Scalable thioarylation of unprotected peptides and biomolecules under Ni/photoredox catalysis† †Electronic supplementary information (ESI) available. See DOI: 10.1039/c7sc04292b

A mechanistically distinct, Ni/photoredox-catalyzed arylation of unprotected, native thiols (e.g., cysteine residues) is reported – a process initiated through a visible light-promoted, hydrogen atom transfer (HAT) event under ambient conditions.


General Considerations
All reactions were carried out under an inert atmosphere of nitrogen or argon in oven-dried glassware, unless otherwise noted. Conventional solvents (THF, Et 2 O, CH 2 Cl 2 , toluene, xylenes) were dried using a solvent system. DMF (99.8% , extra dry) was used as received, and all other reagents were purchased commercially and used as received, unless otherw ise noted. RuCl 3 ·xH 2 O and [NiCl 2 (dme)] were purchased from commercial sources.

Photochemistry:
Photoredox reactions were irradiated with blue LED (~ 470 nm, 32,918 mcd ft -1 ) strips, placed along the inside of a Pyrex® crystallizing dish (125 X 65 mm), and the temperature (~ 30 °C) was controlled using one external desk fan set up ~ 15-20 cm away from this photoreactor bed. The fan was employed to ensure reactions remained at or near room temperature when using LEDs which warmed during the reaction. A modified test tube rack was designed to allow multiple reaction setups simultaneously. LEDs were configured as outlined in the photochemical reactor design (image right). Purple LEDs (near UV, 400 nm) were also tested and set up in an analogous manner.
Reactions run in high-throughput fashion were carried out in a clear vial holder situated over a bed of small blue LEDs ( Figure  SI-1 b and c), and a fan was placed ~ 20 cm away from the reaction plate (not shown). S-3

General peptide purification parameters:
Reverse phase preparative chromatography for the small molecule peptides was performed by mass/UV (254 nm) directed preparatory liquid phase chromatography (Prep-LC), specifically Waters AutoPurification System equipped with Waters 3100 MS. Waters Sunfire, C18, 5 m 19x100 mm column was used, unless otherwise noted. Reverse phase preparatory chromatography for larger polypeptides was performed on a Varian Prostar HPLC system. Analytical HPLC for analysis of polypeptide purity and product formation was performed on an Agilent 1100 series HPLC system. Polypeptide starting materials were analyzed by low resolution electrospray ionization mass spectrometry (ESI-LRMS) on a Waters Acquity Ultra Performance LC connected to a single quadrupole detector (SQD) mass spectrometer. Polypeptide reactions were analyzed using Matrix Assisted Laser/Desorption Ionization-mass spectrometry (MALDI-MS) on a Bruker Ultraflex III mass spectrometer with a time-of-flight detector.

High-Throughput Experimentation:
High Throughput Experimentation (HTE) was performed at the Penn/Merck Center for High Throughput Experimentation at the University of Pennsylvania. All reagents were used as purchased from commercial suppliers. Solvents were purchased anhydrous and used with no further purification. All reactions were performed inside a glovebox operating with a N 2atmosphere (oxygen typically < 5 ppm). Reaction experimental design was aided by the use of Accelrys Library Studio. 2.5 μmol scale reactions for the limiting reagent were carried out in HPLC vial glass inserts, 4 x 21 mm, 50 L, flat bottom) equipped with magnetic tumble stir bars in 96-well reaction plates. Liquid handling was carried out using single and multichannel pipettors (10, 100, 200, and 1000 μL). On completion of solution dosing the plates were covered by a perfluoroalkoxy alkane (PFA) mat, followed by two silicon rubber mats, and an aluminum cover which was tightly and evenly sealed by 9 screws. Reactions were monitored by UPLC-MS. Column: Acquity UPLC HSS C18 1.7 m 2.1x50 mm, using 0.1% TFA MeCN and 0.1% TFA H 2 O as mobile phases. The instrument was equipped with an SQD detector with electrospray ionization (ESCi) source in positive and negative mode. High throughput data analysis was carried out with Virscidian Analytical Studio™ software.

General High-Throughput Experimentation Protocol
Stock solutions of the relevant catalyst or ligand were prepared in DCE or THF. Using a 200 μL pipettor, the catalyst or ligand solutions were dispensed into 1 mL glass vials arrayed in a S-4 96 well microtitre plate according to the required Ni/L loading per reaction screen. The solvent was removed in vacuo. Inorganic bases (if applicable) Cs 2 CO 3 , K 2 CO 3 , K 3 PO 4 , and Na 2 CO 3 were added either by slurry addition of the relevant base (25 mg/mL slurry stock solution in THF) or by manual solid addition using a BioDot™ DisPro MAR Series Adjustable Mass Powder Dispenser. The vials were then charged with tumble stir bars. Using a multichannel pipettor, a solution of aryl bromide and thiol (100 μL) in the relevant screen solvent was added to each 1 mL vial. Stock solutions of each additive (1 M/20 μL in THF) was then added using a 200 μL pipettor. The 96 well reaction plate was covered with the PFA film, sealed, removed from the glovebox, and placed on a bed of blue LEDs (Figure SI-1c). The vials were stirred at the corresponding reaction temperature for the allotted time. The reactions (if required) were cooled to ambient temperature and quenched with a stock solution of 4,4'-dimethylbiphenyl in MeCN (0.04 M, 500 μL) and mixed thoroughly. Aliquots of the quenched reactions were taken (20 μL), further diluted with 700 μL of MeCN, and subjected to UPLC and/or UPLC-MS analysis.

Select Reaction Optimization Studies
Additional studies examining various reaction conditions as deviated from the standard:   Equiv. H2O

S-6
Effect of aryl bromide (4-cyano bromobenzene) loading on the reaction outcome at various time points. In general, with higher aryl bromide loading, the faster the reaction was after 1 h, although decomposition of the arylated GSH to unidentified products was observed after prolonged reaction times (Chart SI-1). These reactions were run in triplicate (averaged), and the optimized reaction conditions/yield can be observed using 1 equiv ArBr after 18 h. The trend-lines overlaid to show the effects of increasing ArBr equiv over temporal extremes (1 and 18 h), all else constant. More dilute reaction conditions were pursued employing 1 equiv of the ArBr and GSH (1.2 equiv) at various time points, which would be ideal when examining small quantities of peptide thiols (Chart SI-2). As depicted, 0.1 M DMF proved optimal with these reagent parameters, and reactivity diminished (nearly linearly) at more dilute concentrations.
Chart SI-2. ArBr loading at various reaction concentrations over time under otherwise standard reaction conditions. Yield as compared to product/IS ratios.
Additional optimization reactions to uncover "dilute conditions" while maintaining fast reaction kinetics (< 4 h) was pursued at various loadings of aryl bromide (reactions run together in duplicate, Chart SI-3). 20 equiv of ArBr was found to optimal in a side-by-side comparison of the reactions after 2 h, although 10 equiv may be substituted in certain situations when ArBr is very expensive. As noted earlier, decomposition of the product was seen with higher ArBr loadings after prolonged reaction times (e.g., 20 h).

Precipitation of desired thioarlyated peptide (only tested with arylated GSH adducts):
The extraction process as detailed above was carried out and repeated 2-4 times. In cases where the thioarylated adduct is sparingly soluble in pure water, this pure product can slowly precipitate given suitable conditions. Following extraction, if solid material was present, it was helpful to let the aqueous solution sit on the benchtop and/or gradually cool (0-5 C) for 2-18 h to encourage maximum precipitation. The thioarylated peptide precipitate was then vacuum filtered through a glass frit or Büchner funnel and gently washed with ice water (no more than 2 mL), then Et 2 O. The filtered solid was analytically pure in nearly all cases.
*NOTE: In many cases, the final thioarylated peptide may not be fully soluble in the resulting DMF/H 2 O solution. Filtering this heterogeneous mixture before prep-LC has resutled in decreased yield as the filtered solid contained a fair amount of desired arylated peptide.

S-9
Gram-scale Thioarylation Reaction with Glutathione 3.5 mmol scale reaction: A 40 mL scintillation vial with air-tight, sealable septum was charged with glutathione (1.07 g, 3.5 mmol, 1.0 equiv, not fully soluble) and the aryl bromide (637 mg, 3.5 mmol, 1.0 equiv), followed by addition of diisopropylammonium bis(catechol)isobutyl silicate ( The now darker, milky-brown solution was transferred to a separatory funnel with deionized H 2 O (35 mL), followed by CH 2 Cl 2 (35 mL). This mixture was vigourously shaken, the layers were allowed to separate, and the organic layer was removed. This sequence was repeated 4 times. The remaining aqueous layer containing solid precipitate was transffered to a 50 mL Erlenmeyer flask and placed in the refrigerator for 7 h. The solid precipitate was vacuum filtered via Büchner funnel, washed with ice water (10 mL), then Et 2 O (30 mL) to afford a light beige solid (1.02 g, 72% yield). S-11  Product %area/IS %area ratios are reported [normalized against IS (10 mol %) giving relative values]. GSH provided uncharacteristically poor results in this HTE format, most likely because of poor solubility, stirring, or a combination of both. Parameters and conditions were explored to improve conversion across the board with GSH to little avail. Notably, bromide X2 was employed GSH under standard reaction conditions (0.1 mmol scale) and the thioarylated adduct was isolated in 50% yield (see manuscript). Thioglucose (tetraacetate used because of commercial availability) and tiopronin were agreeably soluble under the DMF screening conditions, and exhibited exceptional reactivity considering these were more hindered, secondary thiols.

S-12
Complete list of Merck informer halides examined: S-13

General Procedure for Thioarylation of Peptides under "Dilute Conditions"
For peptides and enzyme CoA reactions Reaction preparation and execution: In a separate, oven-dried vial was added a magnetic stirbar, the peptide or thiol of choice (13 µmol, 1 equiv), followed by three vacuum/Ar cycles.
To the vial containing the peptide and stirbar was added 1.3 mL of the DMF stock solution containing the reagents, under Ar, and the vial was sealed with Parafilm. This reaction vessel was placed in a blue LED photoreactor ~4-8 cm away from the irradiation source, and the reaction was allowed to stir for 1-3 h. The temperature of the reaction was maintained at approximately 29 C via an overhead fan (see Figure SI-3c). Reaction progress was carefully monitored via syringe needle by HPLC and/or UPLC-MS, and once complete, the slightly darker solution was either a) diluted with purified H 2 O (~ 0.5 mL) and directly purified by reverese phase preparatory liquid chromatography, or b) diluted with purified H 2 O (1.3 mL) and the aqueous layer extracted with CH 2 Cl 2 three timesthe reamining aqueous layer was subjected to purification.

NOTE:
Reactions with enzyme CoA and the various aryl bromides (20 equiv) were carried out in identical manner to that described above, except 4 equiv of the diisopropylammonium bis(catechol)isobutyl silicate was employed. See line listing reaction details below for additional details. S-14

Thioarylation of Peptide 9 under "Dilute Conditions" (0.68 µmol scale)
A stock solution was prepared as described above in the general procedure, relative to 0.68 µmol (1 equiv) of the starting peptide substrate.*

Reaction preparation and execution:
In a separate, oven-dried HPLC (or equivalent) vial, was added a magnetic stirbar and the thiol peptide 9 (0.68 µmol, 1 equiv). The vial was sealed with a tight fitting cap or septum, further sealed with Parafilm, and three gentle vacuum/Ar cycles ensued (Figure SI-3a). To the vial containing the peptide and stirbar under Ar was added 68 µL of the DMF stock solution containing the reagents (Figure SI-3b). The reaction vessel was placed in a blue LED photoreactor ~4-8 cm away from the irradiation source. The reaction was allowed to stir for 90 min. The temperature of the reaction was maintained at approximately 30 C via an overhead fan (see Figure SI-

Peptide HTE Screen
An HTE, micro-scale reaction screen (< 0.16 µmol, ~0.1 mg peptide/reaction well) in a 24well plate was conducted with thiol peptide 9 (primarily because of the limited availability of the peptide substrate) in an oxygen free glovebox. The 24-well plate conta ining all the reagents (in DMF stock solution, as previously described in the general experimental section) was screwed/sealed shut, free of oxygen, and run outside the glovebox over a bed of blue LEDs for 1 and 3 h. Crude reactions were analyzed by UPLC-MS (results in Table SI-2).
Cleaner reaction profiles to 10 (see traces below) were generally observed under these oxygen-free conditions than the previously optimized conditions on the benchtop.
*Note: It has been previously documented that lower photocatalyst loadings (< 0.25 mol %) can be more effective in certain Ni/photoredox cross-coupling reactions, 6 although this trend was not corroborated in this peptide thioarylation chemistry.

HPLC and MALDI Analysis of Peptide Cys Arylation Reactions
Analytical HPLC of peptide 10 was performed on a Phenomenex Luna C8 column using gradient 5, and an injection volume of 200 μL (flow rate of 1.0 mL/min). Absorbance was monitored at 215, 254, and 325 nm. The collected fractions from the product peak were analyzed by MALDI TOF-MS, and to confirm the selectively of the arylation for the Cys residue, tandem MS/MS fragmentation was performed. The fragmentation pattern for the major peak in peptide 10 (retention time = 22.7 min) shows unambiguous modification at the Cys residue without any apparent modifications at the Trp, His, or Tyr residues (Figure SI-4).
All other impurities in the chromatogram for peptide C do not correspond with a mass consistent with aberrant arylation at other residues or desulfurization of the Cys.

Enzyme CoA Data
A stock solution was prepared as described above in the general dilute reaction condition procedure, relative to 13 µmol (1 equiv) of the enzyme CoA substrate.

Reaction preparation and execution for Enzyme CoA Reactions (14-16):
Reactions with enzyme CoA (13 µmol, 1 equiv, 10 mg) and the various aryl bromides (20 equiv) were carried out in identical manner to that described previously under the dilute reaction conditions procedure , except 4 equiv of the diisopropylammonium bis(catechol)isobutyl silicate (52 µmol, 20.8 mg) were employed. Conversion to product was determined by UPLC-MS vs internal standard, until CoA starting material was consumed (90 min). Purification of the CoA adducts was attempted by reverse-phase prep LCMS in 0.1% TFA buffer (MeCN/H 2 O); however, clean material for spectral data analysis could not be fully ascertained because of product instability. UPLC-MS and HRMS data was collected from the crude reaction mixture following the organic extraction protocol.
CoA aryl sulfide 14 Aliquot from the crude reaction mixture of CoA thioarylate d adduct 16 (in green) with excess ArBr present (in blue) via UPLC-MS.

Cleavage/Arylation of GSH Disulfide (17)
A stock solution was prepared as described above in the general dilute reaction condition procedure, relative to 16 µmol (0.5 equiv) of the disulfide peptide substrate 17.

Reaction preparation and execution for Enzyme CoA Reactions:
Reactions with GSH disulfide 17 (16 µmol, 0.5 equiv) and aryl bromide 2 (20 equiv) were carried out in an identical manner to that described previously under the dilute reaction conditions procedure. Conversion to product was monitored and determined by LCMS/HPLC vs internal standard (90 min). Starting disulfide could still be detected by LCMS. The crude reaction mixture was diluted with H 2 O and extracted (CH 2 Cl 2 ) according to the general extraction protocol. The peptide (aqueous layer) was purified via reverse phase LCMS, to afford the title compound 3 as the TFA salt in 44% yield.
Analytical data matched that of the previously characterized compound.

Select Supplemental HPLC traces for arylated GSH adducts
The above trace profiles the crude reaction mixture of the thioarylated GSH adduct 3. The color-coded peaks designate product and byproducts including catalyst, ligands, and starting material (GSH). Mass chromatograms are included to verify arylated peptide (green, 0.48 min), catechol (red, 0.45 min), and photocatalyst (orange, 1.15 min) , and to assist in the user experience.

S-40
The traces below are aliquots from select reactions of the final aq layers following organic extraction with DCM (upon reaction completion) and before reverse-phase prep-LCMS. Unreacted aryl bromide and GSH can be observed in some cases, but traces are otherwise clean following the extraction protocol.