Desymmetrization via Activated Esters Enables Rapid Synthesis of Multifunctional Benzene-1,3,5-tricarboxamides and Creation of Supramolecular Hydrogelators

Supramolecular materials based on the self-assembly of benzene-1,3,5-tricarboxamide (BTA) offer an approach to mimic fibrous self-assembled proteins found in numerous natural systems. Yet, synthetic methods to rapidly build complexity, scalability, and multifunctionality into BTA-based materials are needed. The diversity of BTA structures is often hampered by the limited flexibility of existing desymmetrization routes and the purification of multifunctional BTAs. To alleviate this bottleneck, we have developed a desymmetrization method based on activated ester coupling of a symmetric synthon. We created a small library of activated ester synthons and found that a pentafluorophenol benzene triester (BTE) enabled effective desymmetrization and creation of multifunctional BTAs in good yield with high reaction fidelity. This new methodology enabled the rapid synthesis of a small library of BTA monomers with hydrophobic and/or orthogonal reactive handles and could be extended to create polymeric BTA hydrogelators. These BTA hydrogelators self-assembled in water to create fiber and fibrous sheet-like structures as observed by cryo-TEM, and the identity of the BTA conjugated can tune the mechanical properties of the hydrogel. These hydrogelators display high cytocompatibility for chondrocytes, indicating potential for the use of these systems in 3D cell culture and tissue engineering applications. This newly developed synthetic strategy facilitates the simple and rapid creation of chemically diverse BTA supramolecular polymers, and the newly developed and scalable hydrogels can unlock exploration of BTA based materials in a wider variety of tissue engineering applications.


Mechanical properties:
Rheological measurements were performed using a DHR-2 rheometer using a 20 mm cone-plate geometry with a 2.002° angle. All measurements were performed at 20 °C. First, the linear viscoelastic region was determined by performing a strain sweep on the 10% (w/v) hydrogels. We found that storage and loss modulus are independent of strain below 10% strain and that hydrogels from C6C12 and C12C12 clearly showed hydrogel dominant behavior (G'>G"), while the hydrogel from C6C6 shows liquid dominant behaviour at this frequency (1 rad/s, Figure S54). A rejuvenation process was carried out to remove any mechanical history within gels. We choose a 400% strain amplitude at frequency 1 rad/s. Strain amplitude is well into the non-linear regime of the samples. Following the rejuvenation process, a time sweep at 1% strain was carried to observe the ageing kinetics and allow sample equilibration. Subsequently, a frequency sweep was carried out from 0.01 rad/s to 627 rad/s at 1% oscillation strain followed by oscillatory strain amplitude sweeps 1 to 1000% at an angular frequency of 1 rad/s. For self-healing measurements, a stepstrain experiment was carried out, the strain was varied between 400% and 1% (at 1 rad/s) for rupture and recovery phase of the sample.

ATDC5 chondrocytes cell culture:
ATDC5 were expanded at 37 °C in humidified incubator using a growth medium consisting of high glucose Dulbecco's modified eagle medium (DMEM) with 10% fetal bovine serum, 1% penicillin/streptomycin. The medium was changed after three days and ATDC5 were used for viability studies between 75-85% confluency.
Cell encapsulation in BTA supramolecular hydrogelators ATDC5 (chondrocytes) were encapsulated in 200 µL BTA hydrogel at a concentration of 5 million cells per mL of hydrogelator. For encapsulation of cells within hydrogels, 100 µL of hydrogel was transferred to the well plate, centrifuged to form a uniform layer at the bottom of the 48-well plate (non-treated for cell culture). Cell suspension in 25 µL of media was spread on top of the gel and centrifuged at 80 RCF for cells to sediment in a gel. After centrifugation, the other 100 µL of the hydrogel was added on top of the hydrogel and gently mixed using a spatula. On top of each gel, 200 µL media was added and the gels were incubated at 37°C. Dulbecco's Modified Eagle's Medium-F12 (DMEM-F12, low glucose) supplemented with 10% (v/v) fetal bovine serum (FBS) and 1% (v/v) P/S) was used in this study.

ATDC5 viability in BTA supramolecular hydrogelators:
LIVE/DEAD™ Viability/Cytotoxicity Kit (Thermo Fisher Scientific) was used for evaluating cell viability. For live-dead staining, the hydrogels were transferred into a 35 mm glass-bottom dish (ibidi, Germany), washed with PBS, and a solution of calcein-AM (final concentration 1 µM) and ethidium homodimer-1 (final concentration 2 µM) was added to each gel. The gels were incubated for 45-60 minutes in the dark at 37 °C and imaged using an inverted fluorescence microscope (Nikon Eclipse Ti-e) under the conditions of 37 °C and 5% CO2. Images were analyzed using ImageJ.

Synthesis of tris(perfluorophenyl) benzene-1,3,5-tricarboxylate (3):
To a three neck clean and dry 250 mL round bottom flask equipped with a stir bar under nitrogen flow, pentafluorophenol (5.56 g, 30.1 mmol, 4 equiv) was added, followed by anhydrous dichloromethane (DCM, 30 mL). The reaction solution was stirred until complete dissolution of pentafluorophenol. The reaction solution was then placed into an ice bath and allowed to cool. After cooling, dry diisopropylethylamine (DIPEA, 3.89 g, 30.1 mmol, 4 equiv.) was mixed with anhydrous dichloromethane (~80 mL) and added dropwise to the reaction solution under vigorous stirring. Subsequently, benzene-1,3,5-tricarbonyl trichloride (2.0 g, 7.5 mmol, 1 equiv.) dissolved in ~25 mL of dry DCM was added dropwise to the reaction mixture. The reaction mixture was removed from the ice bath after 30 minutes and allowed to stir at room temperature for 4 hours. TLC (ran in DCM) showed two spots at Rf close to zero and at Rf 0.9. The reaction mixture in DCM was concentrated in vacuo, passed through a filter paper, and then through a silica bed twice. The solvent (DCM) was removed under reduced pressure, yielding 3 (BTE-F5Ph) as pure white solid (4.9 g, 91% yield). 1

Synthesis of tris(2,5-dioxopyrrolidin-1-yl) benzene-1,3,5-tricarboxylate (4):
A 50 mL clean and dry single-neck flask equipped with a stir bar was purged with nitrogen. Benzene-1,3,5tricarbonyl trichloride (0.1 g, 0.3 mmol, 1 equiv.) was added under nitrogen atmosphere and the flask was closed with a screw cap. Freshly dried THF (using NaOH pellets) was added to the flask using a cannula and stirred until dissolved. DIPEA (0.38 g, 3 mmol, 8 equiv.) and NHS (0.21 g, 1 mmol, 5 equiv.) dissolved in 7 mL of anhydrous THF was added drop by drop to the reaction mixture using a syringe. The resulting mixture was stirred for 20 hours and then the volatiles were removed by vacuum. TLC showed two spots at Rf equal to ~0.1 and ~0.4 ( Figure S9); however, during separation using silica gel flash column chromatography, 4 was not obtained pure and free NHS seems to eluted with 4 ( Figure S10, 1 showing presence of free NHS. Free NHS appears to be travelling with 4. We were able to separate reaction mixture spots on TLC as shown above in figure S9 but 4 was not obtained pure when separated using silica gel flash column chromatography.

Synthesis of molecule 5:
Under nitrogen atmosphere, 1 (20 mg, 0.11 mmol, 1 equiv.) was dissolved in a dry round bottom flask using 3 mL anhydrous DMF. Color of the reaction mixture was milky white and 1 was not fully soluble in the reaction solvent. DIPEA (12 µL, 0.069 mmol, 0.66 equiv.) and hexylamine (4.6 µL, 0.034 mmol, 0.33 equiv.) was dissolved in 3-4 mL and added drop by drop to the reaction flask using a syringe and under nitrogen atmosphere. The reaction solution was stirred at room temperature for 5 hours. With reaction proceeding solubility of 1 increases and the color of the reaction mixture changes from milky white to light yellow and all the crystals disappeared. TLC of the reaction mixture was taken in different solvent combinations as shown above in Figure S11 and showed a maximum of three spots with Rf close to 0.1, 0.7, and 0.9 in DCM with 10% (v/v) acetonitrile. After 5 hours reaction solvent was concentrated in vacuo and molecules were separated using silica gel flash column chromatography. Using dichloromethane with 5% (v/v) acetonitrile as eluent, 5 was obtained (6.9 mg, 37% isolated yield). 1

Synthesis of molecule 7:
Under nitrogen atmosphere, 1 (20 mg, 0.11 mmol, 1 equiv.) was dissolved in a dry round bottom flask using 5-6 mL anhydrous DMF. The reaction mixture color was milky white and non-transparent and 1 was partially soluble in anhydrous DMF. DIPEA (18 µl, 0.11 mmol, 1 equiv.) and hexylamine (9.2 µl, 0.07 mmol, 0.66 equiv.) was dissolved in 7-8 mL of anhydrous DMF and added drop by drop to the reaction flask using syringe and at room temperature under nitrogen atmosphere. The reaction mixture was stirred for 5 hours and the color of the reaction goes from milky white to yellow and clear without any solid crystals. 7 was separated using silica gel flash column chromatography using DCM with 10% (v/v) acetonitrile. 1

Synthesis of molecule 6:
In dry round bottom flask, 2 (20 mg, 0.11 mmol, 1 equiv.) was dissolved in 3-4 mL anhydrous DMF under nitrogen atmosphere. DIPEA (12 µL, 0.069 mmol, 0.66 equiv.) and hexylamine (4.6 µL, 0.034 mmol, 0.33 equiv.) was dissolved in 3-4 mL and added drop by drop to the reaction flask using syringe and under nitrogen atmosphere. Color of the reaction mixture was milky white and upon adding DIPEA and hexylamine, the reaction solvent turns yellow and the reaction mixture was clear without any solid crystals. After 5 hours, the reaction mixture was dried in vacuo 6 was separated by silica gel flash column chromatography using DCM with 5% (v/v) acetonitrile. 6 was not obtained as pure and it contains starting materials 2, free 4-nitrophenol, which can be seen in Figure S17 1 H NMR spectrum. Molecule 6 was produced in 38% yield from proton 1 H NMR peak integration analysis. In dry round bottom flask, 2 (20 mg, 0.11 mmol, 1 equiv.) was dissolved in 5-6 mL anhydrous DMF under nitrogen atmosphere. DIPEA (18 µL, 0.11 mmol, 1 equiv.) and hexylamine (9.2 µL, 0.07 mmol, 0.66 equiv.) was dissolved in 5-6 mL of anhydrous DMF and added dropwise to the reaction flask under nitrogen atmosphere.

Synthesis of molecule 9:
In the dried round bottom flask under nitrogen atmosphere molecule 3 (15 mg, 0.064 mmol, 1 equiv.) was dissolved in 3 mL of anhydrous DCM and the reaction glass flask was placed into a cold dry ice bath in acetone (-78 °C   Synthesis of molecule 10:  In dry round bottom flask, 3 (50 mg, 0.21 mmol, 1 equiv.) was dissolved in 3 mL of anhydrous DCM under nitrogen atmosphere. The temperature of the solution was maintained to 4°C using cold ice bath. DIPEA solution (49.1 µL, 0.28 mmol, 1.33 equiv.) was dissolved in anhydrous 3 mL anhydrous DCM and added to the reaction flask. Reaction solvent color turns yellow. Subsequently, solution of dodecylamine (13.1 µL, 0.07 mmol, 0.33 equiv.) in 3 mL anhydrous DCM was added dropwise to the reaction flask. The reaction was allowed to stir at 4°C and after 2 hours reaction solvent was removed under reduced pressure and 1 H NMR of the reaction mixture was taken and 11 was produced in 39% yield from proton 1 H NMR analysis. Using DCM solvent on silica gel flash column chromatograph, 11 was separated. 1     Under nitrogen atmosphere and in a dry round bottom flask, 3 (20 mg, 0.084 mmol, 1 equiv.) was dissolved in 3 mL anhydrous DCM. Reaction flask was placed into ice cold bath (4 °C) and then DIPEA (9.84 µL, 0.06 mmol, 0.67 equiv.) dissolved in anhydrous DCM was added into the reaction flask. Later, 5Nb-2MA (3.62 µL, 0.03 mmol, 0.33 equiv.), dissolved in anhydorus DCM, was added to the reaction flask drop by drop over 5-10 minutes using syringe under nitrogen atmosphere. After addition, the reaction solution was allowed to stir for 2 hours and formation of 13 was checked by thin layer chromatogrpahy shown in Figure  S28. The excess of DCM was removed in vacuo and 13 was separated using silica gel flash chromatography using eluent DCM as white powder in 40% isolated yield (7.2 mg) as white powder. 1

Synthesis of molecule 14:
A dry round bottom flask was loaded with molecule 3 (500 mg, 2.11 mmol, 1 equiv.) under nitrogen atmosphere and dissolved in 40 mL anhydrous DCM. Subsequently solution of anhydrous DIPEA (182 µL, 2.11 mmol, 1 equiv.) in 2 mL anhydrous DCM was added into the reaction flask. Recation flask was set inot ice bath (4 ºC). Solution of 5Nb-2MA (182 µL, 1.41 mmol, 0.67 equiv.) in DCM (25 mL) was added drop wise to the reaction flask in roughly 10 minutes under nitrogen atmosphere. The reaction was stirred for 2.5 hours at 4 ºC and reaction mixture was vacuum dried to remove excess solvent. Nb-BDA was separated running flash column chromatography on silica gel using eluent DCM/acetonitrile (92.5/7.5) by volume. Nb-BDA was obtained as a white powder (202 mg) in 49% isolated yield and 52% yield based on recovered starting material. 1        Desymmetrization using 6-amino-1-hexanol: Figure S33: A) Reaction scheme for desymmetrization of molecule 3 using hexylamine alcohol. B) 1 H NMR (CDCl3) of reaction mixture. Integration peak analysis indicates that there are total 4 molecules in reaction mixture which based on their peak ppm and peak integration analysis are mono-substituted, di-substituted and tri-substituted and leftover starting molecule 3 (BTE-F5Ph).

Computational Details
General Input files and molecular geometries were generated using Gaussview. Low energy structures were optimized in Gaussian 09W employing the B3LYP functional 1,2,3 up to the 6-311+G(d) basis set 4,5 ; in all cases, no negative frequencies were found in the optimized geometries. Atomic charges were calculated via the NBO partitioning method 6 which also allows for the calculation of orbital occupancies. Electrophilicity values (ω) 7 were calculated according to the relationship ω = µ 2 /2η, where µ is the electronic chemical potential and η is the global hardness. The values of µ and η can be approximated, applying Koopman's theorem 8  Data presented (aside from entry 1) are derived from quantum chemical calculations (at the B3LYP/6-311G* level) unless specified otherwise. Structures 9a and 10a consisted of a truncated methyl-substituted amide. Data is presented as the average of multiple esters unless specified. b 13 C NMR chemical shifts for the ester carbonyl carbons (in CDCl3). c Occupancy of the ester carbonyl π * orbital from NBO population analysis. d Calculated global electrophilicity index. e Electrostatic potential at nuclei (EPN).

Synthesis of molecule 17
To a clean and dry round bottom flask was added 12 (25 mg, 0.035 mmol, 1 equiv.) and dissolved in anhydrous dichloromethane (DCM, 4 mL) under inert atmosphere. To a stirring reaction solution was added dry DIPEA (13.5 µL, 0.077 mmol, 2.1 equiv.) and 5Nb-2MA (9.4 µL, 0.074 mmol, 2.1 equiv) solution in anhydrous DCM (3 mL) was added to the reaction flask. After overnight stirring excess DCM was removed in vacuo and 12 was obtained via column chromatography using eluent methanol/dichloromethane (5/95) by volume. Isolated yield of the reaction was 87% (20.1 mg). 1 H NMR spectrum is shown in Figure S39. It was not possible to clearly evaluate peak splitting pattern for some of the peaks in 1 H NMR and we leave all the information in the 1 H NMR spectrum. 1

Synthesis of molecule 18:
A clean and round bottom flask equipped with a stir bar was charged with 12 (25 mg, 0.035 mmol, 1 equiv.) and dissolved in anhydrous dichloromethane (DCM, 4 mL) under inert atmosphere. The reaction solution was stirred until 12 dissolved and then to a stirring solution was added solution of DIPEA (13.5 µL, 0.08 mmol) and 3-Azido-1-propanamine (0.01 g, 0.074 mmol, 2.1 equiv.) in anhydrous DCM (3 mL) was added. After overnight stirring at RT, the reaction mixture was vacuum dried to remove excess solvent (DCM) and 12 was obtained in 85% yield (19 mg) by running flash column chromatography on silica gel using eluent methanol/dichloromethane (5/95) by volume. 1

PEG-CDI synthesis:
Bishydroxy PEG (20 kg/mol) was dried via azeotropic distillation using toluene. Dried bishydroxy PEG (30 g, 3 mmol, 1 equiv.) was dissolved in 100 mL of anhydrous 1,4-dioxane at 37 °C under N2 atmosphere. Vacuum dried (at 50 °C for 3 hours) carbonyl diimidazole (CDI, 0.08 g, 0.5 mmol, 5 equiv. per OH) dissolved in 30 mL of anhydrous 1,4-dioxane was added to the reaction flask under nitrogen inert atmosphere. The reaction mixture was stirred at 37 °C for 3 hours. The reaction mixture was precipitated out in excess cold diethyl ether twice and dried overnight in a vacuum oven rotavap at 40 °C. The product was obtained as a white solid with a 96% (29 g) yield. 1

PEG-Bisamino dodecane synthesis:
PEG-CDI (27 g, 1.34 mmol, 1 equiv.) was vacuum dried at 60 ºC for 3 hours and dissolved in 270 mL anhydrous DMF. The polymer solution was added dropwise to the solution of 1,12 diaminododecane (3.5 g, 19 mmol, 14 equiv.) dissolved in 275 mL of anhydrous DMF and maintained at 70 ºC. The reaction mixture at 70 ºC was stirred for 24 hours under nitrogen atmosphere. The reaction mixture was concentrated by removing DMF and precipitated in excess cold diethyl ether. The product was again dissolved in DCM and precipitated out in excess cold diethyl ether. The product was obtained as a white solid in 97% yield. The mass of the polymer is 25760 g/mol with Đ of 1.

Synthesis of hydrogelator C6C6
In a dry round bottom flask, 10 (0.2 g, 0.34 mmol, 1.1 equiv.) was dissolved in 8mL anhydrous DCM and DIPEA (0.06 g, 0.44 mmol, 1.5 equiv.) was added into the reaction flask. Subsequently, PEG bisaminododecane (3.2 g, 0.3 mmol, 1 equiv.) solution in anhydrous DCM (5-6 ml) was added dropwise to the reaction flask. The reaction mixture was stirred for 40 hours at room temperature (~20 °C) under nitrogen atmosphere. Excess solvent was removed in vacuo and the crude reaction mixture precipitated in excess cold diethyl ether obtaining BTA 1 hydrogelator, as a white powder in 98% yield. Second purification was done by dialyzing the sample in methanol against methanol to remove any unreacted small molecule impurities. Hydrogelator C6C6 was obtained in 89% yield with the mass of 24344 g/mol and with Đ of 1.  15 (0.03 g, 0.043 mmol, 1.1 equiv.) was dissolved in 3 mL anhydrous DCM and DIPEA (0.008 g, 0.06 mmol, 1.5 equiv.) was added to it. PEG bisaminododecane (0.4 g, 0.04 mmol, 1.0 equiv.) was weighed in dried flask, dissolved in anhydrous DCM (3-4 mL) and added to the flask. The reaction continued to run for 40 hours at 20 °C under nitrogen inert atmosphere. The reaction mixture was concentrated using vacuum and precipitated out using excess cold diethyl ether. The same procedure was repeated twice. The product was dried in a vacuum oven at 45 °C for 3 hours and the white product was obtained in 75% yield. Hydrogelator C6C12 was obtained with mass 21207 g/mol with Đ of 1.  12 (C12C12, 0.23 g, 0.32 mmol, 1.1 equiv.) was dissolved in 8 mL anhydrous DCM and DIPEA (0.056 g, 0.44 mmol, 1.5 equiv.) was added to it. PEG bisaminododecane (3 g, 0.29 mmol, 1 equiv.) was weighed in dried flask, dissolved in anhydrous DCM (5-6 mL) and added to the flask. The reaction continued to run for 40 hours at 20 °C under an inert atmosphere. Reaction mixture was first purified by precipitation in cold diethyl ether and the pure molecule was obtained in 98% isolated yield as a white powder. Later also sample was dissolved in methanol and dialyzed against methanol for further purification. The pure molecule was obtained in 84% yield. Hydrogelator C12C12 was obtained in 84% yield with the mass of 23573 g/mol and with Đ of 1.

Vial inversion experiment:
Mechanical Properties: Figure S53: Vial inversion experiment to determine flow behaviour of hydrogels at 10% (w/v). Hydrogelator C6C6, C6C12 and C12C12 are labelled as HH (orange), DH (green) and DD (red). Flow behavior is qualitative test and provide information on viscoelastic behaviour of hydrogels. Hydrogelator C6C6 (HH) roughly started to flow after 20 minutes of vial inversion. A small flow can be seen for hydrogelator C6C12 after 1680 minutes (28 hours) and hydrogelator C12C12 did not flow even after 4320 minutes (72 hours).  Figure S55: Self-healing and moldability of hydrogelators C6C6 (orange), C6C12 (green), and, C12C12 (red). All hydrogelators are self-healing instantaneously when hydrogels pieces are placed and pressed together with spatula. Also hydrogels are moldable since they can adopt to different mould shapes when subjected to stress. Live-dead cell area was calculated using Image J by stacking z stacks live and dead channel images were processed separately for calculating % of cell areas. The number of aggregates counted were between 190 and 500 for all BTA hydrogelators.