Automated Flow Peptide Synthesis Enables Engineering of Proteins with Stabilized Transient Binding Pockets

Engineering at the amino acid level is key to enhancing the properties of existing proteins in a desired manner. So far, protein engineering has been dominated by genetic approaches, which have been extremely powerful but only allow for minimal variations beyond the canonical amino acids. Chemical peptide synthesis allows the unrestricted incorporation of a vast set of unnatural amino acids with much broader functionalities, including the incorporation of post-translational modifications or labels. Here we demonstrate the potential of chemical synthesis to generate proteins in a specific conformation, which would have been unattainable by recombinant protein expression. We use recently established rapid automated flow peptide synthesis combined with solid-phase late-stage modifications to rapidly generate a set of FK506-binding protein 51 constructs bearing defined intramolecular lactam bridges. This trapped an otherwise rarely populated transient pocket—as confirmed by crystal structures—which led to an up to 39-fold improved binding affinity for conformation-selective ligands and represents a unique system for the development of ligands for this rare conformation. Overall, our results show how rapid automated flow peptide synthesis can be applied to precision protein engineering.


Materials and Methods
Safety Statement: No unexpected or unusually high safety hazards were encountered.

Synthesis of FKBP51 FK1 variants by AFPS
Five FKBP51 16-140 variants were synthesized by Automated Flow Peptide Synthesis in the Pentelute Group according to the established conditions described in (1,2).

Automated flow peptide synthesis
Approximately 130 mg of H-Rink amide ChemMatrix resin (loading 0.18 mmol/g) was loaded into a fritted syringe (6 mL), swollen in amine-free DMF for 15 min and washed alternatingly with DCM and DMF.The beads were loaded into the reactor of the in the Pentelute Lab, selfbuild automated flow peptide synthesizer ("Peptidator") (2).The general flow rate was 40 mL/min, the temperature in the loop was 90 °C and 85-90 °C in the reactor.The first step was a 60 s wash step at elevated temperatures with 40 mL of amine-free DMF.The couplingdeprotection cycle was repeated for all additional monomers used in the sequences.
The shown sequence represents the recombinantly FKBP51 16-140  For the synthetic variants the presented sequence was used as template.Additional modification specific to the variants are summarized in the table below.
Table SI1: Alterations in protein sequence specific to the variants.After synthesis, the beads (loading: 0.0221 mmol, 1 eq.) were washed thoroughly with amine free DMF (3 x 5 mL) and DCM (3 x 5 mL) before drying them in a vacuum chamber.5.1 mg Pd(PPh3)4 (0.00442 mmol, 0.2 eq.) was dissolved in 4 mL DCM and 54 µL of PhSiH3 (0.442 mmol, 20 eq.) was added.The mixture was added to the syringe containing the dried beads.The suspension was incubated for 30 min at room temperature.This step was performed twice.The beads were then washed with DCM (3 x 5 mL) and subsequently DMF (3 x 5 mL).

Lactamization
The beads (loading: 0.0221 mmol, 1 eq.) were mixed with a solution of 57.6 mg PyAOP (0.1105 mmol, 5 eq.) and 80 µL DIPEA (0.442 mmol, 20 eq.) in 2 mL amine free DMF.The reaction mixture was incubated for 4 h at room temperature on a rolling device.The beads were washed thoroughly with DMF for 3 x 5 min and subsequently with DCM for 3 x 5 min.The beads were then dried in a vacuum chamber.

Cleavage
The dried beads were mixed with cleavage solution to remove the remaining protecting groups and release the proteins from the beads.For approximately 200 mg of beads, 5 mL of cleavage solution (82.5 % TFA, 5 % water, 5 % phenol, 5 % thioanisole, 2.5 % EDT) was added and kept on a rolling device for 3 h at room temperature.The proteins were precipitated in 40-45 mL ice cold diethyl ether and the precipitate was collected by centrifugation while the supernatant was discarded.Traces of diethyl ether and TFA were removed as much as possible by evaporation with a nitrogen gas flow.The proteins were dissolved in 50 % acetonitrile (0.1% TFA) in water and dried by lyophilization.

LC-MS
To check the mass of the synthesized proteins, a 1 mg/mL solution was prepared and diluted with 50 % HPLC acetonitrile in water with 0.1% TFA as additive to a 0.1 mg/mL solution.The solutions were filtered and measured on an Agilent 6545 Accurate-Mass Q-TOF LCMS system.
The used solvent composition is for solvent A MQ water with 0.1% formic acid as additive and for solvent B HPLC grade acetonitrile with 0.1% formic acid.Method used: C3-1-95 in 15 min using the Agilent Zorbax 300SB-C3 column (2.1 mm × 150 mm, 5-μm particle size).

Analytical HPLC
The purity of the samples was analyzed by analytical HPLC.The measurements were performed on Agilent Technologies 1200 Series instrument with a Kinetex® 2.6 µm (C18, 100 Å, 100 x 2.1 mm) column, with solvent A being MQ water and solvent B being HPLC-grade acetonitrile with 0.1 % TFA as additive for each solvent.The flow rate was 0.375 mL/min and a linear gradient of 5-65 % solvent B in 60 min was used for all samples.For sample preparation a 1 mg/mL protein solution in 50 % solvent B in solvent A was prepared and filtered.5 µL of each sample was injected.

ETD measurements
For top-down electron transfer dissociation (ETD) measurements, the lyophilized proteins were dissolved in 200 mM ammonium acetate (pH 7, LC-MS grade) at a final concentration of 30 µM.To remove small molecules from the synthesis, samples were buffer exchanged 5-times with 200 mM ammonium acetate buffer (pH 7, LC-MS quality) at 14,000 x g for 30 min using analysis was performed in two ways: manually and software-assisted.For the manual evaluation the spectra were accumulated, smoothed with Savitzky-Golay-filter with a smoothwindow of 3 and a 2 times repetition, and centered with the median.The observed masses were compared with the theoretical mass list of possible fragments, which were calculated with Proteomics Toolkit (4).The second evaluation was done with the software MASH Native (5).
Therefore, the UniDec algorithm (6) accumulates the spectra and then fits them with eTRASH (7) with a peak background ratio of 2, S/N Threshold of 2, maximum charge state of 17, mass range 100 to 2000 m/z, and a maximum mass of 13950 Da.In addition, we made sure that at least 4 isotopes of a fragment were present for it to be identified as a fragment.We looked for c-and z-fragments, as these are the typical fragments generated in ETD (8,9).

Size exclusion chromatographynative & denatured
The refolded proteins and recombinant protein stock solutions were buffer-exchanged four times in 200 mM ammonium acetate buffer (pH 7, LC-MS grade) at 14,000 x g for 30 min using 3k Amicon ® Ultra centrifugal filters (Merck).The buffer-exchanged protein solutions were diluted with 200 mM ammonium acetate buffer to an initial concentration of 10 µM.For the denatured measurements the protein solutions were diluted to a final protein concentration of

Ion mobility mass spectrometry-native & denatured
The refolded proteins and recombinant protein stock solutions were buffer-exchanged four times in 200 mM ammonium acetate buffer (pH 7, LC-MS grade) at 14,000 x g for 30 min using 3k Amicon ® Ultra centrifugal filters (Merck).The buffer-exchanged protein solutions were diluted with 200 mM ammonium acetate buffer to an initial concentration of 10 µM.The samples were measured with direct-infusion nanoESI on a Synapt G2-S from Waters. 10 µL of protein solution was loaded in a self-pulled nanoESI capillary (Micropipette Puller Model P-97 from Sutter Instrument).The Capillary Voltage was between 1.0 and 2.0 kV, Sampling Cone 20 V, Source Offset 50 V, Source Temperature 30°C, Cone Gas Flow 200 L/h, and Purge Gas Flow 300 mL/h for the nanoESI source.For recalibration, we performed an internal calibration using the masses of the analytes with a first-order regression.
For the IM-MS measurements, we used helium at 180 mL/min as the cooling gas and nitrogen at 90 mL/min as the drift gas for the IMS cell.Wave velocities of 800, 1000 and 1150 m/s were used in the ion mobility cell.The wave height was 40 V each time with a constant wave velocity of 141 m/s and a wave height of 6 V in the transfer cell.The DC bias voltage of the trap was 30 V. The quadrupole with the RF generator of 800 kHz was set to RF only, and the TOF was in resolution mode.The mass range was from 400 to 5000 m/z.Before the measurements, the MS was mass calibrated with sodium iodide to 5000 m/z.
For CCS calibration, we measured bovine ubiquitin, equine cytochrome c, and equine holomyoglobin with the same settings.A logarithmic regression model was chosen for calibration (10).The peak value is the highest data point, while the weighted average is based on equation ( 1). ( Where xi is the collision cross section associated with arrival time i and yi is the corresponding intensity. For the denatured IM-MS measurements the samples were diluted to a protein concentration of 100 nM with 0.2 % formic acid in 95:5 Water:Acetonitrile with (LC-MS grade).
The IM-MS measurements were performed with an HDX setup from Waters.This included a PAL RTC Autosampler from Leap, a UHPLC with µBinary Pump and Auxiliary Pump, the HDX Manager and a Synapt XS from Waters.The column installed was a Acquity UPLC Protein BEH C4 from Waters which is based on SiO2 particles (1.7 µm particle size, 300 Å pore size, 50 x 2.1 mm) and an oven temperature of 25°C.50 µL of sample was injected.
Eluent A was 0.2% formic acid in water (LC-MS grade) and eluent B was 0.2% formic acid in acetonitrile (LC-MS grade).The gradient started at 95% A and 5% B and was held for 1 min.
Between 1 min and 5 min, it was increased to 50% B.Over the course of 5 to 10 min, 100% B was reached and held for 4 min.Within 6 s, the gradient was set back to 5% B for 6 min to reequilibrate the column.Subsequently, the sample was ionized with an ESI source in positive mode.The source voltage was 3 kV, with a Cone Voltage of 50 V, a Source Offset of 20 V, at a Source Temperature of 70°C and Desolvation Temperature of 200°C.We set the Desolvation Gas Flow to 550 L/h, the Nebuliser gas flow to 6 bar and the cone gas flow to 100 L/h.The Synapt XS uses traveling-wave ion mobility (11,12).For the IM-MS measurements we used helium at around 3.4 mbar as cooling gas and nitrogen at 2.9 mbar as drift gas for the IMScell.Wave velocities of 500, 600 and 700 m/s were used.The wave height was held at 40 V with a constant wave velocity of 141 m/s and wave height of 6 V in the transfer cell.The trap DC bias was 45 V.The quadrupole with the RF generator of 300 kHz was set to RF only and the TOF was in resolution mode.The mass range was from 100 to 2000 m/z.Before the measurements the MS was mass calibrated with sodium iodide up to 5000 m/z.
We used the LockSpray with leucine-enkephalin as a reference, where every minute for one second the reference was measured, with the masses 120.0813 m/z, 278.1141 m/z, 397.1876 m/z, 425.1825 m/z and 556.2771 m/z ( 13) with a first-order regression.
For the CCS calibration we measured with the same settings separate solutions with 10 µM protein in 200 mM ammonium acetate buffer of bovine ubiquitin, equine cytochrom c, and equine holo-myoglobin.The mobilograms were extracted with CIUSuite 2 (14).A logarithmic regression model was chosen for calibration (10).The top value is the highest data point, while the weighted average was based on Equation (1).

Protein production and purification
Chemically competent E. coli BL21 (DE3) cells were transformed with a plasmid containing the gene construct for FKBP51 16-140 (A19T, C103A, C107I, 16-140 AA) with an N-terminal His-and Sumo-Tag.Therefore, 10 ng of the plasmid was mixed with chemically competent E.coli BL21 (DE3) cells and incubated on ice for 30 min.The heat shock is performed for 45 s at 42°C, following the addition of 250 µL SOC media and an incubation of 2 h at 37°C.The preculture was prepared by adding the transformed cells to a flask containing 200 mL of LB media and 0.1 % v/v kanamycin which was incubated over night at 37°C and with 150 rpm. 4 L of LB media were inoculated to an OD600 of 0.1 with the preculture.As soon as a OD600 of approximately 0.5 was reached, the culture was induced by the addition of 1 mM isopropyl 1thio-D-galactopyranoside and incubated over night at 25°C and with 150 rpm.The cells were harvested by centrifugation (10 000 g, 10min, 4°C) and stored at -20°C until lysis.the supernatant.To achieve a higher purity, the sample was also purified by SEC using the HiLoad 16/600 Superdex 75 pg column with a flow rate of 0.5 mL/min (Buffer: 20 mM HEPES pH 8, 20 mM NaCl).The fractions containing the pure protein were concentrated using Amicon ® Ultra 2 mL centrifugal filter.The protein was aliquoted and stored at -80°C.

Fluorescence polarization
To determine the binding affinity of our low affinity tracer (Fluorescein conjugated SAFit1analogue, without Top-group) and the respective protein variants, fluorescence polarization binding assays were performed.In a 384-well assay plate (black, flat bottom, low-binding) a serial dilution of the protein variants under investigation were prepared in FP-Assay Buffer (20 mM HEPES pH 8, 150 mM NaCl, 0.015 % Triton X-100).In case of an active site titration the tracer was added at an end concentration of 30 nM, and 1 nM for recording binding curves.
The plates were incubated for 30 min at room temperature and the fluorescence polarization was measured on a plate reader.The fluorescence polarization values were normalized with respect to the highest signal.       and lactam-bridged variants at charge state 8+, 11+ and 18+ under denaturing conditions.The collision cross section increases with a higher charge state, showing a similar denaturing behavior for all variants.
compared to the wildtype were made: -N-terminus starts at position A16 -The first three amino acids are Tag remains -A19T mutation for crystallography (3) -Cysteines are abolished: C103A and C107I -M48 and M97 are replaced by Norleucin replaced by Fmoc-Glu(OAll)-OH -For K58 the building block Fmoc-Lys(Alloc)-OH is used -The N-terminus is protected by incorporating Boc-Gly-OH F67E/K60 (i, i+7) -F67 is replaced by Fmoc-Glu(OAll)-OH -For K60 the building block Fmoc-Lys(Alloc)-OH is used -The N-terminus is protected by incorporating Boc-Gly-OH F67E/K60Orn (i, i+7) -F67 is replaced by Fmoc-Glu(OAll)-OH -K60 is replaced by the building block Fmoc-Orn(Alloc)-OH -The N-terminus is protected by incorporating Boc-Gly-OH F67E/K60Dab (i, i+7) -F67 is replaced by Fmoc-Glu(OAll)-OH -K60 is replaced by the building block Fmoc-Dab(Alloc)-OH -The N-terminus is protected by incorporating Boc-Gly-OH Orthogonal on-bead deprotection of Alloc and OAll groups Merck 3k Amicon ® Ultra centrifugal filters.At the end, 50 µL protein solution in 200 mM ammonium acetate buffer was obtained and mixed with 22 µL acetonitrile, 0.3 % formic acid, giving a ratio of 70:30 H2O:ACN and about 0.1 % formic acid.The samples were measured solution was loaded in a self-pulled nanoESI capillary (Micropipette Puller Model P-97 from Sutter Instrument).Settings were as follows: Capillary Voltage between 1.0 and 1.5 kV, Sampling Cone 50 V, Source Offset 20 V, Source Temperature 100 °C, Cone Gas Flow 200 L/h, and Purge Gas Flow 300 mL/h for the nanoESI source.The ETD reagent used was 1,4-dicyanobenzene with a Make-Up gas flow of 50 mL/min, flowing through the Glow Discharge at a current of 55 µA and a voltage of 0.9 kV.In the process, the trap cell was filled with anion radicals at an interval of one second for a duration of 0.1 seconds.The pressure inside the helium filled trap cell was around 5.5•10 -2 mbar with a trap RF of 450 kHz. 10 -2 mbar and a collision energy of 5 V was used.The MS/MS experiment was performed to isolate a charge state using a quadrupole equipped with a 300 kHz RF generator and set to an LM resolution of 14 and a HM resolution of 16.The TOF was set in resolution mode with a mass range of 50 to 2000 m/z, and measurements were performed for at least 20 minutes.Before the measurements, the instrument was calibrated with sodium iodide up to 5000 m/z.Internal recalibration was performed based on the fragments c4 (z = 1, 314.18232 m/z), z7 (z = 1, 691.3541 m/z), c8 (z = 1, 744.38869 m/z), z16 (z = 2, 934.9819 m/z), c19 (z = 2, 1001.5023m/z), c23 (z = 2, 1200.14837m/z) and z30 (z = 1, 1630.8362 m/z), if they were present.The 1 µM with 0.2 % formic acid in 30:70 acetonitrile:water (LC-MS grade).The SEC-MS measurements were performed with an HDX setup from Waters.This included a PAL RTC Autosampler from Leap, a UHPLC with µBinary Pump and Auxiliary Pump, the HDX Manager and a Synapt XS from Waters.The column installed was a bioZen SEC-2 from Phenomenex with pure SiO2 particles (1.8 µm particle size, 150 Å pore size, 150 x 2.1 mm) and an oven temperature of 20°C. 2 µL of sample was injected and the run was performed at an isocratic gradient of 35 µL/min with 50 mM ammonium acetate in water (LC-MS grade) for the native and 0.2 % formic acid in water (LC-MS grade) for the denatured measurements.After the chromatography the sample was ionized with an ESI source in positive mode.The source voltage was 3 kV, with a Cone Voltage of 50 V, a Source Offset of 20 V, at a Source Temperature of 70°C and Desolvation Temperature of 200°C.We set the Desolvation Gas Flow to 550 L/h, the Nebuliser gas flow to 6 bar and the cone gas flow to 100 L/h.The MS experiment was conducted without isolation or fragmentation in resolution mode of the TOF and a mass range of 100 to 5000 m/z.Before the measurements the instrument was calibrated with sodium iodide up to 5000 m/z.For the recalibration we used an internal recalibration with the analyte masses at different charge states.

The 4 -Figure SI2 :
Figure SI2: Structures of the used tracers.A: SAFit1 based tracer coupled to fluorescein.B: Low affinity tracer (without top group) coupled to fluorescein.

Figure SI4 :
Figure SI4: ETD Fragmentation map from top-down measurements.ETD fragmentation map from top-down measurements of recombinant FKBP51 16-140 (A), synthetic FKBP51 16-140 (B), F67E/K58 (i, i+9) (C), F67E/K60Orn (i, i+7) (D) and F67E/K60Dab (i, i+7) (E).The numbers on the right side correspond to the FKBP51 16-140 sequence.A-E: Typical c-or z-fragments for ETD are indicated in blue.In green are the c-fragments with an H2O-loss.The FKBP51 sequence starts with the alanine residue in position 16.The purple amino acids indicate the modifications C103A and C107I as well as the exchange of methionine in norleucine (Nle) in positions 48 and 97.Positions 60 and 58 are highlighted in yellow, position 67 in orange, and the black line indicates the formed lactam-bridge.

Figure SI5 :
Figure SI5: ETD Fragmentation map from top-down measurements.Note that the data in this Figure are the same as in FigureSI4, but the residue numbering follows that of the FKBP5116-140 construct, rather than that of the full-length protein.Numbers on the left count up, while numbers on the right count down, in order to facilitate tracking the cleavage position associated with c-and z-fragments, respectively.A: ETD fragment map of recombinant FKBP5116-140 , with a sequence coverage of 100 %, a residue cleavage of 62.50 %, and a total number of 49 c-fragments and 44 z-fragments.B: ETD fragment map of synthetic FKBP51 , with a sequence coverage of 100 %, a residue cleavage of 66.41 %, and a total number of 40 c-fragments and 50 z-fragments.L36 and L85 are norleucine.C: ETD fragment map of F67E/K58 (i, i+9), with sequence coverage of 90.6%, residue cleavage of 50.00 %, and a total of 25 c-fragments and 39 z-fragments in blue.L36 and L85 are norleucine.Shown in green are the c-fragments with a loss of H 2 O (-18.0153Da).Including these increases the c-fragments to 28 and the sequence coverage to 100 %.The black line indicates the lactam bridge formed.D: ETD fragment map of F67E/K60Orn (i, i+7), with sequence coverage of 90.60 %, residual cleavage of 47.65 %, and total number of 20 c-fragments and 41 z-fragments in blue.L36 and L85 are norleucine.K48 has a mass difference of -14.0156Da as a result of the replacement of lysine by ornithine.The black line indicates the lactam bridge formed.E: ETD fragment map of F67E/K60Dab (i, i+7), with sequence coverage of 91.40%, residue cleavage of 55.47 %, and total number of 22 c-fragments and 49 z-fragments in blue.L36 and L85 are norleucine.K48 has a mass difference of -28.0313Da because of the replacement of lysine by diaminobutyric acid.Shown in green are the c-fragments with a loss of H 2 O (-18.0153Da).Including these increases the number of c-fragments to 24 and the sequence coverage to 100 %.The black line indicates the lactam bridge formed.For B-E a mass difference of -0.984 Da is observed since the C-terminus is an amide due to synthesis instead of a carboxylic acid.

Figure SI6 :
Figure SI6: MS spectrum of all variants analyzed in the SEC-MS experiment (size exclusion chromatography mass spectrometry) under denaturing conditions.

Figure SI7 :
Figure SI7: MS spectrum of all variants analyzed in the SEC-MS experiment (size exclusion chromatography mass spectrometry) under native conditions.

Figure SI 9 :
Figure SI 9: Mobilograms of the recombinant, synthetic FKBP5116-140 and lactam-bridged variants at charge state 8+, 11+ and 18+ under denaturing conditions.The collision cross section increases with a higher charge state, showing a similar denaturing behavior for all variants.
.75 M guanidine HCl) under stirring.The plate was incubated for 1 h at 4°C before combining the fractions and exchanging the buffer to 20 mM HEPES pH 8, 20 mM NaCl and concentrating the samples using Amicon ® Ultra 2 mL centrifugal filters (MWCO 10kDa).
1 mg of the synthetic protein was dissolved in 250 µL denaturing buffer (6 M guanidine HCl, 100 mM Tris pH 8) and incubated at room temperature for 15 min.In 25 wells of a 96-well plate 10 µL of the protein solution was added slowly to 190 µL of Refolding Buffer (50 mM Tris-Cl pH 8.5, 9.6 mM NaCl, 0.4 mM KCl, 2 mM MgCl2, 2 mM CaCl2, 0.5 M arginine, 0.4 M sucrose, 0 For the lysis, the cells were resuspended in 15 mL lysis buffer (20 mM HEPES pH 8, 300 mM NaCl) before 2 mM EDTA, 1 mM PMSF and 1 mg/mL Lysozym were added.The incubation time was 2 h at 4°C on a rolling device.Afterwards 0.1 mg/mL DNase and 100 mM MgSO4 were added, and the tube is shaken at 4°C on a rolling device for another hour.Additionally,