Discovery of phosphotyrosine-binding oligopeptides with supramolecular target selectivity

We demonstrate phage-display screening on self-assembled ligands that enables the identification of oligopeptides that selectively bind dynamic supramolecular targets over their unassembled counterparts. The concept is demonstrated through panning of a phage-display oligopeptide library against supramolecular tyrosine-phosphate ligands using 9-fluorenylmethoxycarbonyl-phenylalanine-tyrosine-phosphate (Fmoc-FpY) micellar aggregates as targets. The 14 selected peptides showed no sequence consensus but were enriched in cationic and proline residues. The lead peptide, KVYFSIPWRVPM-NH2 (P7) was found to bind to the Fmoc-FpY ligand exclusively in its self-assembled state with KD = 74 ± 3 μM. Circular dichroism, NMR and molecular dynamics simulations revealed that the peptide interacts with Fmoc-FpY through the KVYF terminus and this binding event disrupts the assembled structure. In absence of the target micellar aggregate, P7 was further found to dynamically alternate between multiple conformations, with a preferred hairpin-like conformation that was shown to contribute to supramolecular ligand binding. Three identified phages presented appreciable binding, and two showed to catalyze the hydrolysis of a model para-nitro phenol phosphate substrate, with P7 demonstrating conformation-dependent activity with a modest kcat/KM = 4 ± 0.3 × 10−4 M−1 s−1.


MATERIALS
Fmoc-FpY with 95.0% purity was purchased from CSBio Co (CA, USA), the P7 peptide with 98.0% purity from China Peptides (China) and Caslo, the Ph.D.-12 phage display library, 2738 E.coli strain and M13 phage control from New England BioLabs (MA, USA), the primers from GenScript (NJ, USA) and the Chymotrypsin from Sigma Aldrich (NJ,USA).

Phage Display
The Ph.D.-12 phage display peptide library kit (New England BioLabs) was used to select peptide binders with supramolecular recognition and phosphatase activity. For the bio-panning, the phage library (10 11 pfu/mL) containing 10 9 phage different clones was mixed with 20 mM Fmoc-FpY in 200 µL of 100 mM phosphate buffer pH 8.0 and incubated for 48h at room temperature. The phage particles were recovered by centrifugation (18000xg, 15 min), and then washed five times with Tris Buffered saline -0.1% (v/v) Tween 20 (TBS-T) buffer. The Fmocpeptide aggregates associated with the phages was removed by digestion with 200 µL of 10 mg mL-1 Chymotrypsin (Sigma-Aldrich) 30 min at 37ºC. The eluted phages were then amplified by infection into E. coli strain ER2738, followed by precipitation with 20% PEG-2.5M NaCl. The amplified and purified phages (input) and the eluted phages (output) were tittered on LB plates containing tetracycline, X-gal and IPTG. Three rounds of biopannig were performed. The phage supernatant products from the 3 rd output plate was amplified and sequenced by GenScript.

Binding Assays by Colorimetric ELISA
The selected phage clones were amplified and then incubated with Fmoc-FpY at the same conditions of biopanning. The last wash of Fmoc-phage aggregates was performed in coating buffer 100 mM bicarbonate/carbonate pH 8.6, immobilized onto a 96-well microplate (Maxisorb, Nunc) and then incubated overnight at 4ºC. After washing twice with TBS, 300 µL of Blocking Buffer (1% BSA in TBS-T 0.05% v/v) was added in each well and incubated 1h at room temperature. The respective wells were then washed with TBS-T (0.05% v/v) (5x) and incubated with 200 µL anti-M13 antibody monoclonal conjugated with HRP (1:2000) during 1h at room temperature. The wells were again washed with blocking buffer (4x) and once with TBS-T (0.05% v/v). The ABTS substrate (100 µL) was then added to each well and the absorbance at 405 nm was monitored in a SpectraMax i3 (Molecular Devices). Wild-type M13 phage has been used as a positive control of ELISA assay and as a negative control of Fmoc-FpY binding. The experiments were performed in triplicate.

Kinetic Phosphatase Assays of the phage clones
The amplified phage clones and free peptide stock solutions (1mM and 0.5mM) were re-suspended in 100 mM phosphate buffer pH 8.0. The kinetic measurements were performed in a Jasco V-660 spectrophotometer monitoring absorbance of the product (p-nitrophenol) at 405 nm at 25ºC using a 1 cm quartz cuvette. p-Nitrophenol phosphate (pNPP) was used from a 100 mM stock solution to obtain a final concentration of 10 mM. A calibration curve was performed for different pNP concentrations ranging from 0 to 0.02 mM, for the quantification of pNP released. In case of kinetic assays with phage clones, a volume of 50 µL of phage solution previously amplified was added to 850 µL of PBS and 100 µL of pNPP stock solution. The initial velocity rate of pf each phage clone was obtained by using the data from the linear phase of pNP formed overtime (240h). The rate of catalysis was determined according to equation Rate (h -1 ) = [Product formed] (mM) / ([P7] (mM) x time (h)). The experiments were performed in triplicate.

Phage Amplification
The phage plaques obtained from the 3 rd round output plate were inoculated in 1-mL 2738 E.coli bacterial overnight culture diluted 1:100 and incubated at 37ºC at 220 rpm for 5h. The culture was then centrifuged at 12,000xg for 1 min. The supernatant was then removed to a fresh micro-centrifuge tube and re-centrifuged (12,000xg, 1min, 4°C). 80% of the supernatant was then transferred to a fresh tube and precipitated overnight with 150 µL of 20% PEG /2.5 M NaCl. The precipitated phage was then centrifuged at 12000xg and 4°C for 15 min and the supernatant was removed, and the pellet was re-centrifuged to remove residual supernatant. The pellet was re-suspended in 1 mL TBS. The solution was transferred to a new tube and spin at 12000xg, 4ºC for 5 min to pellet residual cells. The phage supernatant was precipitated with 200 µL of 20%PEG-2.5M NaCl and incubated on ice for 1h. The phage solution was centrifuged at 12000xg and 4°C for 15 min. The supernatant was again discarded, and the pellet was re-spin at 12000xg and 4°C for 5 min. The transparent pellet was re-suspended in 200 µL of 100 mM phosphate buffer pH 8.0.

TEM of Phage samples
TEM imaging was performed at the Advanced Science Research Center (ASRC), CUNY Imaging Facility using a FEI, TITAN Halo TEM operating at 300 Kv. Images were recorded in the low-dose mode (20 e − Å-2) on FEI CETA 16M camera (4,096 × 4,096 pixels). Carbon-coated grids were purchased from Electron Microscopy Sciences. The grids were firstly glow discharged in air for 30 s. 4.5 µL of sample was added to the grid and blotted down using filter paper after 60 s. The grids were then washed twice with distilled water and blotted with filter paper after 30 sec, to completely remove 100 mM sodium phosphate pH 8.0 buffer. For double negative staining, 4.5 µL of 2% aqueous uranyl acetate was applied twice (30 s each time) and the mixture blotted again using filter paper to remove excess. The dried grids coated with sample were then imaged.
The NMR experiments were acquired in a Bruker Avance III 600 spectrometer equipped with a triple-resonance cryoprobe (TCI). Spectra were processed using software TOPSPIN (Bruker Biospin, Karlsruhe, Germany) and peaks assigned with Sparky (TD Goddard and DG Kneller, Sparky 3, University of California, San Francisco, USA).

Diffusion ordered spectroscopy
Diffusion ordered spectroscopy (DOSY) spectra of P7 were acquired at 50 µM and 750 µM concentration using the pulse sequence from the Bruker library (ledbpgppr2s). The spectra were recorded with 2k scans for the lower concentration sample and 256 scans for the high concentration sample. The diffusion time was adjusted to 80 ms and the duration of the encoding/decoding gradient was calibrated to 4ms. The pulse gradient was increased from 2 to 95% of the maximum gradient strength using a linear ramp in 32 gradient steps. 32k data points in the F2 dimension were collected. DOSY experiments were acquired in a Bruker Avance III 400 spectrometer equipped with a triple resonance probe (TXI) and analyzed with TOPSPIN.

Circular Dichroism
The far-UV CD spectra (190-250 nm) was acquired with an Applied Photophysics Chirascan™ qCD spectrometer at 25ºC in a 1 × 1 mm square quartz cuvette under nitrogen flow, response of 3 s/nm and bandwidth of 1 nm. Each spectrum represents an average of three scans.

Molecular Dynamics Simulations
MD simulations of p7 monomer peptide were performed in GROMACS 2016.5 simulation package 1 using the AMBER99SB-ILDN force field 2 . The Lys and Arg residues were modelled in the protonated state, with a positively charged N-terminal and an amidated C-terminal. Starting from the extended conformation, the peptide was placed in a cubic box of 197 nm 3 . The box was then solvated with TIP3P [3] water molecules and 3 chloride counter-ions were added to neutralize the total charge with long-range electrostatics being treated with the Particle-Mesh Ewald algorithm 4 . The system was then energy-minimized in two steps to remove atom clashes and bond contacts: first by a steepest descent minimization algorithm (max 2000 steps), followed by a conjugated gradient algorithm (max 1000 steps). Equilibration of solvent molecules was made for 5 ns in a NPT ensemble, with the V-rescale thermostat at 300 K (time-constant of 1.6 ps), the isotropic Berendsen barostat at 1 bar (time-constant 5 ps) and with positional restrains for all peptide heavy-atoms with the LINCS algorithm with a force constant of 1000 kJ/mol/nm 2 . The integration step was 2 fs and coordinates were saved each 20 ps, Production runs in the NPT ensemble were made for 250 ns in 3 independent replicates in a total of 750 ns.
Cluster analysis of the 3 replicates was made using the method reported by Daura et al. 5 . The root means square RMSD matrix of 7501x7501 elements (equally spaced 100 ps frames) was calculated, with a minimum of 100 structures per cluster and a cut-off of 3 Å between backbone atoms. For each cluster, the centroid structure was obtained.

AFM Imaging
Samples were prepared by diluting 20 µl of the sample to a total volume of 100 µl of solution in deionized water.
Then it was pipetted on a freshly cleaved mica sheet (G250-2 Mica sheets 1″ × 1 ″ × 0.006″ (Agar Scientific Ltd)) attached to an AFM support stub and left to air dry overnight in a dust-free environment, prior to imaging. For AFM measurements, samples of the assembled peptides were diluted 7.5-fold in deionized water in a total volume of 100. 15 µL of each of the diluted sample was deposited on top of a freshly cut mica substrate (G250-2 Mica sheets 1″ × 1 ″ × 0.006″, Agar Scientific Ltd) and let air dry overnight in a dust-free environment, prior to imaging. The images were obtained by scanning the mica surface in air under ambient conditions using a Multimode 8 and a FastScan Microscope (Bruker) operated in tapping mode. The AFM scans were taken at a resolution of 512 × 512 pixels. The images were analysed using NanoScope Analysis software Version 1.40.

Determination of the Critical Aggregation Concentration (CAC)
A stock solution of 2.5mM pyrene in methanol was prepared and then diluted 20-fold in methanol. 50 μl of the diluted pyrene solution was added to each of the different concentrations of peptide P7. The concentrations used were 0.05, 0.25, 0.75, 1.25, 2.5 and 5mM 6 .These samples were excited at 334 nm in a Jasco FP-6500 spectrofluorometer. The ratio of the first (I) and third peak (III) λmax values was plotted against peptide P7 concentration to determine the critical aggregation concentration 7 .