Structural Elucidation of 2-(6-(Diethylamino)benzofuran-2-yl)-3-hydroxy-4H-chromen-4-one and Labelling of Mycobacterium aurum Cells

Abstract

Trehalose conjugates of 3-hydroxychromone (3HC) dyes have previously been utilized as fluorescence labels to detect metabolically active mycobacteria with a view to facilitating point-of-care detection of mycobacterial pathogens, especially Mycobacterium tuberculosis. We subjected the 3HC dye 2-(6-(diethylamino)benzofuran-2-yl)-3-hydroxy-4H-chromen-4-one (3HC-2) to a combined X-ray crystallography and density functional theory (DFT) study, and conducted preliminary fluorescence labelling experiments with the model organism Mycobacterium aurum. In the crystal, 3HC-2 exhibits an s-cis conformation of the chromone and the benzofuran moieties about the central C–C bond. According to DFT calculations, the s-cis conformer is about 1.8 kcal mol⁻¹ lower in energy than the s-trans conformer. The solid-state supramolecular structure features hydrogen-bonded dimers and π…π stacking. Fluorescence microscopy revealed fluorescence of M. aurum cells treated with the dye trehalose conjugate 3HC-2-Tre in the GFP channel. It was concluded that s-cis is the preferred conformation of 3HC-2 and that the generally considered non-pathogenic M. aurum can be labelled with the fluorescence probe 3HC-2-Tre for convenient in vitro drug screening of new antimycobacterial agents.

1. Introduction

With a total of 1.6 million deaths and an estimate of 10.6 new cases worldwide in 2021, tuberculosis (TB) remains the leading bacterial killer and a public health threat [1]. The etiologic agent of TB is primarily Mycobacterium tuberculosis. Infections caused by non-tuberculous mycobacteria (NTM), which are mostly considered opportunistic pathogens, are also on the rise globally [2,3]. Point-of-care detection of mycobacteria based on low-cost microscopy methods [4] could help prevent and combat mycobacterial infections. In this context, Kamariza et al. recently reported on 3-hydroxychromone dye trehalose conjugates for the fluorescence labelling of mycobacterial cells [5]. As illustrated in Figure 1, exogenous trehalose molecules can be mycolylated at position 6 to give trehalose monomycolates (TMM), which are incorporated into the mycomembrane. A solvatochromic 3-hydroxychromone dye appended to trehalose as a fluorophore group appears to be tolerated by the converting enzymes antigene 85 (Ag85), which enables visualization of metabolically active mycobacteria.

Mycobacterium aurum is a fast-growing non-tuberculous mycobacterium [6], which has been used as a surrogate bacterium in anti-TB drug discovery [7,8,9], although its suitability as a model organism for M. tuberculosis has been called into question [10]. M. aurum is generally considered non-pathogenic, but a documented rare case of keratitis attributable to M. aurum has been reported in the medical literature [11]. We have also used M. aurum as a test bacterium in early-stage antimycobacterial drug discovery [12,13,14]. Therefore, we became interested in possible applications of the fluorescence probe 3HC-2-Tre (Figure 2) [5] for the labelling of M. aurum cells, as this could be a useful tool for in vitro testing of new antimycobacterial agents.

Whereas in-depth spectroscopic investigations of the dye 3HC-2 can be found in the literature [15,16], its crystal and molecular structure appears to be hitherto unpublished, as revealed by a search of the Cambridge Structural Database (CSD) [17] via the WebCSD interface [18] in April 2023. Therefore, we subjected 3HC-2 to X-ray crystallography, density functional theory (DFT) calculations, and Hirshfeld surface analysis in order to better understand its features. In this contribution, we report the molecular structure of 3HC-2 in the crystal, a computational study of its conformational preference, its supramolecular structure in the solid state, and the preliminary results of fluorescence labelling experiments with M. aurum cells using the dye trehalose conjugate 3HC-2-Tre.

2. Results and Discussion

2.1. Structural Description of 3HC-2

Compound 3HC-2 (Figure 3a) was synthesized as described in the literature [5]. In brief, 2′-hydroxyacetophenone was reacted with 6-(diethylamino)benzofuran-2-carbaldehyde [16], followed by treatment with hydrogen peroxide to give the desired 3-hydroxychromone derivative 3HC-2. Intense yellow crystals of 3HC-2, as shown in Figure 3b, grew from a solution in heptane/ethyl acetate. X-ray crystallography revealed that the compound crystallized solvent-free in the triclinic system, centrosymmetric space group P-1, with two molecules in the unit cell (Z = 2).

Figure 4 shows the molecular structure of 3HC-2 in the crystal. One of the ethyl groups exhibits disorder over two positions. As expected, the ten-membered chromene system is planar (r.m.s. deviation 0.0108 Å) as is the nine-membered benzofuran moiety tethered to C2 (r.m.s. deviation 0.0107 Å). The molecule adopts an s-cis conformation about the C2–C2′ bond. The O1–C2–C2′–O1′ torsion angle is −13.76(16)° in the chosen asymmetric unit, and the angle between the mean planes through the chromene and benzofuran moieties is 14.42(4)°. An s-cis conformation was encountered in the related 3-(furan-2-yl)-2-hydroxy-4H-chromen-4-one (CSD refcodes: IJUCEW and IJUCEW01) [19,20] and its nicotinic acid ester (MASGAS) [21]. Interestingly, an s-trans conformation was found in the crystal structures of the 2-thiophenyl derivatives ((5-(3-hydroxy-4-oxo-4H-chromen-2-yl)-2-thienyl)methylene)malononitrile (MUGGEC) [22] and 2-(thiophen-2-yl)-4H-chromen-4-one-3-O-2,3,4,6-O-tetraacetyl-β-d-glucopyranoside (QICLIA) [23].

A CSD search (April 2023) for crystal structures containing a 2,3-diether-butadiene moiety with a central acyclic C-C single bond revealed 107 entries. Of these, 80 structures exhibited an O–C–C–O torsion angle between 163 and −164°, and for 21 structures the O–C–C–O torsion angle was between −26 and 20°, as observed for 3HC-2. Intermediate torsion angle values in six structures can likely be attributed to packing effects or the steric bulk of substituents (NENLAT −60.2° [24], AHAYEO −47.3° [25], CIVXUC 37.9° [26], SILXAP 49.5° [27], HEJRUL 76.3° [28], DADHOJ 127.2° [29]). The relatively small number of structures with an O–C–C–O torsion angle around 0° prompted us to calculate the optimized structure of the free molecule of 3HC-2 using DFT methods. The minimum energy structure of the free molecule exhibited an O1–C2–C2′–O1′ torsion angle of 0°. Figure 5 shows a superposition of the 3-hydroxychromone moieties of the molecular structure in the crystal and the DFT-optimized molecular structure of the free molecule of 3HC-2, illustrating the conformational difference between the two structures, which can be attributed to packing effects in the crystal (vide infra). A relaxed surface scan was subsequently calculated in order to explore the conformational flexibility of 3HC-2 about the central C2–C2′ bond. This revealed a rotational barrier of ca. 7.5 kcal mol⁻¹ and the energy of the s-trans conformer was about 1.8 kcal mol⁻¹ higher than that of the local ground state structure adopting the s-cis conformation (Figure 6).

The supramolecular structure of 3HC-2 in the solid state features centrosymmetric O–H⋯O hydrogen-bonded dimers (Figure 7). The 3-hydroxy groups each act as a hydrogen bond donor towards the chromone carbonyl oxygen atom of the other molecule in a dimer, resulting in a ${\mathrm{R}}_2^2\left(10\right)$ motif [30]. The corresponding hydrogen bond parameters (

Table 1. Hydrogen bond geometry for 3HC-2 (Å, °).

D–H⋯A 1	d(D–H)	d(H⋯A)	d(D⋯A)	<(DHA)
O2–H2⋯O3i	0.857(14)	1.902(15)	2.7146(13)	157.9(17)

¹ Symmetry code: (i) −x, −y, −z + 2.

) are characteristic of strong hydrogen bonds [31]. The same intermolecular hydrogen bond motif was observed in the crystal structures of the unsubstituted 3-hydroxychromone (HOHHIW) [32] and in the aforementioned IJUCEW and MUGGEC. Whereas the chromone oxygen atom O1 does not exhibit contacts shorter than the sum of van der Waals radii in the crystal, the benzofuran oxygen atom O1′ is approached by a methyl hydrogen atom of the disordered ethyl group of an adjacent molecule. Figure 8 shows the Hirshfeld surface of 3HC-2 and the corresponding 2D fingerprint plot, revealing the dominance of short out of the plane O⋯H contacts, resulting from the intermolecular O–H⋯O hydrogen bonding described above and the π⋯π stacking of the molecules.

2.2. Labelling of Mycobacterium aurum Cells

The dye trehalose conjugate 3HC-2-Tre (Figure 2) for the labelling of M. aurum cells was prepared from anhydrous trehalose and 3HC-2 as described by Kamariza et al. [5]. In brief, trehalose was brominated in the 6-position by a variation of the Appel reaction using N-bromosuccinimide and triphenylphosphine, followed by acetylation. Subsequently, a nucleophilic substitution reaction with 3HC-2, after deprotonation of the 3-hydroxy group, followed by deacetylation afforded the anticipated 3HC-2-Tre. We then incubated M. aurum cells with 100 µM 3HC-2-Tre in liquid growth medium for 3 h at 36 °C and subjected the sample to fluorescence microscopy. As shown in Figure 9, our preliminary results demonstrate fluorescence of 3HC-2-Tre-labelled M. aurum cells with λ_ex = 485 nm and λ_em = 510–531 nm filter sets. It is interesting to note that Kamariza et al. came to the conclusion that 3HC-2-Tre-labelling of Mycobacterium smegmatis, another generally considered non-pathogenic, fast-growing mycobacterium and model organism for the pathogen M. tuberculosis [33], was not specific to the trehalose pathway (vide supra) [5]. Exploring the mechanism of the labelling of M. aurum cells with 3HC-2-Tre, however, is beyond the scope of this preliminary investigation.

3. Materials and Methods

3.1. General

Compounds 3HC-2 and 3HC-2-Tre were prepared following the procedures reported by Kamariza et al. [5]. The synthesis of the aldehyde precursor to 3HC-2, i.e., 6-(diethylamino)benzofuran-2-carbaldehyde, from 3-dimethylaminophenol is described in Ref. [16]. Experimental details of the preparation of 3HC-2-Tre can be found in the Supplementary Materials. The starting materials 3-dimethylaminophenol (97.69%, BLD Pharmatech GmbH, Kaiserslautern, Germany), 2′-hydroxyacetophenone (99.94%, BLD Pharmatech GmbH), and anhydrous trehalose (>98.0%, TCI, Chuo-ku, Tokyo) were purchased and used as received.

3.2. X-ray Crystallography

A few crystals of 3HC-2 suitable for single-crystal X-ray diffraction were obtained by chance when the remainder of a heptane/ethyl acetate solution resulting from flash chromatography evaporated to dryness in a tube. The crystals were coated with perfluoropolyether PFO-XR75 and mounted using a MiTeGen cryo-loop. The X-ray diffraction data were measured on a Bruker AXS D8 Venture diffractometer, equipped with an Incoatec IµS Diamond microfocus X-ray source, Incoatec multilayer optics, and a CMOS Photon III detector. The APEX4 software was used to control the diffractometer [34]. The raw data were processed with the SAINT software [35] and corrected for absorption effects with SADABS-2016/2 [36] using the Gaussian method based on indexed crystal faces.

The crystal structure was solved with SHELXT [37] and refined with SHELXL-2019/3 [38]. Anisotropic atomic displacement parameters (ADPs) were introduced for all non-hydrogen atoms. The diethylamino group was affected by positional disorder, which was taken into account by a split model for one ethyl group. Refinement of the ratio of occupancies by means of a free variable resulted in 0.531(4):0.469(4). Similar distance (SADI) and enhanced rigid-bond restraints (RIGU) [39] were applied to the disordered ethyl group. Except for the hydroxy hydrogen atom H2 and H3′ on the benzofuran moiety, hydrogen atoms were placed in geometrically calculated positions with C_aromat–H = 0.95 Å, C_methylene–H = 0.99 Å, and C_methyl–H = 0.98 Å and refined using a riding model with U_iso(H) = 1.2 U_eq(C) (1.5 for methyl groups). The initial torsion angle of the methyl group of C10 was determined in a circular Fourier calculation and subsequently refined while maintaining a tetrahedral structure. H2 and H3′ were located in difference Fourier maps and refined with the O2–H2 and C3′–H3′ distance restrained to target values of 0.84(2) and 0.95(2) Å, respectively, and with U_iso(H) = 1.2 U_eq(C, O). Crystal data and refinement details are listed in

Table 2. Crystal data and refinement details for 3HC-2.

Empirical Formula	C21H19NO4
Mr	349.37
T (K)	100(2)
λ (Å)	0.71073
Crystal system, space group	Triclinic, P-1
a (Å)	7.2022(5)
b (Å)	8.4003(5)
c (Å)	15.0621(10)
α (°)	99.969(4)
β (°)	94.673(4)
γ (°)	111.188(2)
V (Å3)	826.48(9)
Z, ρcalc (g cm−3)	2, 1.404
μcalc (mm−1)	0.098
F(000)	368
Crystal size (mm)	0.236 × 0.112 × 0.020
θ range (°)	2.665–30.555
Reflections collected/unique	91202/5061
Rint	0.0553
Observed reflections [I > 2σ(I)]	3946
Data/restraints/parameters	5061/29/261
Goodness-of-fit on F2	1.027
R1 [I > 2σ(I)]	0.0483
wR2 (all data)	0.1418
Δρmax, Δρmin	0.471/−0.556

. Structure pictures were drawn with Diamond [40]. Hirshfeld surface analysis was carried out using CrystalExplorer [41].

3.3. Computational Methods

DFT calculations were performed with ORCA (version 5.0) [42] with a B3LYP/G VWN1 hybrid functional (20% HF exchange) [43,44,45], using a def2-TZVPP basis set [45]. Optimization of the structure was completed using the BFGS method from an initial Hessian according to Almoef’s model with a very tight self-consistent field convergence threshold [46]. Calculations were made on the free molecule of 3HC-2. A relaxed surface scan was carried out varying the crystallographic O1–C2–C2′–O1′ torsion angle from 0 to 359° in 1° steps using a def2-TZVPP basis set and optimizing as above with a tight self-consistent field convergence criterion. Structures near the transition states were subsequently optimized using a def2-TZVPP basis set and a very tight self-consistent field convergence criterion, as for the optimized structure. The optimized local minimum-energy structures exhibited only positive modes and the transition states each exhibited one imaginary mode. Cartesian coordinates of the DFT-optimized structure of 3HC-2 can be found in the Supplementary Materials. Structure pictures were generated with Mercury [47].

3.4. Microbiology

M. aurum DSM 43999 was cultivated via the inoculation of a single colony from a fresh streak plate (Middlebrook 7H10 agar) into 10 mL of Middlebrook 7H9 liquid medium supplemented with 10% ADS [5% (m/v) bovine serum albumin fraction V, 0.81% (m/v) sodium chloride, 2% (m/v) dextrose in purified water] and 0.05% polysorbate 80. The culture was grown to an OD₆₀₀ between 0.2 and 0.8 for the labelling experiment by incubation at 36 °C with shaking (50 rpm). After dilution to an initial concentration of 5 × 10⁷ CFU/mL (OD₆₀₀ 0.1 = 1 × 10⁸ CFU/mL) with growth medium, the culture was transferred to a clear flat-bottom 96-well plate (Sarstedt, 83.3924.500) with 100 µL per well. Subsequently, 1 µL of 3HC-2-Tre (10 mM in DMSO) was added to each well to achieve a final dye concentration of 100 µM. After homogenization by pipetting, the plate was incubated for 3 h at 36 °C with shaking (50 rpm).

3.5. Fluorescence Microscopy

After incubation (see Section 3.4), the contents of the wells were homogenized by pipetting and 1 µL of each well was transferred to a black clear flat-bottom 96-well plate (Greiner bio one, 6550909) filled with 200 µL phosphate-buffered saline per well. Fluorescence microcopy was performed with a Thermo Scientific (Waltham, MA, USA) CellInsight CX5 instrument. Samples were excited at 485 nm and imaged with the GFP channel (510–531 nm).

4. Conclusions

We have structurally characterized the 3-hydroxychromone dye 3HC-2 in the solid state by X-ray crystallography. In the crystal, the molecule exhibits an s-cis conformation with an O–C–C–O torsion angle of ±13.76(16)°. DFT calculations on the free molecule gave an O–C–C–O torsion angle of ca. 0°, and revealed that the s-cis conformer was approximately 1.8 kcal mol⁻¹ more stable than the s-trans conformer, with a rotational barrier of ca. 7.5 kcal mol⁻¹. Preliminary labelling experiments demonstrated that M. aurum cells incubated in the presence of the dye trehalose conjugate 3HC-2-Tre could be detected by fluorescence microscopy in the GFP channel. We expect that this will facilitate in vitro testing of new antimycobacterial agents using this test bacterium, which is generally regarded as non-pathogenic.