Synthesis, antibacterial and anticancer activity, and docking study of aminoguanidines containing an alkynyl moiety

Abstract Two series of aminoguanidines containing an alkynyl moiety were designed, synthesised, and screened for antibacterial and anticancer activities. Generally, the series 3a–3j with a 1,2-diphenylethyne exhibited better antibacterial activity than the other series (6a–6k) holding 1,4-diphenylbuta-1,3-diyne moiety antibacterial activity. Most compounds in series 3a–3j showed potent growth inhibition against the tested bacterial strains, with minimum inhibitory concentration (MIC) values in the range 0.25–8 µg/mL. Compound 3g demonstrated rapid and persistent bactericidal activity at 2 × MIC. The resistance study revealed that resistance of the tested bacteria towards 3g is not easily developed. Molecular docking studies revealed that compounds 3g and 6e bind strongly to the LpxC and FabH enzymes. Moreover, excellent activity of selected compounds against the growth of cancer cell lines A549 and SGC7901 was also observed, with IC50 values in the range 0.30–4.57 µg/mL. These findings indicate that compounds containing the aminoguanidine moiety are promising candidates for the development of new antibacterial and anticancer agents.


Introduction
Multidrug-resistant (MDR) Gram-negative bacterial strains have arisen against all antibiotics in clinical use. Infections caused by these MDR bacteria threaten global public health 1,2 and are associated with high mortality rates. New antibacterial drugs with novel chemical scaffolds and targets are urgently required to combat infections due to drug-resistant strains 3,4 .
One of the validated antibiotic targets against Gram-negative bacteria is UDP-3-O-(R-3-hydroxymyristoyl)-N-acetylglucosamine deacetylase (LpxC), an essential enzyme catalysing the first committed step in the lipid A biosynthetic pathway. Because there is no gene homology in humans, inhibiting LpxC could result in the death of bacteria without causing side effects in the body 5 . Based upon these facts, LpxC has become a promising target for developing novel therapeutics against MDR Gram-negative pathogens [6][7][8] .
Studies of the enzymatic mechanism underlying LpxC have identified a hydrophobic tunnel that binds a myristate fatty acyl chain of the natural substrate and leads to a Zn 2þ active site responsible for deacetylation 9,10 . Since the discovery of L-161240 as the first LpxC inhibitor in the 1990s 11 , many LpxC inhibitors have been reported as antibacterial agents [6][7][8] . Most LpxC inhibitors share common structural features that mimic the natural Zn 2þ -binding substrate: (1) a hydroxamate head group, (2) a central linker, and (3) a lipophilic tail 12 . Among the well-characterised compounds, threonyl-hydroxamate derivatives, such as CHIR-12 and LPC-009, are representative LpxC inhibitors [13][14][15] . The hydroxamate group of these compounds occupies the active site, and their diphenyl acetylene or phenyl-diyne group penetrates the hydrophobic passage 16,17 .
Aminoguanidine is functional group with a high polarity and capacity for hydrogen bonding with many critical amino acid residues as well as metal ions. Many aminoguanidine derivatives exhibit antitumor activity via the formation of complexes with metal ions. In our previous work, some aminoguanidine derivatives were reported as potent antibacterial agents against Grampositive bacteria and Gram-negative bacteria including MDR clinical isolates [18][19][20] . The binding of these compounds involves a specific interaction with the b-ketoacyl-acyl carrier protein synthase III (FabH) enzyme 20 . By analysing the SAR of these compounds, it can be concluded that aminoguanidine combined with a hydrophobic moiety in the form of azomethine imine is the common structural requirement for their antibacterial activity.
In this contribution, we built upon the above observations by replacing the threonyl-hydroxamate group in the lead compounds CHIR-12 and LPC-009 with aminoguanidine ( Figure 1), yielding two series of aminoguanidines 3a-3j and 6a-6k. We anticipated that this design would promote binding with both FabH and LpxC, resulting in high and broad-spectrum antibacterial activity.
All the synthesised compounds were characterised by 1 H-NMR, 13 C-NMR, and high-resolution mass spectrometry, then evaluated for their antibacterial activity against 12 bacterial strains. To further characterise the antibacterial effects of compound 3g, the propensity for the development of bacterial resistance was determined and a bactericidal time-kill assay was carried out. Molecular docking studies of representative compounds 3g and 6e with LpxC and FabH were performed to understand the binding mechanism. Considering the reported anticancer activity of numerous compounds containing the guanidine moiety 21,22 , the anticancer activity was also evaluated for some selected compounds against two cancer cell lines (A549 and SGC7901). In this work, some aminoguanidines were discovered with promising antibacterial and anticancer activities.

Instruments and reagents
All the reagents and solvents were purchased from Aladdin (Shanghai, China) or Sinopharm Chemical Reagent Co. Ltd. (Shanghai, China), and were used as received. Melting points were determined in open capillary tubes and are uncorrected. Reaction courses were monitored by thin-layer chromatography (TLC) on silica gel-precoated F 254 plates (Merck, Darmstadt, Germany). Developed plates were examined with UV lamps (254 nm). Nuclear magnetic resonance spectroscopy was performed on an AV-400 spectrometer (Bruker, Zurich, Switzerland) operating at 400 MHz for 1 H and 100 MHz for 13 C and using DMSO-d 6 as the solvent and tetramethylsilane as the internal standard. Matrix-assisted laser desorption/ionization time-of-flight mass spectrometry (MALDI-MS) experiments were performed on a Bruker ultrafleXtreme MALDI-TOF/TOF mass spectrometer (Bruker Daltonik GmbH, Leipzig, Germany) equipped with a smartbeam II laser (1000 Hz).
After stirring overnight at room temperature, the resulting mixture was concentrated under vacuum. The residue was applied onto a silica gel column eluted with 1-2% ethyl acetate in petroleum ether to afford a yellow solid.
To a stirred solution of hydrazinecarboximidamide carbonate (124.03 mg, 0.91 mmol, 1.4 eq) in water (5 ml) was added NaOAc (74.78 mg, 0.91 mmol, 1.4 eq). After stirring for 0.5 h at room temperature, a mixture of 4-((substituted-phenyl)buta-1,3-diyn-1yl)benzaldehyde (0.65 mmol, 1 eq) in EtOH (5 ml) was added. Then the resulting solution was refluxed overnight. The reaction mixture was diluted with water (15 ml) and then cooled to room temperature. After stirred for 3 h, large amount of solids was precipitated. The solids were collected by filtration, washed with EtOH (2-0.5 ml), and then dried in an oven under reduced pressure to afford an off white solid.

Evaluation of antibacterial activity in vitro
Test bacteria were grown to mid-log phase in Mueller-Hinton broth (MHB) or Tryptone Soya Broth (TSB) and diluted 1000-fold in the same medium. Bacteria (10 5 CFU/mL) were inoculated into MHB or TSB and dispensed at 0.2 ml/well into a 96-well microtiter plate. As positive controls, gatifloxacin, moxifloxacin, norfloxacin, oxacillin, and penicillin were used. Test compounds were prepared in DMSO, the final concentration of which did not exceed 0.05%. The minimum inhibitory concentration (MIC) was defined as the concentration of a test compound that completely inhibited bacterial growth after 24 h incubation at 37 C. Bacterial growth was determined by measuring the absorption at 630 nm using a microplate reader. All experiments were carried out three times.

Evaluation of bacterial resistance development
To evaluate the propensity for developing bacterial resistance, one of the compounds with high antibacterial activity (3g) was used. The initial MIC values of 3g were determined against bacteria Staphylococcus aureus CMCC 25923 and Escherichia coli CMCC 44568, using norfloxacin and colistin, respectively, as antibiotic controls. Subsequently, serial passaging was initiated by transferring bacterial suspension grown at the sub-MIC of the compound/antibiotics (MIC/2) to a new plate and subjecting it to another assay to determine the (new) MIC. After 22 h incubation, cells grown at the sub-MIC of the test compound/antibiotics were once again transferred and the MIC determined. The process was repeated for 20 or 30 passages for S. aureus and E. coli, respectively. The MIC for 3g, norfloxacin and colistin was plotted as a function of time in days (number of passages) to determine the propensity of bacterial resistance development 24 .

Time-kill assay
Methicillin-resistant S. aureus ATCC 33591 grown in MHB was used to determine time-kill kinetics. Bacterial suspensions (10 5 CFU/mL) containing test compounds (norflocaxin, compound 3g) at final concentrations of 1 Â MIC and 2 Â MIC were incubated at 37 C with shaking. Aliquots (10 lL) were removed from the cultures after 0, 0.5, 1, 2, 3, 4, 6, 8 and 12 h, serially diluted 1000-fold in nutrient solution, and plated onto sterile Mueller-Hinton agar medium. Plates were then incubated for 24 h at 37 C, the number of CFU was counted, and the total bacterial log 10 CFU/mL was determined.

Evaluation of cytotoxicity activity in vitro
A lung cancer cell line (A549), gastric cancer cell line (SGC7901) and human hepatocytes (L02) were used to test the anticancer activity and cytotoxicity of the new compounds. 5-Fluorouracil (5-FU) was used as the positive control against cancer cell lines. The A549, SGC7901 and L02 cells were grown in Dulbecco's modified Eagle's medium (DMEM) supplemented with 10% foetal bovine serum (FBS) and antibiotics (100 U/mL penicillin-streptomycin). Cells at 80-90% confluence were split by trypsin (0.25% in PBS; pH 7.4), and the medium was changed at 24 h intervals. The cells were cultured at 37 C in a 5% CO 2 incubator. The cells were passaged three times, then approximately 1 Â 10 4 cells were seeded into each well of a 96-well plate and allowed to incubate to allow attachment of the cells to the surface. After 24 h, the medium was replaced with DMEM supplemented with 10% FBS containing various concentrations (0.1, 0.3, 1, 3, 10, and 30 lg/mL) of test compounds. Each concentration was tested in triplicate. After 48 h treatment, 20 mL of CCK-8 solution was added to each well and the optical density measured at 450 nm after 3 h using a microplate reader. The IC 50 values were defined as the concentrations inhibiting 50% of cell growth.

Docking studies
Molecular docking of compounds 3g and 6e to Pseudomonas aeruginosa LpxC (PDB code: 3P3E), E. coli LpxC (PDB code: 3P3G) and the E. coli FabH-CoA complex structure (PDB code: 1HNJ) was carried out using the DS-CDOCKER protocol implemented through the graphical user interface of the Discovery Studio software (version 2019). The structures of 3P3E, 3P3G, and 1HNJ were downloaded from Protein Data Bank. The three-dimensional structures of 3g and 6e were constructed using Chem3D Ultra 12.0 software (Chemical Structure Drawing Standard; CambridgeSoft Corporation, Waltham, MA, 2010) and was energetically minimised using the MMFF94 force field with 5000 iterations and a minimum RMS gradient of 0.10. For protein preparation, the hydrogen atoms were added and water and impurities were removed. The 3D structure of 3g or 6e was placed during the molecular docking procedure. Types of docking interactions of the proteins with tested compounds were analysed and ranked, and those with maximum binding energy were selected to analyse the interaction patterns.
Spectroscopic analyses of all synthetic compounds fully supported their depicted structures. Taking compound 3a as an example, the 1 H-NMR spectrum yielded two singlets due to N-H of guanidyl at 5.65 and 6.04 ppm, which were assigned to two NH 2 groups involved in the tautomerism of the guanidyl group. Peaks corresponding to the aromatic protons of terminal benzene ring were observed in the range 7.42-7.57 ppm. Two doublets (J ¼ 8.2 Hz) due to aromatic protons of the para-substituted phenyl ring were observed at 7.50 and 7.72 ppm. The absorption peak of the C-H in imine was found at 7.99 ppm. The 13 C-NMR spectra also identified 12 carbon nuclei in different chemical environments, which was also consistent with the structure of 3a. Moreover, high-resolution mass spectrometry of 3a displayed an [M þ H] þ signal at m/z 263.1287, in agreement with its molecular weight of 263.1291.

In vitro antibacterial activity
The evaluation of antibacterial activity of all compounds (3a-3j and 6a-6k) was performed by a serial dilution method to determine the MIC against a panel of pathogens comprising five  Tables 1 and 2. Gatifloxacin, moxifloxacin, norfloxacin, oxacillin, and penicillin were used as positive controls.
In general, the inhibitory activity of the new target compounds against Gram-positive strains was better than that against Gramnegative strains. Compared with compounds 6a-6k with a 1,4diphenylbuta-1,3-diyne moiety, compounds 3a-3j containing a 1,2-diphenylethyne moiety exhibited good to excellent antibacterial activity against all strains except P. aeruginosa CMCC 27853. Pseudomonas aeruginosa are able to rapidly develop resistance to multiple classes of antibiotics, making the treatment of infectious diseases becomes more challenging. The outer membrane porin OprD and the multidrug efflux pumps of P. aeruginosa represent the main barriers for drug entry into the cell. The decrease or loss of antibacterial activity of most compounds synthesized against P. aeruginosa CMCC 27853 may be due to the above facts 25 .
All compounds in series 3a-3j showed potent inhibitory effects against the five Gram-positive strains with MICs in the range 0.25-8 lg/mL, except compound 3j, which showed inhibitory activity at 128 lg/mL against B. subtilis CMCC 63501. In general, compounds 3e, 3f, 3g, 3j, containing 2-Cl, 3-Cl, 4-Cl, and 4-Br, respectively, had greater inhibitory activity against Gram-positive strains. Compounds 3a-3j were less active against the four Gramnegative strains, with MICs ranging from 1 mg/mL to more than 128 mg/mL. For E. coli CMCC 25922 and CMCC 44568, and P. aeruginosa CMCC 10104, the change of substituents had no obvious effect on the inhibitory activity, while for the P. aeruginosa CMCC 27853, the compounds with H, 3-F, 4-F groups had more potency than compounds with other substituents. Among the nine strains, compounds 3a-3j presented the most potent inhibitory activity against E. faecalis CMCC 29212 with a MIC of 0.5 or 1 lg/mL, which is comparable to gatifloxacin, moxifloxacin and norfloxacin (MIC ¼ 1 mg/mL) and is 256 or 128-fold more potent than oxacillin and penicillin (MIC ¼ 128 mg/mL). Compound 3g showed the most potent inhibitory activity against B. subtilis CMCC 63501 (MIC ¼ 0.25 lg/mL).
Compounds in the series of 6a-6k displayed different degrees of inhibitory activity (MICs ranging from 0.25 mg/mL to more than 128 mg/mL) against the five Gram-positive strains and two Gramnegative strains (E. coli CMCC 25922 and P. aeruginosa CMCC 10104), while all of them were insensitive towards E. coli CMCC 44568 and P. aeruginosa CMCC 27853. The structure-activity relationship showed that the para-position substituents reduce the antibacterial activity, giving compounds 6d, 6g, 6i a MIC !128 against almost all of the tested strains. Compounds 6b and 6e, with F and Cl in the ortho position, showed the most potent inhibitory activity against S. aureus CMCC(B) 26003 (MIC ¼ 0.25 lg/mL). Compound 6a, without substituents on the terminal phenyl group, was more effective against P. aeruginosa CMCC 10104 than other compounds in this series.

Propensity to develop bacterial resistance
Bacterial resistance against most antibiotics presents a major problem of current times [26][27][28][29] . Therefore, it is crucial to evaluate the potential emergence of bacterial resistance against these biocides. The propensity for development of bacterial resistance of the synthesised compounds was evaluated by using the most active compound (3g) against both Gram-positive S. aureus and Gramnegative E. coli 30,31 . Norfloxacin, an antibiotic generally used to treat the Gram-positive infections, was used as a positive control for S. aureus, whereas colistin, a lipopeptide antibiotic active against Gram-negative bacteria, was used as a positive control for E. coli. These antibacterial agents were repeatedly challenged against bacteria at their sub-MIC values to allow bacteria to develop resistance. Resistance is usually defined as a > 4-fold increase in the original MIC 32 . Interestingly, little change in the MIC of compound 3g was observed over 20 generations. In comparison, an approximately 32-fold and 4 or 8-fold increase in MIC was observed for norfloxacin and colistin, respectively ( Figure 2). The above results indicated that bacteria do not develop resistance against this type of aminoguanidine within the experimental time period.

Bactericidal time-kill kinetics
To study the bactericidal activity of the promising compound 3g, next we carried out in vitro time-kill assay against MRSA (starting bacterial concentration of 5.5 log10 CFU/mL) at two different concentrations (1 Â MIC and 2 Â MIC) using norfloxacin as a control ( Figure 3). Compound 3g was rapidly bactericidal at 2 Â MIC (>5 log10 CFU/mL reduction) after 2 h and its bactericidal activity persisted for 12 h. In the case of norfloxacin, a concentrationdependent activity was seen at 1 Â MIC to 2 Â MIC, but its effects were bacteriostatic, not bactericidal ( Figure 3). These results clearly demonstrate the superiority of compound 3g over the commonly used antibiotic norfloxacin in killing MRSA bacteria.

Molecular modelling
LpxC is an essential enzyme in the lipid A biosynthetic pathway. Developing novel LpxC inhibitors has been an important approach to obtain new antibacterial drugs targeting Gram-negative pathogens. To gain insight into the molecular interactions of these compounds with LpxC, the co-crystal structures of 3g and 6e complexed with P. aeruginosa LpxC and E. Coli LpxC were obtained. Generally, the 1,2-diphenylethyne group in the series of 3a-3j and 1,4-diphenylbuta-1,3-diyne moiety in the series of 6a-6k functionally interacted with many of the hydrophobic residues in the lipophilic tunnel. The aminoguanidine was responsible for forming hydrogen bonds with several amino acid residues. The aminoguanidine of compound 3g bound to the active site Zn 2þ ion of P. aeruginosa LpxC (Figure 4(A)) and formed hydrogen bonds with ASP241, HIS264, MET62, and GLU77 that line this polar region. In the interaction of P. aeruginosa LpxC with compound 6e (Figure 4(B)), the phenyl ring of 1,4-diphenylbuta-1,3-diyne moiety showed interactions with critical amino acid residues LEU18, ILE197, and VAL216 via aromatic stacking and hydrophobic intermolecular forces. The guanidine group of compound 6e acted as a hydrogen bond donor in the interaction with the carbonyl group of ASP241, HIS264 and MET62.
Interactions of compounds 3g and 6e with E. coli LpxC are shown in Figure 5(A,B), respectively. The aminoguanidine of compound 3g was bound to the active site Zn 2þ ion and forms hydrogen bond interactions with ASP242 and HIS265. The 1,2diphenylethyne group was responsible for forming various aromatic stacking interactions and hydrophobic intermolecular forces with LEU18, ALA215, VAL217, GLY210, SER211, ALA215, and ILE198. In addition, the Cl atom attached to the terminal phenyl group showed hydrophobic interactions with amino acid residues PHE212 and MET195 that increased the binding force with E. coli LpxC. The aminoguanidine of compound 6e was also bound to the zinc atom of E. coli LpxC, as well as the amino acid residues ASP242, GLU78, and HIS265 via hydrogen bond interaction. Aromatic stacking interactions and hydrophobic force interactions were additionally formed between the hydrophobic 1,4-diphenylbuta-1,3-diyne moiety of 6e and LEU18, CYS207, ALA215, LEU62, MET195, GLY210, SER211, PHE212, and VAL217.
For comparison, the interactions of compound 6e with P. aeruginosa LpxC and E. coli LpxC were superimposed on the co-crystallised LPC-009. As seen in Figure 6, compound 6e (red) showed similar interactions as LPC-009 (green). This suggests that binding of 6e to LpxC may be, at least in part, responsible for its antibacterial activity.
It is likely that targets other than LpxC might be involved in the broad-spectrum antibacterial activity of the synthesised compounds against Gram-positive and Gram-negative bacteria. The FabH receptor is a condensing enzyme that plays key roles in fatty acid biosynthesis 33 . The potential interaction of some hydrazine compounds with FabH prompted us to investigate the molecular interactions of the representative compounds 3g and 6e with FabH receptor (PDB ID: 1HNJ) 20,34,35 .
As shown in Figure 7(A), residues including ASP27, THR28, ARG151, VAl212, ALA216, ILE250, and ALA246, were involved in the binding of 3g to the active site of E. coli FabH. The C ¼ N and NH groups of compound 3g were involved in the interaction with ASP27, THR28, ARG151 via hydrogen bonding, while the terminal phenyl group and its attached chlorine atom formed hydrophobic interactions with VAl212, ALA216, ILE250, and ALA246. The binding of 6e with E. coli FabH resembled that of 3g. As shown in Figure 6(B), residues ASP27, THR28, ARG151, TRP32, VAl212, ALA216, ILE250, and ALA246, were involved in the interactions with 6e in E. coli FabH enzyme.

Anticancer activity
The cytotoxic effects of compounds 3e, 3f, 3g, 3i, 6a, 6b, 6e and 6k were evaluated using two human cancer cell lines (A549and SGC7901) and one human normal cell line (L02). The IC 50 values of the tested compounds and 5-FU are shown in Table 3. All the tested compounds showed excellent activity against the investigated cancer cells (IC 50 ¼ 0.30-4.57 mg/mL); however, no correlation between the substituents and the cytotoxic activity could be    identifies. The highest activity against A549 cells was exhibited by compound 6b (IC 50 ¼ 2.22 mg/mL), although it is not comparable to that of 5-FU, which exhibited IC 50 of 0.88 mg/mL. The highest activity against SGC7901 cells was exhibited by compound 3f (IC 50 ¼ 0.30 mg/mL), followed by compounds 3e and 6k (IC 50 ¼ 0.45, 0.55 mg/mL, respectively). All the tested compounds showed higher cytotoxic activity against SGC7901 cells than 5-FU. Compounds 3e, 3f, 3g, 3i, 6a, 6b, 6e and 6k showed IC 50 values in the range of 10.25-20.85 mg/mL against the normal cell line L02, comparing to the IC 50 value of 8.44 mg/mL of 5-FU. This result indicated that these compounds have low toxicity towards normal cells in comparison to cancer cells, suggesting potential for a good therapeutic index.

Conclusion
In the present work, a series of novel aminoguanidines containing an alkynyl moiety were synthesised and characterised. The antibacterial and anticancer activities of these compounds were screened, with Gram-positive bacteria being more susceptible than Gram-negative ones. The aminoguanidine derivatives (3a-3j), having a 1,2-diphenylethyne, exhibited greater antibacterial activity than 6a-6j with a 1,4-diphenylbuta-1,3-diyne moiety. Compounds 3a-3j showed potent inhibitory activity against the selected bacterial strains with MIC values in the range of 0.25-8 lg/mL, including the multidrug resistant strains. Among them, compound 3g was the most promising, having superior activity to oxacillin and penicillin against the tested MDR strains. Resistance of the tested bacteria towards 3g was not easily developed and this compound was rapidly bactericidal. Furthermore, 3 g exhibited significant anticancer activity against lung (A549) and gastric cancer cells (SGC7901), with IC 50 values of 4.42 and 1.01 mg/mL, respectively, and low-toxicity towards normal cells. To understand the binding pattern, molecular docking of representative compounds 3g and 6e was performed, demonstrating that they bind strongly to the LpxC enzyme and FabH enzyme. These findings indicate that compounds containing the aminoguanidine moiety are promising candidates for the development of new antibacterial and anticancer agents.