Synthetic α‐Helical Peptides as Potential Inhibitors of the ACE2 SARS‐CoV‐2 Interaction

Abstract During viral cell entry, the spike protein of SARS‐CoV‐2 binds to the α1‐helix motif of human angiotensin‐converting enzyme 2 (ACE2). Thus, alpha‐helical peptides mimicking this motif may serve as inhibitors of viral cell entry. For this purpose, we employed the rigidified diproline‐derived module ProM‐5 to induce α‐helicity in short peptide sequences inspired by the ACE2 α1‐helix. Starting with Ac‐QAKTFLDKFNHEAEDLFYQ‐NH2 as a relevant section of α1, a series of peptides, N‐capped with either Ac‐βHAsp‐[ProM‐5] or Ac‐βHAsp‐PP, were prepared and their α‐helicities were investigated. While ProM‐5 clearly showed a pronounced effect, an even increased degree of helicity (up to 63 %) was observed in sequences in which non‐binding amino acids were replaced by alanine. The binding affinities of the peptides towards the spike protein, as determined by means of microscale thermophoresis (MST), revealed only a subtle influence of the α‐helical content and, noteworthy, led to the identification of an Ac‐βHAsp‐PP‐capped peptide displaying a very strong binding affinity (KD=62 nM).


4
According to a procedure by V. Hack et al., [1] a solution of 14.5 g (111 mmol, 1.00 eq.) (S)-trans-4hydroxyprolin (3) in 110 ml methanol was cooled to 0 °C and 12.0 ml (166 mmol, 1.50 eq.) thionyl chloride were added dropwise. The solution was stirred at room temperature for 16 hours, after which the solvent was evaporated under reduced pressure. The resulting hydrochloride was suspended in 145 ml acetonitrile and cooled to 0 °C. 31.0 ml (222 mmol, 2.00 eq.) triethylamine and 26.6 g (122 mmol, 1.10 eq.) Boc2O were added, and the mixture was stirred at room temperature for 22 hours. After addition of 450 ml CH2Cl2, the layers were separated and the organic layer was washed three times with each 300 ml of 1.0 M aqueous KHSO4 solution, dried over MgSO4 and concentrated in vacuo to afford 26.9 g (110 mmol, 99%, Lit. [1] : 98%) of product 4 as a colorless oil. Rf

1-(tert-Butyl)-2-methyl-(2S,4R)-4-(tosyloxy)pyrrolidine-1,2-dicarboxylate (5) 4 5
According to a procedure by V. Hack et al., [1] to a solution of 25.5 g (104 mmol, 1.00 eq.) hydroxyproline 4 in 170 ml CH2Cl2 were added 43.5 ml (312 mmol, 3.0 eq.) triethylamine and the mixture was cooled to 0 °C. After addition of 43.6 g (229 mmol, 2.20 eq.) para-toluenesulfony chloride the mixture was stirred at room temperature for 20 hours. Once the reaction was determined complete via TLC, 120 ml sat. aqueous NaHCO3 were added, the layers were separated, and the aqueous layer extracted three times with each 120 ml CH2Cl2. The combined organic layers were dried over MgSO4, the solvent evaporated under reduced pressure and the residue was purified by column chromatography (SiO2, EtOAc/cHex 1:4) to afford 36.4 g (91.1 mmol, 88%, Lit. [1] : 88%) of product 5 as a yellow solid. According to a modified procedure by V. Hack et al., [1] to a solution of 11.9 g (25.0 mmol, 1.00 eq.) tosylate 5 in 36 ml anhydrous DMSO under inert atmosphere were added 4.38 g (77.6 mmol, 3.10 eq.) previously dried NaCN. The solution was heated to 55 °C and stirred for 72 hours, after which reaction control via NMR showed full conversion of the starting material. After addition of 20 ml of brine and 20 ml of water the mixture was extracted five times with each 50 ml MTBE. The combined organic layers were dried over MgSO4 and the solvent evaporated under reduced pressure to afford 3.67 g (14.4 mmol, 95%, Lit. [1] : 94%) of crude product 6 as a colorless solid, which was used in the next step without further purification.  To a solution of 4.90 g (19.3 mmol, 1.0 eq.) nitrile 6 in 39 ml of a 1:2 mixture of methanol and 50% aqueous acetic acid were added 11.3 g (96.3 mmol, 5.0 eq.) of a 50% suspension of Raney nickel in water. The reaction flask was flooded with hydrogen gas and the mixture vigorously stirred at room temperature for a total of 4 hours, while carefully monitoring the progress of the reaction via GC-MS. Upon complete conversion the hydrogen gas was removed, and the reaction mixture diluted with 30 ml water. The aqueous layer was extracted three times with each 50 ml ethyl acetate, the combined organic layers were dried over MgSO4, and the solvent evaporated under reduced pressure to obtain 3.53 g (13.7 mmol, 71%) of crude product 7 as a colorless oil as a mixture of isomers (10:1 cis/trans; determined via GC-MS), which was used in the next step without further purification. Rf = 0.21 (SiO2,

(2S,4R)-1-(tert-butoxycarbonyl)-4-vinylpyrrolidine-2-carboxylic acid (2) 8 1
To a solution of 1.04 g (4.07 mmol, 1.0 eq.) methyl ester 8 in 12 ml of a 2:1 mixture of THF and water, 179 mg (4.27 mmol, 1.05 eq.) Lithiumhydroxide monohydrate were added, and the mixture was stirred at room temperature for 18 hours. Once completion of the reaction was determined by TLC, the solution was diluted with 30 ml MTBE then acidified at by addition of 10 ml of 10% aqueous KHSO4 solution to pH = 3. The layers were separated, and the aqueous phase was extracted two times with each 30 ml MTBE. The combined organic layers were dried over MgSO4, and the solvent was evaporated under reduced pressure to afford 976 mg (4.05 mmol, >99%) of product 1 as a colorless solid as a mixture of isomers (

Boc[ProM-5]OtBu
A solution of 500 mg (1.19 mmol, 1.0 eq.) dipeptide 9, 68 mg (0.36 mmol, 0.30 eq.) copperiodide and 202 mg (0.238 mmol, 0.2 eq.) Grubbs II in dry CH2Cl2 under inert atmosphere was heated to 40 °C for 18 hours after which full conversion of the starting material was determined via TLC. After cooling to room temperature, three spatulas of each active charcoal and quadrasil were added and the solution filtrated over a celite pad which was washed with EtOAc. The filtrate was concentrated in vacuo and the residue was purified via column chromatography (SiO2, EtOAc/cHex 1:2). The resulting grey solid was dissolved in EtOAc and again 1 spatula of each active charcoal and quadrasil were added. The solution was filtered over celite once more and the solvent evaporated under reduced pressure to yield 407 mg (1.04 mmol, 87%, Lit.

Solid Phase Peptide Synthesis (SPPS)
All peptide sequences containing natural amino acids were prepared by a MultiSynTech Syro I automated peptide synthesizer via solid phase peptides synthesis. For this, 30 mg of rinkamide resin by Merck loaded with NovaPEG linker with a surface concentration of 0.48 mmol/g were used, resulting in potential 15 µmol of peptide chains. All automatized coupling reactions were performed in DMF as solvent using equimolar amounts (8.00 eq.) of Fmoc-protected amino acids, diisopropylcarbodiimide (DIC) and ethyl cyanohydroxyiminoacetate (oxyma). Side chain functional groups were protected using acid labile protecting groups. At the end of each coupling cycle a cleavage of the Fmoc protecting group was performed using 30% piperidine in DMF.
Manual coupling of N-terminal Fmoc-ProM-5-OH, Fmoc-L-Proline-OH or Fmoc-L-bHAsp-tBu-OH to peptide sequences synthesized as described above was performed by adding a solution of Fmocprotected amino acid (2.0 eq.), [1-[bis(dimethylamino)methylene]-1H-1,2,3-triazolo-[4,5-b]-pyridinium 3oxide hexafluorophosphate (HATU) (2.00 eq.) and NN-diisopropylethylamine (DIPEA) (2.00 eq.) in 300 µl DMF/CH2Cl2 (9:1) to the resin containing the respective peptide. The resin was shaken at room temperature for 2 hours, washed with DMF, CH2Cl2, MeOH and Et2O and dried. To determine coupling success, a Kaiser test was performed. One drop each of a solution of ninhydrin in ethanol (0.05 g/ml), a solution of phenol in ethanol solution (4 g/mL) and a solution of potassium cyanide in pyridine (0.02 mM) was added to the dried resin. The resin was then incubated at 95 °C for 5 min upon which a color change to blue indicates the presence of free amino functions and therefore an incomplete coupling. Upon complete coupling an Fmoc deprotecting was performed via stirring the resins in 30% piperidine in DMF, before the next coupling could be performed.
After a completed peptide synthesis, the free N-terminus was acetylatedby addition of 20 eq. acetic anhydride and 20 eq. DIPEA in 300 µl CH2Cl2 to the resin, followed by shaking for 30 min at room temperature and then washing with DMF, CH2Cl2, MeOH and Et2O.
The peptides were cleaved from the resin by addition of 1 ml a mixture of trifluoro acetic acid (TFA), triisopropylsilane and water (95:2.5:2.5) and shaking at room temperature for 3 hours. The resin was filtered from the solution and washed with 0.2 mL TFA and the combined filtrates were added to 10 ml cold Et2O and stored at -20 °C for 16 hours, upon which the peptides precipitated. They were then washed several times with cold Et2O through centrifugation, dissolved in tBuOH/water (1:4) and lyophilized. For purification, preparative RP-HPLC using a Hitachi Elite LaChrom system with a Macherey Nagel VP 250/8 Nucleodur 100-5 C18ec column was used. As solvents 0.1% aqueous TFA and acetonitrile were used with a linear gradient (30%à60% acetonitrile over 30 minutes) and a flow rate of 1.5 ml/min. The acetonitrile was removed from all relevant fractions using a Horizon Technology Xcel Vap in an air flow gradient from 880 to 1640 mbar in 20 minutes at 65 °C before lyophilization. The product identity was confirmed by LC-ESI-MS analysis using a Merck Chromolith Performance RP-18e endcapped 100-4.6 mm HPLC column coupled to a ThermoScientific LTQ-XL linear ion trap mass spectrometer (gradient: 20%à70% acetonitrile in 0.1% aqueous formic acid over 15 minutes). Purities were determined via integration of peaks in the UV chromatogram.

Characterization by Circular Dichroism (CD)
Circular dichroism spectra were recorded using a J715 Spectropolarimeter by JASCO at wavelengths from 180 to 260 nm in steps of 0.2 nm. For each spectrum 4 sets of measurements were recorded, and a baseline subtraction of the solvent was performed. A quartz cuvette with a thickness of 1 mm was used. Before measurement solutions of peptides in the respective measurement medium were prepared in concentrations of 60 µM. As medium 10 mM phosphate buffer at pH = 7.4 was used. The measured ellipticity was converted into the mean residue ellipticity for number of residues per molecule of 22 using the formula: Obtained spectra were smoothed using the web server-based plotting tool CAPITO [2] applying a Savitzky-Golay filter and plotted using the software Origin by OriginLabs. The raw data was analyzed by the online-tool DichroWeb [3] using the CDSSTR method to determine fractional α-helicities.                 Helical Wheels were drawn for helices resulting from bHAsp-PP coupled peptides (P-1-2, P-2-2, P-3-2) and reference peptides containing the corresponding parent amino acid sequences (P-1-3, P-2-3, P-3-3) to investigate the effect of the N-cap. The plots were produced using the online tool EMBOSS: pepwheel [5] with 18 steps and 5 turns. Polar or negatively charged residues are shown as red diamonds, positively charged residues as black octagons, residues with aliphatic side chains as blue squares and other hydrophobic residues in purple. β-Homoaspartic acid was replaced by glutamic acid for these plots.

Procedure for binding affinity measurements
For binding affinity measurements, His-tagged wildtype RBD (2019-nCoV) from SARS-CoV-2 (Sinobiological) Spike protein from recombinant expression in mammalian cells was kindly provided by Dr. Coskun (TU Dresden). The provided stock solution was diluted with buffer to 10 µM, rebuffered in sodium carbonate aqueous solution at pH 8.0, and incubated in the dark at 300 µM of the red dye included in the Protein Labelling Kit RED-NHS 2nd Generation from NanoTemper. Size exclusion chromatography allowed to obtain the labelled RBD free of unreacted dye. From this concentrated labelled RBD solution a stock of 10 nM was prepared using PBS + 0.05% Tween 20 as diluent.
Microscale thermophoresis (MST) was then measured by making 16 sequential 1:1 dilutions of each peptide, using PBS + 0.05 % (v/v) Tween 20 as diluent, each with a final volume of 10 µL. Each peptide dilution series were then shortly incubated with 10 µL of the 10 nM labeled RBD stock solution, therefore always keeping a 5 nM concentration of the labeled target protein in every sample. MST measurements were conducted at 22 °C with a Monolith NT.115 Pico instrument (NanoTemper Technologies), at an excitation power of 20% and a MST power of 40%, the signal was evaluated 1.5 seconds after start of the infrared laser. These conditions were kept constant for all samples. All measurements were performed at least by triplicate. Peptides P-1-1, P-1-2, P-1-3, P-2-3, P-3-3 did not give any significant response in the MST measurements and were considered as "non-binding". Accordingly no data can be provided in these cases.