition of the desired bis‐electrophilic linker (10 equiv.). The mixture was stirred at rt and the reaction was followed by HPLC and LC–MS until the disappearance of the starting oligomer. Then, the reaction was quenched with an aqueous solution of 5 % TFA and lyophilized. The crude stapled foldamer was purified by RP‐HPLC using the appropriate gradient and lyophilized to give the desired compound.

X‐Ray Crystallography

Cocrystal of hDM2 with Ligand 3

Crystallization assays were performed with the hDM2 protein in complex with a 2‐fold excess of the oligourea‐peptide hybrid 3. Crystals were grown by sitting drop‐vapor diffusion at 293 K, mixing equal volumes (0.2 nL) of protein complex at 5.6 mg/mL and the reservoir solution, from a solution consisting of 1 M Na₂HPO₄/K₂HPO₄ pH 6.9. Data were collected on a single flash‐cooled crystals at 100 K in a cryoprotectant consisting of the mother liquor and 30 % glycerol on the PX1 beamline at the SOLEIL synchrotron). Data were processed with XDS. The structure was solved by molecular replacement with PHASER and refined with PHENIX and iterative model building in COOT.

Cocrystal of hDM2 with Ligand 8 a

Crystallization assays were performed with the hDM2 protein in complex with a 2‐fold excess of the stapled oligourea‐peptide 8 a. Crystals were grown by sitting drop‐vapor diffusion at 293 K, mixing equal volumes (1 μL) of protein complex at 5 mg/mL and the reservoir solution, from a solution consisting of 2.5 M Ammonium Sulfate, 100 mM MES pH 5.5. Data were collected on a single flash‐cooled crystals at 100 K in a cryoprotectant consisting of the mother liquor and 30 % glycerol on the PX1 beamline at the SOLEIL synchrotron). Data were processed with XDS. The structure was solved by molecular replacement with REFMAC and refined with BUSTER and iterative model building in COOT6.

HTRF Assay

Evaluation of the inhibitory activity of compounds on p53‐hDM2 interaction was determined by measuring the loss of interaction using a time resolved fluorescence energy transfer (TR‐FRET) detection method (Hybrigenics Services). The concentration of the peptides and analogues was determined by dosing UV absorbance of the indole side chain (λ=280 nm). hDM2‐GST protein (Hybrigenics Services; GST‐Flag‐89v5; batch #6) and p53‐His protein (Hybrigenics Services; YHX‐His‐1509v3; batch #7) were produced in E. coli and were purified by affinity columns. Fluorescence donor (anti‐GST antibody labeled with europium cryptate ‐Cisbio; ref 61 GSTKLA, batch 101A) was mixed with hDM2‐GST protein in a reaction buffer (PBS pH 7.4, DTT 2.5 mM, Tween 0.075 %+KF 400 mM). Fluorescence acceptor (anti‐His antibody labeled with terbium cryptate; Cisbio; ref 61 HISXLA; batch 61A) was mixed with p53‐His protein in the same reaction buffer. Solutions of proteins/antibodies were mixed in microplates at 25 nM (1 : 1 ratio) per well. The mixture was incubated overnight at 4 °C with reaction buffer (basal control), tested compounds or the reference inhibitor. A microplate reader PHERAstar (BMG) was used to measure the fluorescence transfer at λ_ex=337 nm and λ_em=620 nm and 665 nm. The ratio of the signal measured at 665 nm on the signal measured at 620 nm is used to determine the loss of p53‐hDM2 interaction using the following formula:

Where, “ratio” is the 665/620 fluorescence ratio, “sample” is the signal in presence of HTRF antibodies and “background” denotes the HTRF antibodies in buffer only. The inhibitors have been tested at several concentrations to generate a concentration‐response curve from which IC₅₀ value and SE are calculated by linear regression using GraphPad Prism.

Cell Culture

The human colon cancer cell lines p53wt HCT116 and Halo‐GFP‐Mito HeLa cells were cultured in DMEM (PAN BIOTECH). All media were supplemented with 10 % heat‐inactivated FCS and 1 % penicillin‐streptomycin. Human Colorectal Carcinoma HCT116 cells (CVCL_0291) were used at passages 47–50. The HCT116 cells were obtained from ATCC (CCL‐247). The cells were maintained under fully humidified conditions at 37 °C with 5 % CO₂. Cells were routinely tested for mycoplasma contamination by PCR.

CAPA

The Halo‐GFP‐Mito cells used for CAPA are HeLa cells stably expressing a fusion of HaloTag, GFP, and a mitochondria‐targeting peptide from the C‐terminal domain of the ActA protein from Listeria monocytogenes. ^[30] This fusion anchors HaloTag to the outer mitochondrial membrane, oriented in the cytosol. Cell populations expressing HGM were isolated to obtain individual colonies. CAPA was performed as previously described with slight modifications. ^[31] The day before starting the experiment, Halo‐GFP‐Mito HeLa cells were seeded in a 96‐well plate. To begin the experiment, media was aspirated and 100 μL of DMEM (Gibco) without serum was added. 25 μL of compounds were added at the desired concentration to the cells and were incubated at 37 °C for 4 hours. Supernatants were aspirated and cells were washed with Opti‐MEM (Gibco) for 15 minutes. Opti‐MEM was aspirated and 50 μL Opti‐MEM with 5 μM ct‐TAMRA dye was added to the cells for 15 minutes. The medium was aspirated and cells were washed again for 15 minutes with Opti‐MEM without phenol red. This washing step was repeated a second time. The supernatant was aspirated, cells were trypsinized using Trypsin 0.05 % – EDTA 0.1 % solution without phenol red and resuspended in 150 μL ice‐cold PBS(−Ca²⁺‐Mg²⁺)/FCS 50 %. 80 μL of cell suspension were typically analysed per well using BD LSRFortessa™ Cell Analyzer flow cytometer. Cells were gated, data were processed with FlowJo software and plotted using GraphPad Prism.

Western Blots

Immunoblot analysis was conducted as previously described. ^[32] Briefly, HCT116 cells in the exponential growth phase were seeded in 6‐well plates one day before the experiment. Subsequently, they were exposed to specified concentrations for 6 hours without serum and then lysed using 100 μL of radioimmunoprecipitation assay (RIPA) buffer. Denaturing polyacrylamide gel electrophoresis was employed, and the proteins were transferred onto nitrocellulose membranes using a wet transfer process. Following this, the membranes underwent a blocking step with 5 % BSA in Tris‐buffered saline containing 0.1 % Tween 20 (TBS‐T) for 1 hour at room temperature. Primary antibodies, such as anti‐p53 (Santa Cruz, sc‐71817) and anti‐GAPDH (ThermoScientific, MA5‐15738), were applied overnight at 4 °C. After washing with TBS‐T, the membranes were exposed to an anti‐mouse infrared dye secondary antibody (LI‐COR, 926‐32212) at room temperature for 1.5 hours. The signal was detected using a LI‐COR scanner

RNA Extraction and RT‐qPCR

Total RNA was isolated using TRIzol (MRC, TR 118), following the manufacturer's guidelines. Subsequently, RNA underwent reverse transcription using High‐Capacity cDNA Reverse Transcription Kit (Applied Biosystems, 4368814) and T100 Thermal Cycler (Bio‐Rad). This resulting cDNA was utilized for real‐time PCR with specific primers and SYBR Green PCR Master Mix (Takyon, UF‐RSMT‐B0701), as previously described. ^[33] Quantitative PCR data were collected using the AriaMx Real‐time PCR System (Agilent). Expression levels were normalized to GAPDH. The primers used in this study are p53 (Merck), hDM2 (Sigma‐Aldrich), p21 (Merck) and GAPDH (Eurogentec).

MTS Cell Viability Assay

HCT116 Cells (3×10³ cells/well) were seeded in 96‐well plates with the indicated inhibitors at various concentrations in serum‐free media for 8 hours, followed by addition of serum (10 %) and 64 hours of culture. Cell viability was assessed using the CellTiter 96® AQueous One Solution Cell Proliferation Assay (Promega), as previously described ^[34] and according to the manufacturer's instructions. At the end of the experiment, absorbance was measured at 490 nm using the Tecan Infinite 200 PRO (Life Sciences) plate reader. The percentage of viable cells was calculated by comparing treated cells to non‐treated cells. The data were plotted using GraphPad Prism.

Statistical Analysis

The statistical analyses were performed using Prism software (GraphPad). The Student t‐test was employed. A significance level of p‐values <0.05 was considered, and the experimental data were presented as the mean±standard deviation of the mean.

Conflict of Interests

The authors declare no conflict of interest.

Supporting information

As a service to our authors and readers, this journal provides supporting information supplied by the authors. Such materials are peer reviewed and may be re‐organized for online delivery, but are not copy‐edited or typeset. Technical support issues arising from supporting information (other than missing files) should be addressed to the authors.

Supporting Information

Neuville M., Bourgeais M., Buratto J., Saragaglia C., Li B., Galeano-Otero I., Mauran L., Varajao L., Goudreau S. R., Kauffmann B., Thinon E., Pasco M., Khatib A.-M., Guichard G., Chem. Eur. J. 2025, 31, e202403330. 10.1002/chem.202403330

Data Availability Statement

The structures of compounds 3 and 8 a in complex with hDM2 reported in this manuscript have been deposited in the PDB with accession codes : PDB ID 9FQL and 9GFK, respectively.