Genetically Encoded Aminocoumarin Lysine for Optical Control of Protein–Nucleotide Interactions in Zebrafish Embryos

The strategic placement of unnatural amino acids into the active site of kinases and phosphatases has allowed for the generation of photocaged signaling proteins that offer spatiotemporal control over activation of these pathways through precise light exposure. However, deploying this technology to study cell signaling in the context of embryo development has been limited. The promise of optical control is especially useful in the early stages of an embryo where development is driven by tightly orchestrated signaling events. Here, we demonstrate light-induced activation of Protein Kinase A and a RASopathy mutant of NRAS in the zebrafish embryo using a new light-activated amino acid. We applied this approach to gain insight into the roles of these proteins in gastrulation and heart development and forge a path for further investigation of RASopathy mutant proteins in animals.

S3 0.5 µL each) were added. Reaction mixtures were incubated for 1 hour at 37 °C. The caPKA gene was amplified using primers 5 and 6 from a gene fragment (Twist Bioscience) and the NRAS G60E gene was amplified with primers 7 and 8 from a gene fragment (Twist Bioscience) using PCR (50 µL reaction) with recombinant Taq polymerase (Thermo Scientific EP0402) following the manufacturer's protocol. The following thermocycler conditions were used: 95 °C for 3 min, 34 cycles of 95 °C for 30 sec, 55 °C for 30 sec, 72 °C for 60 sec, then 72 °C for 5 min. The PCR reactions were purified using the GeneJET PCR Purification kit (Thermo). The caPKA PCR product was digested with BamH1 and EcoR1 (NEB, 1 µL each in 1x CutSmart buffer) or the NRAS G60E gene product was digested with Cla1 and EcoR1 (NEB, 1 µL each in 1x CutSmart buffer) for 1 hour at 37 °C. The product was purified by agarose gel electrophoresis and gel extraction as described above. T4 DNA Ligase (1 µL) and 10x T4 DNA ligase buffer (1 µL, NEB) were added to a solution of the insert gene (37.5 ng) and digested, purified pCS2 (50 ng) in MilliQ water (8 µL), and the reaction mixture was incubated at room temperature for 1 hour. Top10 chemically competent cells were transformed with the T4 DNA ligase reaction mixture (5 µL) in the same manner as detailed above. Plating, colony selection, miniprep, and Sanger sequencing were performed as detailed above to yield pCS2-caPKA and pCS2-NRAS-G60E.

Expression and purification of sfGFP-Y151ACK and sfGFP-Y151AC2K in bacteria
Top10 cells transformed with pBK-EV12 2 and sfGFP-Y151TAG-PylT 3 were inoculated from a glycerol stock into LB media (Miller formulation, 2 mL) containing kanamycin (50 µg/mL) and tetracycline (25 µg/mL). After overnight growth (37 °C, 280 rpm), the saturated culture (0.9 mL) was diluted in LB media (45 mL) containing kanamycin (50 µg/mL) and tetracycline (25 µg/mL). The resulting cell suspension was mixed thoroughly by pipetting it up and down five times, and was split in half by pipetting equal portions (20 mL each) into three separate 125 mL Erlenmeyer flasks. The cultures were incubated at 37 °C, 280 rpm until reaching OD600 0.3. At this time, ACK or AC2K (200 µL of a 100 mM stock solution in DMSO for a 1 mM final concentration) were added to two cultures, and DMSO vehicle (200 µL) was added to the other. The flasks were wrapped with aluminum foil to exclude light, and were returned to the incubator for 25 min, after which arabinose (200 µL of 20% (w/v) solution for a 0.2% final concentration) was added to induce protein expression. The cultures were incubated overnight (37 °C, 280 rpm), then were transferred to 50 mL conical tubes and were pelleted by centrifugation (4,500 rcf, 10 min, swinging bucket rotor). The pellets were washed with PBS (20 mL) and were resuspended in 5 mL of ice-cold lysis buffer (50 mM sodium phosphate, pH 8.0) containing Triton X-100 (0.1%) and lysozyme (1 S4 mg/mL). After gentle mixing on ice for 60 min, the cell suspensions were sonicated on ice (6 cycles of 30 s on, 30 s off, 45% amplitude, Fisher Scientific 550 Sonic Dismembrator with microtip). The resulting lysates were clarified by centrifugation at 4 °C (13,000 rcf, 10 min). The supernatants were transferred to 15 mL conical tubes and 100 µL of Ni-NTA resin (Qiagen) was added to each sample. The mixture was gently rocked on ice for 2 h. The resin was collected by centrifugation at 4 °C (500 rcf, 10 min), washed with lysis buffer (2 x 400 µL), washed with lysis buffer supplemented with imidazole (20 mM, 2 x 400 µL), and eluted with lysis buffer supplemented with imidazole (250 mM, 400 µL). The purified protein was analyzed by SDS-PAGE (10%) stained with Coomassie Brilliant Blue G (1 mg/mL in 7:2:1 dH2O:methanol:acetic acid).

Whole-protein MALDI-TOF-MS
Sample preparation. A µC18 resin Zip-Tip (Millipore-Sigma Cat. No. ZTC18M) was placed on a P20 micropipette set to 10 µL per the manufacturer's instructions. The following procedures were then carried out without allowing the tip to dry out once it was wetted. The resin was wetted by drawing up HPLC-grade MeCN (10 µL), and expelling 8-9 µL of it onto a Kimwipe, and repeating the process 4-5 times. The resin bed was then equilibrated with 10 washes of 0.1% TFA in Milli-Q water by the same procedure. A sample of sfGFP-Y151ACK (10 µL) in elution buffer (see above) was mixed with 20% TFA in Milli-Q water (2 µL) in a microcentrifuge tube. The sample-TFA mixture was then loaded onto the resin by pipetting it back-and-forth, up through the resin and back into the microcentrifuge tube 10-15 times, taking care to avoid fully drying out the resin each time. The resin was washed with 0.1% TFA in Milli-Q water (10 x 10 µL), each time expelling 8-9 µL of the wash onto a Kimwipe. SA elution solution (3 µL) consisting of 10 mg/mL sinapic acid in 1:1 MeCN:0.1% TFA (aq) was pipetted into another PCR tube. The purified protein sample for MS analysis was eluted by passing through the tip back-and-forth 15 times.

MALDI-TOF-MS analysis.
A stainless steel target plate (Bruker MSP 96) was pre-spotted with SA elution solution (0.5 µL). The plate was dried in open air for 10 min, and the resulting crystals were gently wiped with a Kimwipe, leaving some on the target. The eluted sample solution was pipetted portion wise on top of the wiped matrix crystals (3 x 1 µL), allowing the sample to air-dry completely in a dark drawer for 10 min between each addition. Samples were analyzed on a Bruker Daltonics UltrafleXtreme MALDI-TOF-MS in linear positive mode.

Whole-protein ESI-MS
Analysis was performed on a Thermo Scientific Q-Exactive Orbitrap mass spectrometer connected to a Dionex Ultimate 3000 UHPLC system. The sample was analyzed with a ProSwift RP-10R 1 mm x 5 cm column, flow rate 200 μL/min and a 26→80% MeCN in 0.1% HCOOH gradient over 30 min. The mass spectrometer was operated in positive-ion mode with a capillary voltage of 3.5 kV and resolution set to 17,500. Sheath gas, auxiliary gas, and sweep gas flow rates were 35, 10, and 5 L/min, respectively. Source temperature was 250 °C. The S-lens RF level voltage was 50 V and the ion transfer tube temperature was 250 °C. The instrument was tuned and calibrated with Pierce LTQ ESI positive ion calibration solution (Thermo Scientific) and the data were collected for m/z values ranging from 500-3000. Xcalibur 3.0.63 and Protein Deconvolution 3.0 software were used for data analysis and deconvolution.

Genetic encoding of aminocoumarin lysines in mCherry-TAG-EGFP-HA reporter in HEK293T cells -imaging and western blot.
HEK293T cells (American Type Culture Collection) were seeded at 160,000 per well into a 6-well plate (Greiner Bio-One 657160) that had been coated with poly-D-lysine (HBr salt, MP Biomedicals, 70-150 kDa). Cells were grown for 18 h (37 ºC, 5% CO2) to ~80% confluency in Dulbecco's Modified Eagle's Medium (DMEM; HyClone Laboratories, GE Life Sciences) supplemented with fetal bovine serum (FBS, 10% (v/v), Sigma-Aldrich), penicillin (100 U/mL, Corning Cellgro), and streptomycin (100 µg/mL, Corning Cellgro). To transfect the cells, a solution of linear polyethylenimine (1 mg/mL, LPEI, Polysciences) was diluted to 0.33 mg/mL with prewarmed Opti-MEM media (Gibco). The resulting solution (10 µL) was added to solutions of mini-prepped (Omega Bio-tek) plasmid DNA (1.5 µg each) in Opti-MEM media (200 µL). The plasmids were pE363-mCherry-TAG-EGFP-HA-PylT4 (provided by the laboratory of Jason Chin) as the reporter, and pE323-(IPYE)HCKRS-PylT4 encoding the PylRS and additional copies of PylT. 4 Cell culture media was replaced with antibiotic-free DMEM (1.8 mL) supplemented with FBS (10% (v/v)) in the presence or absence of the UAA (ACK, AC2K, or HCK, all 0.25 mM) or DMSO vehicle (0.25% (v/v)). The transfection reagent mixtures were incubated for 15 min before adding to the cells, at which time 180 µL of the reagent mixture was added per well. The reagent mixture was added by dispensing a small droplet out of the tip of the pipette tip and touching it to the surface of the media over the cells. Once the reagent was added to all wells, the plate was gently swirled for ~5 seconds and placed in the incubator (37 ºC, 5% CO2). After 48 h, cells were washed with warm PBS (2 x 1.25 mL per well) and changed to fresh PBS (2 mL) to check for protein expression by imaging. The fusion protein was visualized by epi-fluorescence microscopy using a Zeiss Axio Observer Z1 (10x objective, Plan-Apochromat 0.4 NA) with a short mercury arc bulb (Bulbman Inc Osram) in a Zeiss HBO 100 W housing and Zen 2 Blue Edition software. Filter cubes were excitation (ex); BP550/25, emission (em), BP605/70 for mCherry; ex, BP470/40; em, BP525/50 for EGFP; and ex, G 365; em, BP 445/50 for HCK, ACK and AC2K.
Cell lysis. After imaging, cells were cooled on ice, PBS was removed, and cells and lysed in icecold mammalian protein extraction buffer (250 µL) (GE Life Sciences) supplemented with Halt Protease Inhibitor Cocktail (Thermo Scientific) on ice with orbital shaking for 20 min. The lysed cells were scraped from the plates using P1000 pipette tips. Lysates were pipetted into microcentrifuge tubes, and were clarified by centrifugation (21,000 rcf, 20 min, 4 ºC).
Mounting. Once the staining was complete, all PBS was removed. The chamber dividers were snapped off, and the rubber gasket beneath it was carefully removed using a razor blade to free the edge and gentle traction with forceps. The slide was air-dried for 2 h in a drawer to exclude light and dust. A small drop of Gelvatol reagent (Center for Biologic Imaging, University of Pittsburgh; protocol for preparation below) was applied to each circle of cells using a 3 mL syringe.
A 22 x 60 mm #1.5 coverslip was gently applied starting at a ~15° angle, taking care to push all air bubbles out to the far edge while lowering the slide to a horizontal position. The mounted slide was set at 4 °C overnight (laying flat). The following day, the edges were sealed by applying a minimal quantity of Entellan ("New" formula, Millipore-Sigma) with a cotton swab to the edges, curing for 5 h at room temperature in the absence of light (in a box), and storing at 4 °C in the exclusion of light thereafter. The slide was equilibrated to room temperature for 30-60 min before imaging to avoid condensation on the surface. Any dust was gently wiped from the slide with a lens tissue, taking care not to put pressure on the samples.
The protocol for Gelvatol preparation is available on the CBI website (https://cbipitt.webflow.io/protocols). To prepare gelvatol, mix water (104 mL), tris (0.2 M, pH 8.5, 212 mL), and glycerol (84 mL) on a hot plate with stirring at 75 °C. Polyvinyl alcohol (Sigma P-8136 must be used; 42.0 g) is added in portions of ca. 5 g each, waiting until complete dissolution between additions (approximately 1 h per addition). The resulting mixture is clarified by centrifugation (5,000 rcf, 15 min), aliquoted, and stored at 4 °C. Aliquouts were stored in 3 or 5 mL syringes with caps or with the orifice plugged with the narrow end of a P2 pipette tip; Gelvatol may be dispensed directly from these syringes onto slides.
Imaging. Cells were imaged with a Zeiss Axio Observer Z1 with an Andor Zyla 4.2 camera, Excelitas X-Cite 120 LED Boost light source, and Slidebook 6 software (3i). Two objectives were used: 40x air (NA 0.6 LD Plan-Neofluar) and 63x (oil immersion, NA 1.4 Plan-Apochromat. Zeiss Immersol 518F immersion oil for fluorescence microscopy (Zeiss Cat. No. 1262466A, refractive index 1.518) was used with the 63x lens. Filter cubes were used for fluorescence microscopy for GFP (Chroma filter 49002, Ex. ET 470/40; Em. ET 525/50), for mCherry (Zeiss filter set 43, Ex. BP 550/25; Em. 605/70), and for ACK and AC2K (Chroma filter 49028, Ex. ET 395/25, Em. ET 460/50). Image deconvolution was performed in Slidebook 6 (3i) with a PSF estimated based on the NA, index of refraction, and working distance of the objective lens when required by the algorithm. For ACK, No Neighbors Deconvolution was performed with a subtraction constant of 0.9, theoretical spacing of 0.5, and edge padding of x: 78, y: 56 (default values). For AC2K, the Autoquant Blind Widefield Deconvolution algorithm was used with the maximum number of iterations set to 10; edge padding and correction for spherical aberration were disabled.

In vitro transcription
To synthesize mRNA, 10x CutSmart buffer (1 µL) and Not1-HF (1 µL, 20 U/µL, NEB) was added to a solution of the corresponding pCS2 plasmid (4 µg) in MilliQ water (8 µL). The reaction mixture was incubated at 37 °C for 1 hour. The reaction mixture was diluted with water (40 µL) to a total volume of 50 µL, and the product was purified by phenol:chloroform:isoamyl alcohol (PCIA) extraction. Briefly, PCIA (50 µL) was added to the reaction mixture and the suspension was vortexed and then centrifuged (16,200 rcf for 1 minute). The top aqueous layer was carefully pipetted off into a new microcentrifuge tube. Sodium acetate (3 M, 1/10 volume of DNA solution, 5 µL) was then added, and the DNA was precipitated with 100% ethanol (2.5x volume of the DNA and sodium acetate solution, 138 µL) at -20 °C overnight, then pelleted at 16,200 rcf for 5 minutes. The DNA pellet was washed with 70% ethanol, pelleted by repeating the centrifugation step, and the supernatant was removed by pipette. The pellet was dried at room temp with the cap open for 5 minutes before dissolving it in MilliQ water (20 µL).

LC-MS decaging assay (pH-dependent)
Sodium phosphate buffer solutions (100 mM) (pH 5.0, 5.5, 6.0, 6.5, 7.0, 7.5, and 8.0) were prepared, and HCK or ACK was diluted to 100 µM in the buffer from a 100 mM DMSO stock. For each pH, duplicate solutions (50 µL each) were transferred into polypropylene autosampler vials (Chemglass, Cat. No. CV-1007-1232). The 405 nm LED (Luxeonstar, Luxeon Z, 675 mW) was held about 3 cm above the solution (350 mW, measured with Thorlabs Power Sensor (S170C) and Touch Screen Power and Energy Meter Console (PM200)), and samples were irradiated for 10 seconds. Samples were diluted to a common matrix with the addition of 10x PBS, pH 7.4 (10 µL) before analysis. Photolysis products were analyzed by LC-MS using a Shimadzu 2020 instrument. A 2.1 mm x 100 mm Hypersil C18 GOLD column (1.9 µm) at 40 °C was used. The injection volume was set to 3 µL, and a 3→45% MeCN in 0.1% HCOOH gradient was run over 10 minutes. EICs were obtained for the UAAs and for the released coumarin products in positive mode. The masses were [M+H] + = 365 for HCK and [M+H] + = 193 for its product, and [M+H] + = 364 and [M+H] + = 192 for ACK. The retention times were 6.01 min for HCK and 5.96 min for the released product, and 5.85 min for ACK and 5.58 min for the released product. Peak height for was calculated automatically in LabSolutions software (Shimadzu) using File > Print Graph Image > PDF Output.

LC-MS comparison assay (pH 7.4)
Methylation. Since the lysine methyl ester provided a much stronger signal in the LC-MS than the free amino acid, aliquots of ACK and HCK (20 μL, 2.0 nmol) were added to dry MeOH (200 μL) in flame-dried glass vials. The solutions were cooled to 0 °C and trimethylsilyl-diazomethane (20 μL, 40.0 nmol, 2 M in Et2O) was added slowly. The vials were wrapped in aluminum foil and allowed to stir at room temperature overnight. Upon completion the volatiles were evaporated and the resulting DMSO solutions were used directly in the assay. Samples were diluted with addition of 10x PBS, pH 7.4 (15 µL) before analysis to recapitulate the matrix from the pH-dependent assay. Photolysis products were analyzed by LC-MS using a Shimadzu 2020 instrument. A 2.1 mm x 100 mm Hypersil C18 GOLD column (1.9 µm) at 40 °C was used. The injection volume was set to 10 µL, and a 3% → 45% MeCN containing 0.1% HCOOH gradient was run over 10 minutes. EICs were obtained for the methylated ACK and HCK and for the released Lys-OMe in positive ion mode. The masses were [M+H] + = 379 for HCK-OMe, [M+H] + = 378 for ACK-OMe, and [M+H] + = 161 for Lys-OMe. The retention times were 6.97 min for HCK-OMe, 6.83 min for ACK-OMe, and 1.52 min for Lys-OMe. Peak height was calculated automatically in LabSolutions software (Shimadzu) using File > Print Graph Image > PDF Output.
Data Analysis. Initial concentrations of HCK-OMe and ACK-OMe were determined by measuring the concentration of HCK and ACK before and after reaction with trimethylsilyl-diazomethane. The difference represents the HCK-OMe and ACK-OMe concentration. This was achieved by acquiring standard curves from stock solutions (0, 25, 50, and 100 μM) that were prepared in the same way as above. Using this method, we found that the concentration of HCK-OMe was about 76 μM and the concentration of ACK-OMe was about 59 μM before irradiation. The concentration of Lys-OMe released from each irradiation time was calculated using a standard curve for commercially available Lys-OMe (0, 25, 50, and 100 μM). Finally, the percentage of released Lys-OMe for each reaction was obtained by dividing the concentration of Lys-OMe by the concentration of HCK-OMe and ACK-OMe before irradiation and multiplied by 100%.

Generating HCKRS docking models
The docking of HCK and ACK into the HCKRS binding pocket were performed using Chimera (UCSF) and Autodock Vina software. First, the PylRS synthetase protein model was opened in Chimera (PDB: 2Q7H) and the appropriate HCKRS pocket mutations were introduced using the "change rotamer" function (Tools>Structure editing>Rotamers, Y271A, L274M). The "Dock prep" function was performed using the default conditions (Tools>Structure editing>Dock prep), then the Autodock Vina tool was opened (Tools>Surface/Binding analysis>Autodock Vina). The PDB model of HCK or ACK was selected as a ligand, and the HCKRS as the receptor, then the docking selection box was adjusted around the UAA binding pocket (as visualized by the pyrrolysyl UAA residing in the pocket from the original PDB model). The pyrrolysyl UAA was then deleted, leaving an open pocket for docking. The docking process was performed, generating several models. Only the model with the correct orientation of the UAA in the pocket was used for comparison between HCK and ACK.

Synthesis of caged 7-aminocoumarin lysine (ACK)
7-Aminocoumarin lysine 1 was synthesized from 3-aminophenol via the six step reaction sequence shown in Scheme S1. The first four steps, adapted from the literature, 5 were to construct the intermediate 7-aminocoumarin alcohol 4 from 3-aminophenol. The alcohol 4 was reacted with 1,1'-carbonyldiimidazole, followed by Boc-L-lysine-OH, to yield the coumarin-lysine 3 in 74% yield over two steps. Subsequent Boc group deprotection using TFA in the presence of TES as a cation scavenger led to the 7-aminocoumarin lysine amino acid 1 as a TFA salt in 86% yield.
Scheme S1. Synthesis of the 7-aminocoumarin lysine 1.  (1): TFA (0.5 mL) was added to a solution of 3 (88 mg, 0.20 mmol) and triethylsilane (64 µL, 2.2 mmol) in DCM (0.5 mL), and the resulting mixture was stirred at room temperature under a nitrogen atmosphere for 45 min. The reaction mixture was concentrated under vacuum. The residue was dissolved in MeOH (2 mL) and concentrated under reduced pressure. This process was repeated one more time to remove residual TFA. Finally, the crude product was dissolved in a minimal volume of MeOH and precipitated in Et2O under vigorous stirring. The precipitate was collected, and dissolved in a minimal volume of MeOH followed by precipitation in Et2O once more. The resulting precipitate was collected, redissolved in a minimal volume of MeOH, and precipitated in Et2O a third time to furnish coumarin lysine 1 (75 mg, 86%) as an amorphous white solid. 1

Synthesis of fluorescent 7-aminocoumarin lysine (AC2K)
Fluorescent 7-Aminocoumarin lysine 2 was synthesized from 3-bromophenol via the eight step reaction sequence shown in Scheme S2. Pechmann condensation of 3-oxopentanedioic acid and 3-bromophenol in the presence of methanesulfonic acid led to 2-(7-bromo-2-oxo-2H-chromen-4yl)acetic acid 11. The carboxylic group of 11 was then reduced to alcohol 10 using BH3·Me2S. After TBDMS protection of the alcohol, it was subjected to Buchwald-Hartwig amination leading to a Boc protected 7-aminocoumarin derivative 8. Next, the TBDMS group was released with the aid of TBAF and the resulting free alcohol 7 was converted to p-nitrophenyl carbonate 6 by reaction with 4-nitrophenyl chloroformate. Finally, the Boc-L-lysine-OH was reacted with the carbonate 6 to yield the coumarin-lysine 5, which upon treatment with 50% TFA led to desired fluorescent 7-aminocoumarin lysine amino acid 2 as a TFA salt in 86% yield.