Biocompatible Hydrogelators Based on Bile Acid Ethyl Amides

Four novel bile acid ethyl amides were synthetized using a well-known method. All the four compounds were characterized by IR, SEM, and X-ray crystal analyses. In addition, the cytotoxicity of the compounds was tested. Two of the prepared compounds formed organogels. Lithocholic acid derivative 1 formed hydrogels as 1 % and 2 % (w/v) in four different aqueous solutions. This is very intriguing regarding possible uses in biomedicine.


Introduction
Bile acids are a fascinating group of biologically important molecules belonging to the group of steroids. 1,2 Because of their relatively low cost, wide availability, and enantiomeric purity, bile acids are ideal building blocks for gelator molecules. The structure of the bile acids consists of a steroidal backbone and an aliphatic side chain. 3 The facially amphiphatic nature of bile acids arises from their concave hydrophilic face and convex hydrophobic face. Chemically different hydroxyl groups in the concave face and the varying amount of them, as well as the rigid steroidal backbone, are important with regard to bile acids' biological etc. properties. Because bile acids are endogenous compounds 4 , they are perfect starting materials for biological and medical applications. 3,5 Bile acids and their derivatives have various applications especially in biomedicine. There are multiple examples in the field of cancer treatment. [6][7][8][9][10][11] One particularly interesting potential application involves mixing of sodium deoxycholate solutions to Au-nanoparticle (AuNP) solution to create multiple-branched Au-NPs. 8 AuNPs formed have such a strong NIR absorption that they can be used to destroy tumor cells. Many examples include bile acid based compounds as drug carriers [12][13][14][15][16][17] , and even gene 18,19 or RNA 20 carriers. Majority of potential applications of bile acids, or their derivatives, as drug carriers utilize bile acid transportation systems present in organisms/humans to achieve site-specific action of the drug carried. In addition, bile acid derivatives have shown potential as drug absorption modifiers. 21,22 Gels play increasingly important roles in modern life since they appear in myriad of applications ranging from optoelectronics 23,24 and biomedicine [23][24][25] to environmental clean-up. 26,27 Universally a gel is defined as a viscoelastic solid like material which consists of flexible cross-linked network and solvent. Gelator molecules form the cross-linked 3D network that attracts and captures solvent molecules. Gels are divided into different classes based on the solvent used. In hydrogels the solvent is pure water or water solution. If the solvent is organic, the gel is called an organogel, and dried gels are categorized as xerogels.
In the field of supramolecular gels the research focuses mainly on low molecular weight gelators (LMWG). 28 Typically LMWGs form gel networks that resemble fibres. LMWGs consist of a rich variety of molecules to which bile acids and their derivatives enter into.
The first examples of bile acid salts forming hydrogels were reported in the early 20th century. [29][30][31] Despite of that there is a limited amount of bile acid salts and derivatives known to function as hydrogelators. Most of the bile acid based hydrogels reported have been discovered by Maitra and his co-workers. Maitra's research group has reported intriguing gel systems, including cholic acid based luminescent europium containing gels and tripodal cholamide supergelator. This is, however, the first time bile acid alkyl amide derivatives are reported to gelate aqueous solutions.
Gels have potential use in biological applications only if they form in biocompatible solvents. Supramolecular hydrogels are considered to be biologically compatible, moreover they are formed in many cases by naturally occurring molecules which are most likely to be nontoxic. 32 Bile acid based hydrogels are considered to be fully biologically compatible when the gelator molecule itself is not toxic. The full biocompatibility of a hydrogel provides perhaps new potential applications regarding biomedicine, such as drug delivery, artificial tissue engineering, etc.
In relation to the previous work done by our research group [33][34][35][36][37][38][39][40] , in this work we have focused on how the compounds we have prepared behave in aqueous media. As a continuation of the series of bile acid alkyl or functionalized alkyl amide/ester derivatives capable of acting as gelators we report four new bile acid ethyl amide based gelator molecules, which have shown potential in forming hydrogels.
Triethyl amide, ethyl chloroformate, and other reagents used during the synthesis as well as solvents used in chromatography and gelation studies were of analytical grade. Ethyl chloroformate was distilled and 1,4-dioxane was dried over Na prior to use. All other chemicals were used without further purification. The synthetic route to 1-4 is presented in Scheme 1. The mixed anhydride method used has been previously reported by our research group. 41 Scheme 1: The synthetic route leading to compounds 1-4. procedure for the synthesis of the bile acid ethyl amide conjugates 1-4 Reactions were performed under N 2 atmosphere. In a round-bottomed three-necked flask bile acid (5 mmol, 1 eq.) and 1,4-dioxane (42 mL) were cooled on an ice-water bath to +10 ºC. To the cooled solution triethyl amide (6.7 mmol, 1.34 eq.) was added from a dropping funnel, followed by a dropwise addition of ethyl chloroformate (6.7 mmol, 1.34 eq.) in 1,4-dioxane (3 mL). The mixture was stirred at rt for 30 min. Meanwhile in another flask ethyl amide hydrochloride (6.7 mmol, 1.34 eq.) was suspended in DMF (10 mL). Suspension was cooled on ice-salt bath to 0 ºC after which triethyl amide (7.4 mmol, 1.48 eq.) was added from a dropping funnel. The mixture was stirred at rt for 30 min. Ethyl amide in DMF was added dropwise to the bile acid anhydride and stirring of the reaction mixture continued at rt for 20 h. Volatiles were evaporated and the crude product was dissolved in CHCl 3 (100 mL). The crude product was washed with water (2×75 mL), 0.1 M HCl solution (2×75 mL), water (2×75 mL), and brine (2×75 mL). The organic layer was dried (Na 2 SO 4 ), filtered, and the volatiles evaporated under reduced pressure. The crude products were purified by column chromatography (silica gel, DCM:MeOH 96:4 for 1, 90:10 for 2, x:y for 4, and were recorded with a Bruker Avance DRX 500 MHz spectrometer. The spectrometer was equipped with 5 mm diameter broad band inverse detection probehead operating at 500.17 MHz in 1 H and 125.77 MHz in 13 C experiments, respectively. The 1 H NMR chemical shifts are referenced to the signal of residual CHCl 3 (7.26 ppm from internal TMS). The 13 C NMR chemical shifts are referenced to the centre peak of the solvent CDCl 3 (77.0 ppm from internal TMS). A composite pulse decoupling, Waltz-16, has been used to remove proton couplings from 13 C NMR spectra.

Mass spectrometry
Compounds 1-4 were studied by mass spectrometry. Measurements were performed by using Micromass LCT time of flight (TOF) mass spectrometer with electrospray ionization (ESI).
Measurements were conducted using positive ion mode. MassLynx NT software system was used to control the spectrometer, and to acquire and process the data. Flow rate for the sample solutions was 10 µL/min. Sample cone and extraction cone potentials were 40 V and 6 V, respectively. The capillar cone potential varied between 3600 V and 4000 V, RF lens potential was 250 V in all measurements. The desolvation temperature was set to 120 °C and the source temperature to 80 °C.
Stock solutions of compounds 1-4 were 1 mM in acetone. Measurement solutions were prepared from stock solutions by diluting them to 10 µM or 20 µM in methanol. For the measurements of the hydrogels of compound 1, the solution system in question was measured and the spectrum obtained was subtracted from the spectrum of the gel to eliminate the background signal.

Gelation studies
Self-assembly properties of compounds 1-4 were studied by weighing 5 mg to obtain 1 % (w/v) systems or 10 mg to obtain 2 % (w/v) systems of a particular compound in a test tube and adding 500 µl of solvent or solvent mixture in question. The mixture was subjected to ultrasound for ca. 1 min with the cap on and heated with a heat gun until the compound was dissolved or the boiling point of the solvent reached without the cap. The solution or suspension was cooled to rt and the observations with regard gelation were made within 1-2 h after cooling had commenced. Formed solid mass was defined as a gel if there was no solvent flow when test tube was inverted. Possible crystallization and precipitation was observed within days or weeks. Test tubes were stored at rt during the observation time.

SEM
Samples for SEM were prepared by a traditional method. The hot solution containing the gelator in question in an appropriate solvent system was pipetted to the sample stub, and the gel allowed to form while cooling to room temperature. In addition to the traditional method, samples were prepared also directly from previously formed gels. A small amount of a gel in question was scooped with a spatula and the sample stub was thinly brushed with the gel. The gel was allowed to dry on the sample stub in both methods of sample preparation. After drying, all samples were thinly plated with gold with JOEL Fine Coat Ion Sputter JFC-1100. Micrographs were taken with Bruker Quantax400 EDS scanning electron microscope equipped with a digital camera.

Neutral red assay
The neutral red (NR) assay is grounded on the fact that live (non-damaged) cells intake and store NR into their lysosomes. The concentration of the incorporated dye is determined spectrophotometrically at 540 nm after extraction of retained NR into acidic methanolic solution. 42 After the incubation period the cells were first washed with PBS, and subsequently NR solution (0.03%, w/v in PBS) was applied to the cells for 3 h (37 °C, 5 % CO 2 ). Then the cells were washed with a washing solution (formaldehyde (0.125%; v/v), CaCl 2 (0.25%, w/v)), and the retained NR was dissolved in extraction solution (methanol (50%; v/v), acetic acid (1%; v/v)). The absorbance was measured with a microplate reader (Sunrise Remote, Tecan, Austria).

X-ray crystallography
Single crystal X-ray data for 1, 2, and 3 were collected at either 120 or 123 K using Agilent Super-Nova dual source wavelength diffractometer with an Atlas CCD detector using multilayer optics monochromatized CuK α (λ = 1.54184 Å) radiation. The data collection and reduction for 1, 2, and 3 was performed using the program CrysAlisPro. 43 Gaussian face index absorption correction method 43 was used for 1, 2, and 3. All the structures were solved with direct methods (SHELXS) 44 and refined by full-matrix least squares on F 2 using the OLEX2 45 , which utilizes the SHELXL-2013 module. 44 No attempt was made to locate the hydrogens for solvent molecules and heteroatoms, and all the hydrogen atoms were added using ADD H command in OLEX2. Attempts to solve the disorder of N-ethyl chain in compound 3 still engenders B-alerts and requires the usage of constraints (EADP) and restraints (ISOR), which ultimately affects the R 1 -value significantly. As a result, considering the conformational importance of side chain during our discussion, no attempt was made to solve the disorder.

Synthesis
A well-established method, employed several times by our research group 41,46 , was used to synthesize compounds 1-4. Syntheses were relatively fast compared to syntheses that last several days 46 . Purification of compounds 3 and 4 was challenging, probably due to the three hydrophilic functional groups of the steroidal backbone which cause stronger interactions with silica gel when using column chromatography. Yields varied between 34-75 %.   insoluble, G = gel with appr. 0-5 % free solvent, G-= gel with appr. 5-20 % free solvent, G a = gel formed after one day, PG = gel with appr. 50 % free solvent, P = precipitates on cooling.  were rather uniform when compared to each other. In the spectrum of 2 % (w/v) hydrogel of compound 1 in 10:90 H 2 O:acetone, however, a broad band at 3300-3500 cm -1 was observed ( Figure 2), probably due to hydrogen bonding of the gelator molecules. Since, based on X-ray crystallography, hydrogen bonding is important in stabilizing the solid state structures, and since the difference to the spectrum of solid compound 1 is so small, hydrogen bonding in not necessarily the only driving force in the gel network formation.

Microscopy
The hydrogels formed by compound 1 were studied with scanning electron microscope (SEM).
When comparing the SEM images of gels formed by compound 1 in aqueous solutions, it was found that the self-assembly patterns diverge from each other depending on the method of sample preparation. In the case of samples prepared from hot solutions, gel fibers resembled stalks of grass.
However, gelator molecules appeared to form spherical assemblies consisting of fiber bundles resembling old fashioned dusters in samples prepared from gels directly. The assemblies of these bundles were a spitting image of dandelions fluffy seeds. This was most clearly seen in images of

Results of the neutral red assay
The standard mouse cell line Balb/c 3T3, commonly used in accredited tests, was used for the toxicity study. The neutral red (NR) assay was performed for compounds 1-4 (see Equation E1 and Figures S1-S4). It is based on the fact that live (non-damaged) cells intake and store NR into their lysosomes, and the concentration of the incorporated dye can then be determined spectrophotometrically. The toxicity data are expressed as IC 50 ± S.D. (µg/ml) values. Substances efficiency in inhibiting specific biochemical or biological function is described as the half maximal inhibitory concentration (IC 50 ). As can be seen from Table 4, the highest cell viability in the series of compounds was shown by compound 3, which did not form any gels. Compound 1, which formed hydrogels, showed third lowest toxicity in the series. In a previous study performed by us for bile acid-cysteamine conjugates 35 it was observed that the hydroxyl group in position 7α and/or 12α might cause the compound to show toxicity, whereas the non-toxic compounds only had a 3αhydroxyl, 3α-and 7β-hydroxyls, or no hydroxyl groups at all. The current results don't agree with those hypotheses, even though compound 4 bearing no hydroxyls shows the second lowest toxicity in the series.   (Figure 4b). Compound 2, however, crystallized in the triclinic space group P-1 containing three crystallographically independent molecules of 2 (see Table S1) and four acetonitrile molecules in the asymmetric unit. The dihedral angles C13-C17-C20-C22, C17-C20-C22-C23, C20-C22-C23-C24, and C22-C23-C24-N24 (Table 5)      To the best of our knowledge, two host framework arrangements identified within a crystal packing has not been previously reported. Bile acids without hydroxyl groups are often and sometimes difficult to crystallize. For example in the current study, attempts to obtain good quality crystals of compound 4 were unsuccessful. Thin needle shape crystals were obtained after boiling 4 overnight in 1:1 water:methanol solvent ratio (data completeness < IUCr limit). However, the crystal structure shows that the lattice stabilization occurs uniquely at N-H functional group with methanol and water molecules.
Crystal data and X-ray experimental details for compounds 1, 2, and 3 are presented in Table S2.  Table S1: Side chain dihedral angles for three crystallographically independent molecules in compound 2.