Target-Directed Dynamic Combinatorial Chemistry Affords Binders of Mycobacterium tuberculosis IspE

In the search for new antitubercular compounds, we leveraged target-directed dynamic combinatorial chemistry (tdDCC) as an efficient hit-identification method. In tdDCC, the target selects its own binders from a dynamic library generated in situ, reducing the number of compounds that require synthesis and evaluation. We combined a total of 12 hydrazides and six aldehydes to generate 72 structurally diverse N-acylhydrazones. To amplify the best binders, we employed anti-infective target 4-diphosphocytidyl-2C-methyl-d-erythritol kinase (IspE) from Mycobacterium tuberculosis (Mtb). We successfully validated the use of tdDCC as hit-identification method for IspE and optimized the analysis of tdDCC hit determination. From the 72 possible N-acylhydrazones, we synthesized 12 of them, revealing several new starting points for the development of IspE inhibitors as antibacterial agents.


Assessment of amplification factors
The amplification factor for each product of the DCL was assessed at the time of equilibrium of the blank experiment.The relative peak area (RPA) was determined for each experiment; blank and proteintemplated (PT) duplicates, using ACD/Labs (Figure S2-S5, p. 12).All products of the DCL were integrated and their sum set to 100% in order to obtain the respective RPA for each product.The amplification factor was calculated for each PT duplicate individually as ). 2 The mean of the amplification factors and its standard deviation were reported for each product in form of bar charts with error bars, for DCL-1 in the main text (Figure 3A) and for DCL-2, -3, and -4 in Figure S9 (p.17).

General procedure for acylhydrazone formation (GP-3):
To an oven-dried round-bottom flask, under nitrogen atmosphere, the corresponding hydrazide (1.0 eq) and aldehyde (1.0-1.8 eq) were suspended in MeOH (0.02-0.2 M) and refluxed for 1-5 h; its progress was followed on TLC.After the reaction was completed, the solvent was removed under reduced pressure to afford the corresponding acylhydrazone product in 85 -99% isolated yields.

MtbIspE expression and purification
The synthetic gene encoding for MtbIspE was purchased from GenScript, GmbH cloned inside the plasmid pvp008, downstream a N-terminal StrepII tag followed by a TEV site.The plasmid was transformed into RB-competent E. coli Arctic express (DE3).Transformed E. coli cells were grown in LB medium at 37 °C until OD600 reached 0.6; after that, protein expression was induced adding 0.5 mM IPTG and the protein was expressed at 16°C for 48 hours.Cells were harvested by centrifugation at 5000 rpm for 20 minutes.The pellet was resuspended in wash buffer (50 mM HEPES pH 7, 100 mM NaCl, 5 mM MgCl2, 5 mM DTT) and sonicated for cell disruption.The lysate was cleared by centrifugation for 45 minutes at 16000 rpm at 4 °C.The supernatant was syringe-filtered and loaded onto a self-packed StreptActin-HC column equilibrated in wash buffer.MtbIspE was eluted with a single step 40 mL gradient at 100% elution buffer (50 mM HEPES pH 7, 100 mM NaCl, 5 mM MgCl2, 5 mM DTT, 5 mM desthiobiotin).Protein purity was verified via SDS-PAGE.The purification yield was 2 mg/L.

MtbIspE stability study via thermal shift assay
The stability of MtbIspE in the DCL buffer conditions was studied for 10 days by measuring its melting temperature (Tm) using thermal shift assay (TSA).MtbIspE incubated at a final concentration of 0.2 mg/mL in Tris buffer at pH 7.0 with 5%,-and 10% DMSO at room temperature and samples were measured in duplicate at t = 0, 1, 2, 6, 8 & 10 days (10% DMSO sampling stopped after day 6).The experiments were performed in a 96-well PCR plate (Thermoscientific).The final volume per well was 20 µL, consisting of 18 µL of MtbIspE sample (as mentioned above), and 2 µL of GloMelt dye (diluted 20x with H2O).The plate was centrifuged for 1 minute at room temperature at 1200 rpm.The Tm of the protein was measured using a Real-time PCR machine (Step one plus, Applied Biosystem).The conditions of the experiment were adjusted using Step One 2.3 software.The starting temperature, the ending temperature and the heating rate were set as 21 °C, 95 °C and 0.5 °C / min, respectively.The melting curves were analysed using Protein Thermal Shift 1.3 software.Tm of MtbIspE under these conditions was found to range from 50.4 °C to 51.3 °C, with one duplicate of day 6 (5% DMSO sample) discarded due to anomalies in the measurement.As expected MtbIspE is more stable and has higher Tm at 5% DMSO than 10% DMSO.The Tm of MtbIspE with 5% DMSO steadily declines from day 6 (Figure S1)

In silico elucidation of tdDCC hits' binding mode
Docking studies were performed using a pre-release of SeeSAR 13.1 (BioSolveIT, GmbH, Sankt Augustin, Germany), provided for beta-test by the company.Two published MtbIspE crystal structures in complex with ADP and CDP-ME (PDB accession code: 3PYF and 3PYE respectively) were superimposed and the 47 amino-acid in the binding region of both ADP and CDP-ME were used to define active site on 3PYF structure.All 72 possible product combinations (DCL-1, -2,-3, and-4) were docked on the active site defined as described above generating a maximum of 10 poses per molecule using standard docking parameters of SeeSAR software.The 720 poses generated were clustered by fragment and filtered basing on calculated torsion quality and molecular clashes.The binding pose of each fragment in the different molecules was visually analysed to determine whether a tendency in predicted binding pose of such fragments was found (Figure S10 -S12,p.11)

Kinetic turbidimetric solubility
The desired compounds were sequentially diluted in DMSO in a 96-well plate.7.5 µL of each well were transferred into another 96-well plate and mixed with 142.5 µL of PBS.Plates were shaken for 5 min at 600 rpm at room temperature (r.t.), and the absorbance at 620 nm was measured.Absorbance values were normalized by blank subtraction and plotted using GraphPad Prism 8.4.2 (GraphPad Software, San Diego, CA, USA).. Solubility (S) was quantified as the concentration leading to an absorbance value of 0.005 via determining the 'First X' value of the AUC function using a threshold of 0.005."

MtbIspE binding study via MicroScale Thermophoresis
MtbIspE buffer was exchanged to 50 mM HEPES pH 7, 100 mM NaCl, 5 mM MgCl2, 5 mM TCEP before protein labelling.The enzyme was labelled with Cy5 following the protocol indicated by NanoTemper.Labelled MtbIspE was diluted to 400 nM and used mixed 1:1 with 40 µM of each ligand.MicroScale Thermophoresis experiment was performed with a Monolith N.115 (Nanotemper Technologies, GmbH, Munich, Germany) using Standard Capillaries (Nanotemper Technologies, GmbH, Munich, Germany).Each binding event was evaluated at a final concentration of 200 nM MtbIspE and 20 µM of each ligand.All the measurements were performed in triplicates.Data analysis was automatically performed by Monolith NT.115Control and Analysis software (Nanotemper Technologies, GmbH, Munich, Germany).

MtbIspE binding study via thermal shift assay
The shifts in Tm of MtbIspE induced by the compounds were determined via TSA.MtbIspE (15 µM) was incubated with 20 µM of each compound for 30 min in 50 mM HEPES (pH 7.0, 100 mM NaCl, 5 mM MgCl2, 5 mM DTT).After that, SYPRO ORANGE fluorescent dye (Thermo Fisher) was added at a final concentration of 5X.Melting curves of each sample was measured increasing the temperature from 10 °C to 95 °C with an increase of 1 °C/30 sec and recording fluorescence of SYPRO ORANGE after every temperature increase using CFX96 real-time system-C1000 Thermal Cycler (Bio-Rad).Melting temperature (Tm) of each sample were calculated using CFX manager 2.0 software (Bio-Rad).Data analysis and visualization was performed using Excel and GraphPad 9.8.

Determination of E.coli activity
Growth inhibition values were determined in 96-well plates (Sarstedt, Nümbrecht, Germany) against Escherichia coli K12 and E. coli ∆tolC.As bacteria start OD600 0.03 was used in a total volume of 200 µL in lysogeny broth (LB) medium containing the compounds dissolved in DMSO (DMSO concentration in the experiment: 1%) at a concentration of 50 µM.The OD values were measured using a CLARIOstar platereader (BMG labtech, Ortenberg, Germany) after inoculation and after incubation for 18 h at 37 °C with 50 rpm.Based on these values, percent inhibition values were calculated in relation to the DMSO control.

Determination of in vitro anti-tubercular activity and solubility in 7H9 medium
Anti-tubercular tests were performed as previously described. 1In brief, 7H9 complete medium (BD Difco; Becton Dickinson, Maryland,USA) supplemented with 10% OADC (BD), 0.2% glycerol, and 0.05% Tween80 as previously described, 3 was used to culture Mycobacterium tuberculosis (Mtb) strain H37Rv (ATCC 25618) carrying a mCherry-expressing plasmid (pCherry10). 4Cultures were harvested at mid-log phase and frozen in aliquots at -80 °C.Prior to testing aliquots were thawed followed by centrifugation and the pellet was resuspended in 7H9 medium with 10% OADC (without glycerol and Tween80).This was further thoroughly resuspended by passing it through a syringe with a 26-gauge needle to avoid clumping of the bacteria.2×105 CFU (colony forming units) were then cultured in a total volume of 100 μl culture medium (triplicates) to test the non-precipitating compounds for the anti-tubercular activity at the concentrations indicated.For these assays, 96-well flat clear bottom black polystyrene microplates (Corning® CellBIND®, Merck, New York, USA) were used.Each plate had Rifampicin (at 1 μg/ml and 0.1 μg/ml) (National Reference Center, Borstel) as a reference compound.Plates were sealed with an air-permeable membrane (Porvair Sciences, Wrexham, UK) in a 37 °C incubator with mild agitation (TiMix5, Edmund Bühler, Germany).The activity of compounds was determined after 7 days by measuring the bacterial growth as relative light units (RLU) from the fluorescence intensity obtained at an excitation wavelength of 575 nm and an emission wavelength of 635 nm in a microplate reader (Synergy 2, BioTek Instruments, Vermont, USA).Two independent experiments (each in triplicates) were performed, and all values were normalized to untreated control sample (100%) in each experiment.

Cytotoxicity assay
To obtain information regarding the toxicity of our compounds, their impact on the viability of human cells was investigated.HepG2 cells (2x104 cells per well) were seeded in 96-well, flat-bottomed culture plates in 100 µL culture medium (DMEM containing 10% fetal calve serum, 1% penicillin-streptomycin).Twentyfour hours after seeding the cells, medium was removed and replaced by medium containing test compounds in a final DMSO concentration of 1%.Compounds were tested in duplicates at a single concentration or, for CC50 determination, at 8 concentrations that were prepared via 2-fold serial dilutions in 1% DMSO/medium.Epirubicin and doxorubicin were used as positive controls in serial dilutions starting from 10 µM, and rifampicin was used as a negative control (at 100 µM).The living cell mass was determined 48 h after treatment with compounds by adding 0.1 volumes of 3-(4,5-dimethylthiazol-2-yl)-2,5diphenyltetrazolium bromide (MTT) solution (5 mg/mL sterile PBS) (Sigma, St. Louis, MO) to the wells.After incubating the cells for 30 min at 37 °C (atmosphere containing 5% CO2), medium was removed and MTT crystals were dissolved in 75 µL of a solution containing 10% SDS and 0.5% acetic acid in DMSO.The optical density (OD) of the samples was determined photometrically at 570 nm in a PHERAstar Omega plate reader (BMG labtech, Ortenberg, Germany).To obtain percent viability for each sample, their ODs were related to those of DMSO controls.At least two independent measurements were performed for each compound.The calculation of CC50 was performed using the nonlinear regression function of GraphPad Prism 10 (GraphPad Software, San Diego, CA, USA).

Figure S5 .
Figure S5.Product peak assignment for tdDCC analysis.HPLC UV absorbance at 310nm of DCL-3 at 24 h of blank and protein-templated duplicates.The following products were not considered: G2, G3 and L2 (interference with injection peak).

Figure S6 .
Figure S6.Graphical determination of equilibrium in blank DCL-2.A) Evolution of relative peak areas (RPA) of formed acylhydrazones over time.B) Normalised RPAs over time.C) Change of RPA in each time interval.

Figure S7 .
Figure S7.Graphical determination of equilibrium in blank DCL-3.A) Evolution of relative peak areas (RPA) of formed acylhydrazones over time.B) Normalised RPAs over time.C) Change of RPA in each time interval.

Figure S8 .
Figure S8.Graphical determination of equilibrium in blank DCL-4.A) Evolution of relative peak areas (RPA) of formed acylhydrazones over time.B) Normalised RPAs over time.C) Change of RPA in each time interval.

Figure S11 .
Figure S11.Binding poses of hit-structures containing fragment 1, natural substrate CDP-ME and co-factor analog ADP.Molecules are represented with the following colours: C1 in blue; D1 in violet, H1 in yellow, I1 in pink, K1 in green, L1 in brown, CDP-ME and ADP in white.

Figure S12 .
Figure S12.Binding poses of hit-structures containing fragment E and , natural substrate CDP-ME and co-factor analog ADP.Molecules are represented with the following colours: E1 in blue, E4 in red, E6 in violet, CDP-ME and ADP in white.

Figure S13 .
Figure S13.Binding poses of hit-structure E1 and natural substrates.Molecules are represented with the following colours: E1 in blue, CDP-ME and ADP in white.

Figure S31 13 C
Figure S3113  C NMR spectrum of E1.
C NMR spectrum of B6.