Ligand-based discovery of novel N-arylpyrrole derivatives as broad-spectrum antimicrobial agents with antibiofilm and anti-virulence activity

Abstract

The escalating threat of antimicrobial resistance calls for novel therapeutic agents. This study employed a ligand-based design approach to develop three series of N-arylpyrrole derivatives (Va–e, VIa–e, and VIIa–e), refined through molecular modeling. Synthesized compounds were evaluated against ESKAPE pathogens, MRSA, and Mycobacterium phlei. Series Va–e showed the most promise, with compounds Vb, Vc, and Ve outperforming levofloxacin against MRSA (MIC = 4 μg/mL vs. 8 μg/mL). Vc also exhibited activity against E. coli, K. pneumoniae, and A. baumannii, and showed significant inhibition against M. phlei (MIC = 8 μg/mL). Compounds were evaluated for antibiofilm and antivirulence properties, targeting resistance mechanisms linked to infection persistence and dissemination. Most exhibited broad-spectrum biofilm inhibition and antivirulence activity. Cytotoxicity studies revealed selectivity for bacterial cells. ADMET studies supported drug-like properties. Docking studies suggested UPPP inhibition as the potential mechanism. SAR analysis was conducted to support future optimizations.

Graphical abstract

Introduction

Antimicrobial resistance poses a significant threat to global health, as antibiotics are indispensable for basic and modern medical practices. Cancer chemotherapy, invasive surgeries, organ transplants, and complicated deliveries may only be performed without the danger of deadly infections if appropriate anti-bacterial medicines are available¹. Projections suggest a staggering 10 million deaths attributable to infectious disease treatment failure by 2050².

Among the most concerning are the ESKAPE pathogens, a group comprising Enterococcus faecium, Staphylococcus aureus, Klebsiella pneumoniae, Acinetobacter baumannii, Pseudomonas aeruginosa, and Enterobacter species. These bacteria, collectively dubbed “ESKAPE” due to their remarkable ability to escape the effects of antibiotics, have become emblematic of the challenges posed by antimicrobial resistance³. Of particular concern within this group is methicillin-resistant Staphylococcus aureus (MRSA), which exemplifies the significant resistance seen in certain bacterial strains. MRSA has developed resistance to various antibiotics, including beta-lactams, aminoglycosides, glycopeptides, and fluoroquinolones⁴. The global prevalence of MRSA is alarming, with numerous reports indicating a substantial proportion of Staphylococcus infections resistant to methicillin^5,6.

Furthermore, gram-negative pathogens such as Pseudomonas aeruginosa, Acinetobacter baumannii, Klebsiella pneumoniae, and Escherichia coli are increasingly prevalent and demonstrate resistance to conventional therapy. Pseudomonas aeruginosa, in particular, poses a global threat due to its rapid development of resistance during treatment, often facilitated by forming biofilm aggregates^7–9. Similarly, certain strains of Acinetobacter baumannii exhibit multidrug resistance, potentially attributable to their adeptness in deploying diverse resistance mechanisms¹⁰. Epidemic strains of K. pneumoniae have emerged resistant to carbapenem antibiotics¹¹ and fluoroquinolones¹². Moreover, infections caused by Escherichia coli, particularly urinary tract infections, are increasingly resistant to third-generation cephalosporins in certain regions, complicating treatment¹³.

To confront the swiftly escalating crisis of resistance, it is imperative to undertake extensive research and develop innovative antimicrobial agents. This research aims to contribute to global efforts in combating antimicrobial resistance and developing effective therapies for infectious diseases.

A literature review exploring effective chemical moieties with potent anti-bacterial activity, particularly against resistant strains, has identified pyrroles as a promising antimicrobial scaffold. Various pyrrole derivatives have been synthesised and evaluated for their antimicrobial properties. For example, tetra-substituted pyrrole derivatives were synthesised and tested against Escherichia coli and S. aureus (1, Figure 1)¹⁴. Furthermore, diphenyl pyrrole compounds exhibited notable activity against drug-resistant gram-positive and gram-negative bacteria, comparable to or surpassing levofloxacin (2, Figure 1)¹⁵. Notably, pyrrole derivatives have shown significant anti-tubercular activity as well. For instance, BM212, a 2,5-dimethylpyrrole derivative currently undergoing clinical phase II trials, demonstrates potent activity against various mycobacteria, including multidrug-resistant (MDR) tuberculosis strains (3, Figure 1)¹⁶.

Figure 1.

Tetra-substituted pyrrole-based derivatives (1), 1,5-diphenyl pyrrole derivatives (2), antitubercular drug BM212 (3), phenylthiazole lead compound (4), phenylpyrazole derivatives (5), and (thio)semicarbazide derivatives (6,7).

Apart from pyrroles, a highly active phenylthiazole scaffold with significant anti-bacterial effects against gram-positive bacteria, including 18 MRSA and VRE strains, was identified.¹⁷. Structure–activity relationship (SAR) analysis revealed a thiazole nucleus attached to two key structural components: a highly lipophilic moiety at C-2, corresponding to greater potency, and a cationic moiety at C-5 essential for anti-bacterial activity (4, Figure 1)¹⁸. Despite its anti-MRSA activity, this scaffold faced pharmacokinetic limitations^17,19. The central thiazole ring of this scaffold was replaced with a pyrazole ring while retaining the other pharmacophoric features, resulting in compounds with notable activity against gram-positive and gram-negative bacteria (5, Figure 1)¹⁷.

Furthermore, our literature review revealed the bio-potential of semicarbazides and thiosemicarbazide derivatives as non-toxic anti-bacterial agents (6,7) (Figure 1)^20–23. Incorporating these derivatives into our design may yield active anti-bacterial agents.

In this work, we hypothesised that incorporating a dimethylpyrrole ring while maintaining the other pharmacophoric features of the phenylthiazole lead could result in compounds with a broad spectrum (Figure 2). Additionally, we utilised semicarbazide and thiosemicarbazide for the cationic part of our scaffold, aiming to potentially enhance the interaction of the compounds with bacterial targets and improve anti-bacterial potency.

Figure 2.

Design strategy.

A ligand-based preliminary evaluation of the designed compounds using 3DQSAR modelling and ligand pharmacophore mapping protocols was done to predict the potential activity of the newly designed scaffold against MRSA. Subsequently, the designed compounds were synthesised, purified, and structurally confirmed using various analytical and spectral techniques.

The anti-bacterial activities of the synthesised compounds were assessed against a panel of bacteria, including ESKAPE pathogens, including MRSA, and anti-tubercular activity against Mycobacterium phlei. Their antivirulence properties were also assessed using biofilm inhibition and virulence factor assays. Furthermore, the cytotoxicity of the four most potent compounds (Vb, Vc, Vd, and Ve) was assessed.

Additionally, in silico target prediction and docking studies were performed to gain preliminary insights into the potential mechanism of action. ADMET studies were conducted to evaluate the synthesised derivatives’ physicochemical, pharmacokinetic, and toxicity parameters. Finally, a SAR study was established to guide future structural modifications.

Results and discussion

Preliminary evaluation of the designed compounds using computer-aided molecular modelling

Ligand-based computer-aided drug design is used to design and evaluate new drug leads by employing ligands with known activity to help develop new biologically active compounds, allowing us to make better-informed decisions, hence exhausting fewer resources^24,25.

This work used consecutive computer-aided drug design protocols to preliminary evaluate our newly designed compounds. First, we generated a ligand-based 3D QSAR pharmacophore using a dataset of phenylthiazole and phenylpyrazole compounds with known activity (MIC) against MRSA^26–28. The HypoGen algorithm summarised the structural features of these compounds to generate a valid predictive pharmacophore model using the Discovery Studio 2024 software package²⁹. Then, the phenylthiazole lead used to design our molecules¹⁷ was mapped on the generated pharmacophore model (Hypo1) via the ligand pharmacophore mapping protocol for validation. Next, the designed compounds were mapped on the validated pharmacophore model Hypo1 to predict their activity.

3D QSAR pharmacophore model generation

3D QSAR (Three-Dimensional Quantitative Structure–Activity Relationship) pharmacophore model was generated using DS by utilising a dataset of 32 phenylthiazole and phenylpyrazole compounds with known activity (MIC) against MRSA that were extracted from literature^26–28. The data set was randomly divided into 22 compounds as a training set (Figure 3) and 10 as a test set (Figure 4). The HypoGen algorithm generated pharmacophore models with the essential features responsible for the anti-MRSA activity. Ten hypotheses were generated using the HypoGen algorithm and validated for their statistical significance and accuracy of prediction (Table 1). After careful review, Hypo1 was selected as the most valid based on the cost difference, root mean square (RMS) difference, and correlation coefficient hypothesis and used for this study.

Figure 3.

Training set ligands along with their MIC values (μg/mL).

Figure 4.

Test set ligands along with their MIC values (μg/mL).

Table 1.

Statistical parameters of the top 10 generated pharmacophore models.

Hypo no.	Maximum fit	Total cost	Cost distance	RMS	Correlation coefficient (r)	Features
1	9.89	96.60	102.71	1.58	0.90	HBA, HYD, HYD, HYD, RA
2	8.66	101.52	97.797	1.70	0.88	HBA, HYD, HYD, HYD, RA
3	7.35	101.78	97.444	1.68	0.88	HYD, HYD, HYD, PI, RA
4	6.03	123.06	76.305	2.18	0.80	HBA, HYD, HYD, RA
5	6.05	123.15	76.165	2.19	0.80	HYD, HYD, HYD, PI, RA
6	6.25	125.44	73.878	2.22	0.79	HBA, HYD, HYD, HYD, RA
7	7.91	125.46	73.858	2.22	0.82	HBA, HYD, HYD, HYD, RA
8	5.86	127.76	72.174	2.24	0.78	HBA, HBA, HYD, HYD
9	4.92	128.63	70.683	2.31	0.77	HBA, HYD, RA
10	5.66	129.67	69.642	2.26	0.78	HBA, HYD, HYD, HYD, RA

Validation of selected pharmacophore model

The Discovery Studio software employed the HypoGen algorithm to compute three validation cost functions. The first, termed the fixed cost, assumes a straightforward hypothesis model that smoothly accommodates all dataset and library molecules, resulting in the lowest cost³⁰. The fixed cost value for the generated hypothesis is equal to 69.07. Conversely, the null cost represents the highest potential error cost³¹. The null cost for the Hypotheses generated is equal to 199.32. Each pharmacophore hypothesis undergoes an independent calculation of total cost, comprising weight, error, and fixed costs³². Among the 10 generated hypotheses, total costs ranged from 96.6 to 129.67. Hypo1 displayed the most significant cost difference of 102.71 (Table 1), indicating that the pharmacophore model is more than 90% statistically significant³³. Furthermore, Hypo1 also scored the highest correlation coefficient value of 0.90 (Table 1S) and the lowest RMS value of 1.58, both attributed to a superior capacity for biological activity prediction³³.

Furthermore, the predictive ability of the pharmacophore models was assessed by mapping a test set of 10 diverse compounds with known MIC values against MRSA using the Ligand Pharmacophore Mapping tool. A strong correlation (R² = 0.76) between estimated and reported activity confirmed the model’s reliability (Table 2S). The selected pharmacophore model (Hypo1) (Figure 5) consists of five key features: one hydrogen bond acceptor (HBA), three hydrophobic (HYD), and one ring aromatic (RA). Hypo1 was subsequently used to evaluate the novel scaffold.

Figure 5.

Spatial arrangement of the valid pharmacophore model (Hypo1) with the distances and angles displayed.

Mapping of the lead compound on the valid generated model

The phenylthiazole lead compound was also mapped via the Ligand Pharmacophore Mapping protocol in DS on the chosen pharmacophore model (Hypo1). The lead compound successfully mapped into the pharmacophore’s five features (Figure 6) with a high fit value of 7.8 and estimated activity of 7 µg/mL, further validating Hypo1.

Figure 6.

The mapping of the phenylthiazole lead into the generated pharmacophore model (Hypo1).

Ligand pharmacophore mapping

To preliminary evaluate our newly designed scaffold, the proposed compounds were mapped into the generated pharmacophore using the Ligand Pharmacophore Mapping protocol. All the newly designed compounds were mapped successfully into at least four features of the valid generated pharmacophore model, indicating potential anti-bacterial activity (Table 3S). The fit values ranged between 3.3 and 6.8. We noticed that elongating the compound’s lipophilic tail (R₂) allowed it to fit into all five pharmacophore features. This note was incorporated into our design as it might indicate higher activity (Figure 7).

Figure 7.

The mapping of compounds Va and Vc into the generated pharmacophore model.

Chemistry

Scheme 1 includes the synthesis of the intermediates IVa–e. First, the Paal–Knorr condensation reaction of 2,5-hexanedione (I) and different anilines IIa–c in the presence of sulphamic acid as a catalyst yielded the N-aryl-2,5-dimethylpyrrole derivatives IIIa–c³⁴. Second, the N-aryl-2,5-dimethyl-1H-pyrrole-3-carbaldehydes IVa–c and the 2,5-dimethyl-1-aryl-1H-pyrrol-3-yl-ethanones IVd–e were obtained via the simple and convenient Vilsmeier–Haack method in good yields^35,36. All the products were identified using TLC and melting points that matched the reported ones^37–43. The intermediates IVa–e were also confirmed by ¹H NMR and IR spectroscopy, which was in accordance with the literature^39,44–46. The ¹H NMR spectrum of the novel intermediate IVc showed two singlet signals at δ 1.93 and 2.29 ppm corresponding to the methyl groups at position five and two on the pyrrole ring, respectively. The ¹H NMR spectrum also showed one singlet at δ 6.3 ppm corresponding to the pyrrole aromatic proton and an aldehydic proton at δ 9.88 ppm. In addition, the IR spectra for the compound showed signal at 1730 cm⁻¹ for the C═O group.

Scheme 1.

Reagents and conditions: (1) sulphamic acid, rt, 1–6 h; (2) POCl₃, ice cold DMF, 60$\mathrm{°C}$, 4 h, 20% NaOH; (3) POCl₃, ice cold DMA, reflux, 6 h, 20% NaOH.

Scheme 2 shows the reaction of the intermediates IVa–e with aminoguanidine carbonate, thiosemicarbazide HCl, and semicarbazide HCl, thus replacing the acetyl/formyl moiety in the intermediates IVa–e with various cationic tails, such as aminoguanidine, semicarbazide, and thiosemicarbazide, affording the desired final compounds Va–e, VIa–e, and VIIa–e, respectively. The reaction is based on nucleophilic addition on the carbonyl group of the appropriate aldehyde or ketone; the partial positive charge on the carbonyl atom makes it susceptible to attack by a nucleophile like a primary amine forming imine derivatives, also known as Schiff bases (compounds having a C═N function)⁴⁷. The structures of compounds Va–e, VIa–e, and VIIa–e were supported by elemental analyses, high-resolution mass spectrometry (HRMS), and spectroscopic data: IR, ¹H NMR, and ¹³C NMR.

Scheme 2.

Reagents and conditions: (4) aminoguanidine HCO₃, conc. HCl, absolute ethanol, reflux, 20–24 h; (5) thiosemicarbazide HCl, anhydrous sodium acetate, absolute ethanol, reflux, 12 h; (6) semicarbazide HCl, anhydrous sodium acetate, absolute ethanol, reflux, 12 h.

The IR spectra of all series compounds showed IR signals at 3289–3565 cm⁻¹ corresponding to the NH/NH₂ groups. Compounds Va–e showed IR signals at 1582–1670 cm⁻¹ corresponding to the C═N group in the aminoguanidine moiety, while compounds VIa–e showed IR signals at 1482–1521 cm⁻¹ corresponding to the C═S group in the thiosemicarbazide moiety, and compounds VIIa–e showed IR signals at 1681–1687 cm⁻¹ corresponding to the C═O group in the semicarbazide moiety.

The ¹H NMR spectra of compounds VIa–e showed new exchangeable singlet signals at δ 5.51 and 5.8 ppm, corresponding to the NH and NH₂ groups. Compounds VIa–e and VIIa–e displayed one singlet at δ 6.17 ppm corresponding to the NH₂ moiety and another at δ 9.02–9.72 ppm corresponding to the NH group. Furthermore, ¹H NMR for compounds Va–c, VIa–c, and VIIa–c showed the disappearance of the aldehydic proton at δ 9.88 ppm and the appearance of a singlet at δ 7.99–8.75 ppm corresponding to the proton at the CH═N. While compounds Vd–e, VId–e, and VIId–e showed an extra singlet at δ 2.16–2.3 ppm corresponding to the methyl group at R₂.

The Va–e, VIa–e, and VIIa–e structures were further assured by ¹³C NMR spectra, which shared the same essential peaks at δ 140.8–158 ppm corresponding to the C═NH moiety. Compounds Va–e showed a peak at δ 158.3–158.7 ppm corresponding to the C═N group in the aminoguanidine moiety, while compounds VIa–e showed peaks at δ 194.2–199.2 ppm corresponding to the C═S group in the thiosemicarbazide moiety, and compounds VIIa–e showed peaks at δ 158 ppm corresponding to the C═O group in the semicarbazide moiety. Compounds Vd–e, VId–e, and VIId–e showed an extra signal at δ 16.1–16.4 ppm corresponding to the methyl group at R₂.

Biological evaluation

Antimicrobial activity

All the synthesised compounds were tested for their biological activity by determining their MIC values against a panel of ESKAPE pathogens, including MRSA, and anti-tubercular activity against Mycobacterium phlei using the broth microdilution method. In addition, all compounds were tested for their anti-tubercular activity against clinical isolates of Mycobacterium phlei. The anti-bacterial drugs levofloxacin and rifampicin were used as reference standards. Compounds displaying MIC values lower or equal to the reference standard drug were considered promising. The results of the antimicrobial evaluation are presented in Table 2.

Table 2.

The minimum inhibitory concentration (MIC in μg/mL) of the all the synthesised compounds against a panel of susceptible and drug-resistant gram + ve, gram − ve bacteria and clinical isolates of Mycobacterium phlei.

				Gram-positive strains (µg/mL)		Acid-fast strains (µg/mL)	Gram-negative strains (µg/mL)
							Enteric		Non-enteric
Compound	R1	R2	x	MSSA ATCC 29213	MRSA Clinical isolate	Mycobacterium phlei Clinical isolate	E. coli ATCC 25922	K. pneumoniae ATCC 700603	A. baumannii Clinical isolate	P. aeruginosa ATCC 27853
Va	H	H	NH	16	32	128	128	128	128	256
Vb	Cl	H	NH	4	4	16	8	8	16	16
Vc	Ph	H	NH	4	4	8	4	8	8	32
Vd	H	CH3	NH	16	8	64	32	64	32	128
Ve	Cl	CH3	NH	8	4	32	32	16	32	128
VIa	H	H	S	>1024	1024	1024	>1024	1024	1024	1024
VIb	Cl	H	S	256	128	128	512	256	128	1024
VIc	Ph	H	S	64	64	128	512	512	64	128
VId	H	CH3	S	1024	1024	512	1024	1024	256	1024
VIe	Cl	CH3	S	64	64	128	256	512	256	256
VIIa	H	H	O	>1024	512	128	>1024	>1024	512	1024
VIIb	Cl	H	O	256	256	256	512	512	256	128
VIIc	Ph	H	O	256	64	128	128	256	256	256
VIId	H	CH3	O	>1024	1024	512	>1024	1024	1024	1024
VIIe	Cl	CH3	O	64	64	128	1024	512	256	256
Levofloxacin				0.5	8	2	0.05	4	8	4
Rifampicin				0.015	4	0.0125	16	8	8	64

Anti-bacterial activity against gram + ve strains

In the evaluation of anti-bacterial activity against both methicillin-susceptible S. aureus (MSSA) and clinical isolates of MRSA (Table 2), the aminoguanidine series Va–e demonstrated excellent activity against both MSSA and MRSA with MIC values ranging from 4 μg/mL to 32 μg/mL. Compounds Vb, Vc, and Ve outperformed levofloxacin (MIC = 8 μg/mL) with a MIC of 4 μg/mL on MRSA, revealing promising activity as anti-MRSA agents. Vd showed notable activity against MRSA with a MIC of 8 μg/mL, equivalent to levofloxacin. Compound Va displayed the least activity in the series against MRSA with an MIC of 32 μg/mL. In the thiosemicarbazide series, compounds VIc and VIe displayed a MIC value of 64 μg/mL for both MSSA and MRSA, while compounds VIa, VIb, and VId showed higher MIC values (≥512 μg/mL). Similarly, in the semicarbazide series, compounds VIIc and VIIe exhibited a MIC value of 64 μg/mL for both MSSA and MRSA, while VIIa, VIIb, and VIId displayed little to no activity against both strains (≥256 μg/mL).

Anti-bacterial activity against gram − ve strains

All synthesised compounds were tested for their anti-bacterial activity against the enteric gram-negative bacterial strains E. coli and K. pneumoniae and the non-enteric gram-negative bacterial strains A. baumannii and P. aeruginosa (Table 2). In the aminoguanidine series Va–e, Vc exhibited notable activity with MIC values of 4 μg/mL against E. coli, 8 μg/mL against K. pneumoniae, and 8 μg/mL against A. baumannii, comparable to levofloxacin. Vb also displayed significant activity with a MIC of 8 μg/mL against both enteric strains and 16 μg/mL against non-enteric strains. The thiosemicarbazide series VIa–e showed less promising results, as VIb and VIe demonstrated higher MIC values of 256 μg/mL against K. pneumoniae and E. coli, respectively, while VIa, VIc, and VId exhibited MIC values ≥512 μg/mL. The semicarbazide series VIIa–e displayed little to no activity against enteric and non-enteric bacterial strains, with MIC values ranging from 128 μg/mL to 1024 μg/mL.

Anti-tubercular activity against Mycobacterium phlei

The anti-tubercular activity of synthesised compounds against clinical isolates of Mycobacterium phlei was assessed (Table 2). In the aminoguanidine series Va–e, Vc exhibited the most potent anti-tubercular effect with the lowest MIC value of 8 μg/mL, followed by Vb with a MIC of 16 μg/mL. Compounds Va, Vd, and Ve displayed MIC values ranging from 32 μg/mL to 128 μg/mL. For the thiosemicarbazide series, VIa–e, VI, VIc, and VIe demonstrated the lowest MIC value of 128 μg/mL, while VIa and VId exhibited higher MIC values of 1024 μg/mL and 512 μg/mL, respectively. The semicarbazide series VIIa–e showed MIC values ranging from 128 μg/mL to 512 μg/mL.

Anti-virulence activity

Bacterial biofilms are organised communities of bacteria that attach to surfaces and become encased within a self-produced extracellular polymeric substance (EPS) matrix. This matrix, composed of polysaccharides, proteins, nucleic acids, and lipids, provides structural integrity, protection, and an environment conducive to bacterial communication and nutrient exchange⁴⁸ The EPS matrix also acts as a protective barrier, shielding bacteria from environmental stressors such as antibiotics, desiccation, and immune responses⁴⁹. We evaluated the antibiofilm efficacy of the designed compounds at sub-MIC doses by measuring the biofilm density of treated cultures relative to untreated control. Notably, all the compounds exhibited significant, broad-spectrum anti-biofilm activity against biofilms formed by E. coli, S. aureus, and P. aeruginosa (A, Figure 8). Microscopic analysis of biofilms from E. coli, and S. aureus culture treated with the most potent compounds revealed distinct differences in biofilm structure between untreated and treated bacterial cultures. Untreated biofilms exhibited dense, cohesive structures, while drug-treated biofilms appeared disrupted, sparse, and unstable, reflecting effective antibiofilm activity (B, Figure 8).

Figure 8.

Evaluation of the anti-virulence effects of designed compounds. (A) Antibiofilm activity of designed compounds against E. coli, S. aureus, and P. aeruginosa. Biofilm density was quantified and expressed as a percentage relative to untreated controls. Statistical analysis was performed using one-way ANOVA followed by Dunnett’s multiple comparison test (***p < 0.001). (B) Microscopic visualisation of biofilms from E. coli and S. aureus. Scale bar 50 µm. (C) Anti-protease activity of designed compounds evaluated in S. aureus and P. aeruginosa using skimmed milk agar. Protease activity in untreated cultures produced clear inhibition zones, which were entirely absent in drug-treated cultures, confirming complete protease inhibition. U: untreated culture; S: 1% SDS; N: no bacterial lysate.

Proteases represent another critical bacterial virulence factor, facilitating infection by breaking down host tissues and evading immune defences⁵⁰. To evaluate the impact of the designed compounds on protease activity, cultures of P. aeruginosa and S. aureus were treated with the drugs and analysed using the skimmed milk agar method. Supernatants from untreated cultures served as controls, demonstrating full protease activity (C, Figure 8). Protease activity was evidenced by clear inhibition zones surrounding the wells, with larger zones corresponding to higher enzymatic activity. In cultures treated with the compounds, these inhibition zones were completely absent, indicating effective protease inhibition. Collectively, these findings highlight the remarkable ant virulence properties of the designed compounds.

In vitro cytotoxicity assay

Antimicrobial agents should demonstrate low or no toxicity towards mammalian cells^51,52. Ideally, a potential drug should have a relatively high cytotoxic concentration (CC50) and a relatively low active concentration⁵³. Evaluation of the selectivity index (SI) value is considered crucial for determining the therapeutic potential of evaluated compounds⁵⁴. Selectivity index is defined as the ratio of the toxic concentration known as 50% cell cytotoxic concentration (CC50) of a sample against its effective bioactive concentration (MIC)⁵⁵. High values of SI indicate excellent selectivity^55–57.

This work evaluated the cytotoxicity of the four most potent compounds (Vb, Vc, Vd, and Ve) by MTT assay against normal mammalian cell line VERO (African Green Monkey Kidney cells). The CC50 values were calculated for the four compounds, along with the SI of each compound towards MRSA (Table 3). In this work, SI was calculated following this equation:

$$\mathrm{\text{SI}}=\mathrm{\text{CC}}50/\mathrm{\text{MIC}}\mathrm{ }$$

where CC50 is 50% vero cell cytotoxic concentration for 48 h in µg/mL, and MIC is minimal inhibitory concentration against MRSA in µg/mL.

Table 3.

In vitro cytotoxicity screening against normal mammalian cell line VERO for the most potent compounds, their MIC on MRSA, and calculated mean selectivity index.

Cpd	Cytotoxicity	Active Concentration	Mean SI
	Vero CC50 (µg/mL)	MRSA MIC value (µg/mL)
Vb	43.32 ± 2.86	4	10.83
Vc	188.17 ± 6.87	4	47.01
Vd	2.8 ± 0.35	8	0.27
Ve	10.33 ± 0.49	4	2.55

The most potent compound, Vc, demonstrated excellent selectivity towards MRSA with a mean SI of 47.01 and CC50 value of 188.17 ± 6.87 µg/mL; moreover, cell viability was 100% at a concentration of 15.6 µg/mL, which is almost four times its MIC against MRSA (Table 3). Compound Vb also demonstrated good selectivity against MRSA with a mean SI value of 10.83 and CC50 value of 43.32 ± 2.86 µg/mL. Moreover, cell viability was 76.38% at a concentration of 15.6, almost four times its MIC against MRSA (Figure 1S). Compounds Vd and Ve displayed moderate to weak selectivity against MRSA with mean SI values of 0.27, 2.55, and CC50 values of 2.8 ± 0.35 and 10.33 ± 0.49 µg/mL, respectively.

In silico studies

Docking studies and target identification

Bacterial cytological profiling (BCP) previously indicated that the phenylthiazole lead compound exerted rapid bactericidal activity via inhibition of cell wall synthesis, resulting in cell lysis and morphological defects^19,58. Further validation using transposon mutagenesis pinpointed undecaprenyl pyrophosphate phosphatase (UPPP) as the primary target¹⁹.

UPPP plays a crucial role in bacterial isoprenoid biosynthesis, an early and essential step in peptidoglycan formation⁵⁹. Unlike traditional antibiotics such as β-lactams and glycopeptides that interfere with late-stage transpeptidation or crosslinking, UPPP inhibition disrupts the initial stages of cell wall biosynthesis – a mechanism particularly advantageous against resistant strains like MRSA and VRE^60–63

To explore whether our newly designed N-arylpyrrole derivatives act through a similar mechanism, we docked them into the UPPP active site (PDB: 6CB2) using the CDOCKER protocol in Discovery Studio 2024⁶⁴. The phenylthiazole lead showed a -CDOCKER energy of 22.72 and formed four key hydrogen bonds – three involving the aminoguanidine group and Glu49/Glu17, and one involving the thiazole sulphur and Thr28 (Figure 9).

Figure 9.

The 2D interactions of the lead compound and our designed compound Vc in the binding site of UPPP protein (PDB code: 6CB2).

Encouragingly, our designed arylpyrrole derivatives exhibited comparable docking poses and favourable binding energies, suggesting potential UPPP inhibition as a shared mechanism of action (Figure 9, Table 4S). Despite structural differences, the conserved binding interactions imply that these compounds may disrupt bacterial cell wall synthesis by targeting the same pathway.

We therefore hypothesise that the observed broad-spectrum antibacterial, antibiofilm, and antivirulence activities of the N-arylpyrrole series are at least partly mediated through inhibition of UPPP, leading to impaired peptidoglycan biosynthesis and cell wall integrity. This proposed mechanism aligns with the potent activity against MRSA and ESKAPE pathogens and justifies further studies to validate UPPP as a primary molecular target.

ADMET studies

Absorption, distribution, metabolism, excretion, and toxicity (ADMET) are essential parameters to be considered during the drug development process, as they provide a base for the newly designed compounds to be efficient drug candidates^65,66. Therefore, in silico ADMET studies were performed by applying the ADMET descriptors algorithm in Discovery Studio 2024 software package⁶⁷. In which pharmacokinetic parameters such as aqueous solubility^68,69, blood–brain barrier (BBB) penetration⁷⁰, cytochrome P4502D6 (CYP2D6) inhibition⁷¹, human intestinal absorption (HIA)⁷², plasma protein binding (PPB)⁷³ were calculated quantitatively for the four most potent compounds (Vb, Vc, Vd, and Ve) (Table 4).

Table 4.

In silico ADMET screening of the synthesised compounds.

Compound	Solubility level	BBB-level	Absorption level	ALog P98	PSA_2D	CYP2D6# prediction	PPB # prediction
Vb	3	3	0	2.69	79.109	True	True
Vc	2	2	0	3.54	79.109	False	True
Vd	3	3	0	2.01	79.109	False	True
Ve	3	3	0	2.67	79.109	False	True

Comparisons were made with reference level values obtained from Accelrys Discovery Studio (Table 4S)^74,75. Compounds Vb, Vd, and Ve, scoring at level 3, showed low BBB penetration, suggesting a low risk of central nervous system (CNS) side effects. In contrast, compound Vc, scoring at prediction level 2, indicated medium BBB penetration⁷⁰. None of the evaluated compounds, except for Vb, inhibited CYP2D6, thus minimising the likelihood of drug–drug interactions⁷¹. They were also predicted to have high PPB; however, modifications of these compounds could potentially enhance their drug-like properties.

Polar surface area (PSA) plays a significant role in HIA and membrane permeability⁷⁶. Both PSA and AlogP98 were calculated to assess drug bioavailability. Molecules with PSA > 150 or AlogP98 > 7.0 are typically associated with very low absorption and bioavailability⁷⁷. All evaluated compounds exhibited a PSA of 79.10 and AlogP98 values ranging from 2.01 to 3.54, indicating good oral absorption theoretically⁷⁸. The intestinal absorption model, depicted in Figure 10, incorporates 95% and 99% confidence ellipses in the ADMET_PSA_2D, ADMET_AlogP98 plane. These ellipses delineate regions where well-absorbed compounds are anticipated: 95% of well-absorbed compounds are expected within the 95% ellipse, while 99% should fall within the 99% ellipse. All evaluated compounds fell within the 99% ellipse, indicating good absorption.

Figure 10.

Plot of human intestinal absorption (HIA) and blood–brain barrier plot for the newly synthesised hits.

Structure–activity relationship study

A brief SAR study was extracted from the results obtained from the antimicrobial evaluation of the novel N-arylpyrrole derivatives. Judging by the MIC values of the fifteen compounds evaluated, it was found that the aminoguanidine series Va–e were the most active derivatives. In agreement with our preliminary evaluation, compound Vc, featuring a long and highly lipophilic tail (phenyl moiety), stood out as the most potent with promising activity on MRSA (4 μg/mL), E. coli (4 μg/mL), A. baumannii (8 μg/mL), K. pneumoniae (8 μg/mL), and P. aeruginosa (32 μg/mL). Vc also exhibited the lowest MIC value (8 μg/mL) of all tested compounds against clinical isolates of Mycobacterium phlei. Compound Vb, with the second most lipophilic tail (chloride moiety), also displayed promising activity, albeit with slightly higher MIC values. Adding a methyl group at R₂ in compound Ve resulted in one-fold higher MIC values, except for the MIC value against MRSA, which remained the same (4 μg/mL). Compound Va, belonging to the aminoguanidine series Va–e and containing no lipophilic tail (R₁═H), had the least activity against all tested strains; however, the addition of methyl group at R₂ in compound Vd increased the activity, especially against MRSA, where the MIC value of Vd was eightfolds lower than Va (4 μg/mL).

Thiosemicarbazide and semicarbazide series VIa–e and VIIa–e generally exhibited minimal antimicrobial activity. However, compounds VIc, VIIc, VIe, and VIIe, with more lipophilic tails, showed a MIC of 64 μg/mL against MRSA. Increasing lipophilicity at R₁ significantly enhanced antimicrobial activity and adding a methyl group at R₂ further boosted the activity. In conclusion, aminoguanidine substitution, increased lipophilicity at R₁, and adding a methyl group at R₂ are crucial factors for achieving broad-spectrum anti-bacterial activity in the investigated arylpyrrole scaffold. A graphical representation of the SAR is presented in Figure 11.

Figure 11.

Graphical representation of the SAR.

Conclusions

In this study, we synthesised and evaluated three series of N-arylpyrrole derivatives (Va–e, VIa–e, and VIIa–e) for their antimicrobial potential. Series Va–e, particularly compounds Vb, Vc, and Ve, exhibited significant anti-bacterial activity against MRSA and other ESKAPE pathogens. Notably, compound Vc demonstrated superior potency with a MIC of 4 μg/mL against MRSA, outperforming levofloxacin (MIC = 8 μg/mL), and showed activity against E. coli, K. pneumoniae, and A. baumannii. Additionally, Vc displayed notable anti-tubercular activity against M. phlei (MIC = 8 μg/mL). Beyond their direct antimicrobial effects, the compounds exhibited significant antivirulence properties, disrupting biofilm formation in E. coli, S. aureus, and P. aeruginosa at sub-MIC levels and effectively inhibiting bacterial protease activity in P. aeruginosa and S. aureus. Cytotoxicity assessments confirmed Vb and Vc as highly selective for bacterial cells, with Vc displaying an impressive SI of 47.01. In silico ADMET predictions supported favourable pharmacokinetic and toxicity profiles for the synthesised compounds, suggesting their potential as lead candidates for further development. Docking studies further predicted that the antibacterial mechanism involves inhibition of the UPPP enzyme, a key player in early-stage bacterial cell wall synthesis. This may explain their activity against resistant strains such as MRSA. These findings position N-arylpyrrole derivatives, particularly Vc, as compelling leads for the development of new antimicrobial agents. Further studies including structural refinement, resistance development profiling, and in vivo validation are warranted to comprehensively evaluate their therapeutic potential.

Experimental

Preliminary evaluation of the designed compounds using computer-aided molecular modelling

3D QSAR pharmacophore generation

The Discovery Studio V4.1 program performed all the molecular modelling studies. All the compounds were sketched in the Discovery studio program V4.1 and optimised using a CHARMM force field. For the 3D QSAR model generation, a library of 32 phenylthiazole and phenylpyrazole derivatives with reported activity (MIC) against MRSA was extracted from the literature^26–28 and randomly divided into training and test sets. The training set was subjected to the feature mapping protocol in DS to identify distinct chemical features present in the ligands. The features revealed were HBA, hydrogen bond donor (HBD), HYD, and RA. The four features identified were chosen for the 3D QSAR pharmacophore generation protocol. The MIC was selected as the active property, and the energy threshold was retained at 20 kcal/mol throughout the protocol run. The uncertainty value was set to 1.5. This value represents a ratio of the reported value to the minimum and maximum values. Setting the uncertainty value to 1.5 entails that the model can acclimate differences in the experimental MIC values and predict MIC up to 1.5 times³². All the other parameters were left to default.

The HypoGen algorithm utilised in the 3D QSAR pharmacophore generation protocol of DS interpreted the common chemical features related to low or high biological activity in the training set, and 10 hypotheses were generated. Cost analysis was used to choose the best hypothesis from the 10 generated models.

Chemistry

Chemicals were purchased from Sigma-Aldrich (Darmstadt, Germany), Merck (Darmstadt, Germany), Alfa Aesar (Karlsruhe, Germany), and Loba Chemie (Mumbai, India) and were used as such without further purification. Reactions were followed using analytical thin layer chromatography (TLC), performed on Aluminium silica gel 60 F254 TLC plates purchased from Merck (Darmstadt, Germany), with visualisation under UV light (254 nm). Melting points were recorded on Stuart Scientific apparatus and are reported herein uncorrected. Routine ¹H and ¹³C nuclear magnetic resonance spectra were recorded on Bruker 400 MHz spectrometer in δ scale (ppm), using DMSO as solvent and TMS as the internal standard signal at Center for Drug Discovery Research and Development, Faculty of Pharmacy, Ain Shams University, Cairo, Egypt. Infra-red (IR) spectra were recorded using Shimadzu FT-IR 8400s spectrophotometer (Kyoto, Japan) and KBr discs at the Center for Drug Discovery Research and Development, Faculty of Pharmacy, Ain Shams University, Cairo, Egypt. Elemental analyses were conducted on a Thermo Scientific Flash 2000 elemental analyser (Waltham, MA) at the Regional Center for Mycology and Biotechnology, Al-Azhar University, Cairo, Egypt. High-resolution mass spectrometry analyses were performed at the Faculty of Pharmacy, Al-Fayoum University, Faiyum, Egypt, using an Agilent Q-TOF LC/MS system (Santa Clara, CA) equipped with an electrospray ionisation (ESI) source and time-of-flight (TOF) detector.

General procedure for the synthesis of N-aryl-2,5-dimethylpyrrole derivatives (IIIa–c): A mixture of appropriate aniline (2 mmol) (IIa–c) and 2,5-hexanedione (2 mmol, 0.22 g, 0.22 mL) was stirred for 1–6 h at room temperature in the presence of sulphamic acid (0.1 mmol, 0.1 g) as a catalyst. TLC monitored the reaction, and the solvent system used for TLC monitoring was ethyl acetate:hexane (3:7). Then, the reaction mixture was poured on 25 mL of ice water, and the formed precipitate was filtered and crystallised from ethanol, affording pure substituted pyrroles IIIa–c, which were used for the next step without further purification.

• 2,5-Dimethyl-1-phenyl-1H-pyrrole (IIIa). Buff crystals (0.33 g, 98%); MP: 51–52 °C (as reported)^40,41.

• 1-(4-Chlorophenyl)-2,5-dimethyl-1H-pyrrole (IIIb). White crystals (0.4 g, 80%); MP: 47–49 °C (as reported)⁴³.

• 1-([1,1′-Biphenyl]-4-yl)-2,5-dimethyl-1H-pyrrole (IIIc). Dark red crystals (0.45 g, 92.6%); MP: 67–70 °C (as reported).⁴²

General procedure for the synthesis of N-aryl-2,5-dimethyl-1H-pyrrole-3-carbaldehydes (IVa–c): Vilsmeier–Haack reagent was prepared by adding phosphorus oxychloride (POCl₃) (3 mmol, 0.46 g, 0.28 mL) dropwise to ice-cooled DMF (10 mL). A solution of the appropriate pyrrole derivative (IIIa–c) (1 mmol) in DMF (5 mL) was added slowly to the Vilsmeier–Haack reagent. The reaction mixture was allowed to warm to room temperature, then heated to 60 °C for 4 h. The reaction progress was monitored by TLC, which showed a new spot and the disappearance of the starting material. After that, the reaction mixture was cooled to room temperature, poured into crushed ice, and quenched with 20% NaOH, forming the precipitate. The resultant precipitate was filtered and washed with water to get 2,5-dimethyl-1-aryl-1H-pyrrole-3-carbaldehydes IVa–c.

• 2,5-Dimethyl-1-phenyl-1H-pyrrole-3-carbaldehyde (IVa). Grey solid (0.1 g, 86%); MP: 90 °C (as reported)⁷⁹. IR (KBr, cm⁻¹): 1730 (C═O). ¹H NMR (400 MHz, DMSO-d₆) δ (ppm): 1.93 (s, 3H, CH₃), 2.23 (s, 3H, CH₃), 6.3 (s, 1H, pyrrolyl), 7.35 (d, J = 8 Hz, 2H, ArH), 7.52–7.6 (m, 3H, ArH), 9.8 (s, 1H, CH═O), The ¹H NMR is in accordance with literature⁴⁵.

• 1-(4-Chlorophenyl)-2,5-dimethyl-1H-pyrrole-3-carbaldehyde (IVb). Brown solid (0.12 g, 60%); MP: 120 °C. IR (KBr, cm⁻¹): 1730 (C═O). ¹H NMR (400 MHz, DMSO-d₆) δ (ppm): 1.93 (s, 3H, CH₃), 2.23 (s, 3H, CH₃), 6.3 (s, 1H, pyrrolyl), 7.42 (d, J = 8 Hz, 2H, ArH), 7.64 (d, J = 8 Hz, 2H, ArH), 9.8 (s, 1H, CH═O). The ¹H NMR is in accordance with literature⁴⁴.

• 1-([1,1′-Biphenyl]-4-yl)-2,5-dimethyl-1H-pyrrole-3-carbaldehyde (IVc). Grey solid (0.22 g, 80%); MP: 196–200 °C. IR (KBr, cm⁻¹): 1730 (C═O). ¹H NMR (400 MHz, DMSO-d₆) δ (ppm): 1.99 (s, 3H, CH₃), 2.29 (s, 3H, CH₃), 6.3 (s, 1H, pyrrolyl), 7.42–7.51 (m, 5H, ArH), 7.76 (d, J = 8 Hz, 2H, ArH), 7.86 (d, J = 8 Hz, 2H, ArH), 9.83 (s, 1H, CH═O). Anal. Calcd for C₁₉H₁₇NO (275.34): C, 82.88; H, 6.22; N, 5.09; found: C, 82.63; H, 6.41; N, 5.19%.

General procedure for the synthesis of N-aryl-2,5-dimethyl-1H-pyrrol-3-yl-ethanones (IVd–e): Phosphorus oxychloride (3 mmol, 0.46 g, 0.28 mL) was added dropwise to DMA (10 mL) at 0 °C. The mixture was stirred for 30 min at room temperature. Then, one mmol of IIIa–c in DMA (20 mL) was added to the mixture at 0 °C. Subsequently, the reaction mixture was heated at reflux for 6 h, and then after cooling, ice water was added, and the mixture was neutralised with 20% NaOH. The precipitate was filtered, washed with water, and dried, affording the desired compounds IVd–e.

• 1-(2,5-Dimethyl-1-phenyl-1H-pyrrol-3-yl)ethanone (IVd). Greyish solid (0.16 g, 80%); MP: 80–81 °C (as reported)³⁷. IR (KBr, cm⁻¹): 1727 (C═O). ¹H NMR (400 MHz, DMSO-d₆) δ (ppm): 1.93 (s, 3H, CH₃), 2.20 (s, 3H, CH3–C═O), 2.23 (s, 3H, CH₃), 6.39 (s, 1H, pyrrolyl), 7.3 (d, J = 8 Hz, 2H, ArH), 7.52–7.59 (m, 3H, ArH). The ¹H NMR is in accordance with literature^39,46.

• 1-[1-(4-Chlorophenyl)-2,5-dimethyl-1H-pyrrol-3-yl]ethanone (IVe). Yellow solid (0.14 g, 60%); MP: 81 °C, (as reported)^38,39.

General procedure for the synthesis of hydrazinecarboximidamide derivatives (Va–e): To a solution of the intermediates IVa–e (1 mmol) in absolute ethanol (25 mL), 1 mL of concentrated HCl was added, followed by dry aminoguanidine HCO₃ (4 mmol, 0.54). This mixture was heated under reflux for 20–24 h, and the progress of the reaction was monitored by TLC (8:2, DCM:methanol). The reaction mixture was rinsed with sodium bicarbonate solution, and the solid formed was filtered, giving the pure compounds Va–e.

• 2-((2,5-Dimethyl-1-phenyl-1H-pyrrol-3yl)methylene)hydrazinecarboximidamide (Va). Yellow solid (0.22 g, 90%); MP = 195–197 °C. IR (KBr, cm⁻¹): 3473, 3403, 3303 (NH/NH₂), 1656, 1582 (C═N). ¹H NMR (400 MHz, DMSO-d₆) δ (ppm): 1.95 (s, 3H, CH₃), 2.04 (s, 3H, CH₃), 5.51 (s, 2H, NH₂), 5.8 (s, 2H, NH), 6.24 (s, 1H, pyrrolyl), 7.27 (d, J = 8 Hz, 2H, ArH), 7.46 (t, J = 8 Hz, 1H, ArH), 7.54 (t, J = 8 Hz, 2H, ArH), 7.99 (s, 1H, CH═N). ¹³C NMR (100 MHz, DMSO-d₆) δ (ppm): 13.1, 13.9 (CH₃s), 107, 120.9, 126.1, 126.9, 128.3, 128.6, 129.7, 138.4 (Ar-Cs), 148.6 (C═NH), 158.4 (C═N). HRMS (ESI) m/z: calcd for C₁₄H₁₇N₅ [M + H]⁺ 256.1562, found 256.1562. Anal. Calcd for C₁₄H₁₇N₅ (255.15): C, 65.86; H, 6.71; N, 27.43; found: C, 66.04; H, 6.83; N, 27.25%.

• 2-((1-(4-Chlorophenyl)-2,5-dimethyl-1H-pyrrol-3yl)methylene)hydrazinecarboximidamide (Vb). Yellow solid (0.28 g, 98%); MP = 218–220 °C. IR (KBr, cm⁻¹): 3445, 3335 (NH/NH₂), 1670, 1621 (C═N). ¹H NMR (400 MHz, DMSO-d₆) δ (ppm): 1.96 (s, 3H, CH₃), 2.05 (s, 3H, CH₃), 5.23 (s, 2H, NH₂), 5.64 (s, 2H, NH), 6.24 (s, 1H, pyrrolyl), 7.33 (d, J = 8 Hz, 2H, ArH), 7.5 (d, J = 8 Hz, 2H, ArH), 7.98 (s, 1H, CH═N). ¹³C NMR (100 MHz, DMSO-d₆) δ (ppm): 11.2, 13 (CH₃s), 105.5, 118, 128.7, 128.9, 129.8, 130.3, 133, 137 (Ar-Cs), 140.8 (C═NH), 158.7 (C═N). HRMS (ESI) m/z: calcd for C₁₄H₁₆ClN₅ [M + H]⁺ 290.1176, found 290.1177. Anal. Calcd for C₁₄H₁₆ClN₅ (289.11): C, 58.03; H, 5.57; N, 24.17; found: C, 58.2; H, 5.73; N, 22.04%.

• 2-((1-([1,1′-Biphenyl]-4-yl)-2,5-dimethyl-1H-pyrrol-3-yl)methylene)hydrazinecarboximidamide (Vc). Dark brown solid (0.26 g, 80%); MP = 201–204 °C. IR (KBr, cm⁻¹): 3469, 3332 (NH/NH₂), 1684, 1610 (C═N). ¹H NMR (400 MHz, DMSO-d₆) δ (ppm): 2 (s, 3H, CH₃), 2.13 (s, 3H, CH₃), 5.21 (s, 2H, NH₂, exchangeable by D₂O), 5.9 (s, 2H, NH, exchangeable by D₂O), 6.3 (s, 1H, pyrrolyl), 7.3–7.51 (m, 5H, ArH), 7.73 (d, J = 8 Hz, 2H, ArH), 7.8 (d, J = 8 Hz, 2H, ArH), 8.3 (s, 1H, CH═N). ¹³C NMR (100 MHz, DMSO-d₆) δ (ppm): 11.4, 13.1 (CH₃s), 105.4, 114.7, 125.8, 127.2, 127.6, 127.9, 128.9, 129.1, 129.5, 137.3, 139.5, 140 (Ar-Cs), 141.1 (C═NH), 158.5 (C═N). HRMS (ESI) m/z: calcd for C₂₀H₂₁N₅ [M + H]⁺ 332.1878, found 332.1882. Anal. Calcd for C₂₀H₂₁N₅ (331.18): C, 72.48; H, 6.39; N, 21.13; found: C, 72.52; H, 6.48; N, 21.40%.

• 2-(1-(2,5-Dimethyl-1-phenyl-1H-pyrrol-3-yl)ethylidene)hydrazinecarboximidamide (Vd). Yellow solid (0.13 g, 50%); MP = 178–180 °C. IR (KBr, cm⁻¹): 3474, 3403, 3303 (NH/NH₂), 1652, 1606 (C═N). ¹H NMR (400 MHz, DMSO-d₆) δ (ppm): 1.94 (s, 3H, CH₃), 2.15 (s, 3H, CH₃), 2.16 (s, 3H, CH₃–C═N), 5.21 (s, 2H, NH₂), 5.50 (s, 2H, NH), 6.08 (s, 1H, pyrrolyl), 7.25 (d, J = 8 Hz, 2H, ArH), 7.458 (t, J = 8 Hz, 1H, ArH), 7.53 (t, J = 8 Hz, 2H, ArH). ¹³C NMR (100 MHz, DMSO-d₆) δ (ppm): 13.1, 13.9, 16.4 (CH₃s), 107, 120.9, 126.1, 126.9, 128.3, 128.6, 129.7, 138.4 (Ar-Cs), 148.6 (C═NH), 158.4 (C═N). HRMS (ESI) m/z: calcd for C₁₅H₁₉N₅ [M + H]⁺ 270.1717, found 270.1720. Anal. Calcd for C₁₅H₁₉N₅ (269.16): C, 66.89; H, 7.11; N, 26.00; found: C, 67.05; H, 7.27; N, 26.21%.

• 2-(1-(1-(4-Chlorophenyl)-2,5-dimethyl-1H-pyrrol-3-yl)ethylidene)hydrazinecarboximidamide (Ve). Brown solid (0.23 g, 77%); MP = 175–178 °C. IR (KBr, cm⁻¹): 3466, 3321 (NH/NH₂), 1616, 1595 (C═N). ¹H NMR (400 MHz, DMSO-d₆) δ (ppm): 1.96 (s, 3H, CH₃), 2.15 (s, 3H, CH₃), 2.21 (s, 3H, CH₃–C═N), 5.8 (s, 2H, NH₂), 6.01 (s, 2H, NH), 6.11 (s, 1H, pyrrolyl), 7.29 (d, J = 8 Hz, 2H, ArH), 7.6 (d, J = 8 Hz, 2H, ArH). ¹³C NMR (100 MHz, DMSO-d₆) δ (ppm): 13, 13.9, 16.4 (CH₃s), 107.3, 121, 126.3, 127, 129.8, 130.5, 132.9, 137.3 (Ar-Cs), 148.5 (C═NH), 158.3 (C═N). HRMS (ESI) m/z: calcd for C₁₅H₁₈ClN₅ [M + H]⁺ 304.1333, found 304.1333. Anal. Calcd C₁₅H₁₈ClN₅ (303.13): C, 59.30; H, 5.97; N, 23.05; found: C, 59.47; H, 6.11; N, 23.32%.

General procedure for the synthesis of hydrazinecarbothioamide derivatives (VIa–e): A solution of the intermediates IVa–e (1 mmol) in absolute ethanol (20 mL) was added slowly dropwise to a solution of thiosemicarbazide hydrochloride (2 mmol, 0.25 g) and sodium acetate (4 mmol, 0.32 g) dissolved in 15–20 mL of distilled water. The reaction mixture was heated under reflux for 12 h, and the progress of the reaction was monitored by TLC (6:3:1, hexane:ethyl acetate:ethanol). The reaction mixture was rinsed with ice-cold water, and the solid formed was filtered and crystallised from ethanol, giving the pure compounds VIa–e.

• 2-((2,5-Dimethyl-1-phenyl-1H-pyrrol-3yl)methylene)hydrazinecarbothioamide (VIa). Yellow solid (0.25 g, 95%); MP = 243–245 °C. ¹H NMR (400 MHz, DMSO-d₆) δ (ppm): 1.95 (s, 3H, CH₃), 2.23 (s, 3H, CH₃), 6.10 (s, 2H, NH₂), 6.28 (s, 1H, pyrrolyl), 7.27 (d, J = 8 Hz, 2H, ArH), 7.47 (t, J = 8 Hz, 1H, ArH), 7.54 (t, J = 8 Hz, 2H, ArH), 7.84 (s, 1H, CH═N), 9.72 (s, 1H, –NH). The ¹H NMR is in accordance with literature⁸⁰. Anal. Calcd C₁₄H₁₆N₄S (272.11): C, 61.74; H, 5.92; N, 20.57; found: C, 62.02; H, 6.04; N, 20.80%.

• 2-((1-(4-Chlorophenyl)-2,5-dimethyl-1H-pyrrol-3-yl)methylene)hydrazinecarbothioamide (VIb)⁸¹. Yellow solid (0.28 g, 96%); MP = 255 °C. ¹H NMR (400 MHz, DMSO-d₆) δ (ppm): 1.95 (s, 3H, CH₃), 2.2 (s, 3H, CH₃), 6.17 (s, 2H, NH₂), 6.29 (s, 1H, pyrrolyl), 7.33 (d, J = 8 Hz, 2H, ArH), 7.33 (d, J = 8 Hz, 2H, ArH), 7.57 (d, J = 8 Hz, 2H, ArH),7.82 (s, 1H, CH═N), 9.72 (s, 1H, –NH). Anal. Calcd C₁₄H₁₅ClN₄S (306.81): C, 54.81; H, 4.93; N, 18.26; found: C, 55.06; H, 5.12; N, 18.45%.

• 2-((1-([1,1′-Biphenyl]-4-yl)-2,5-dimethyl-1H-pyrrol-3-yl)methylene)hydrazinecarbothioamide (VIc). Brown solid (0.23 g, 70%); MP = 224–226 °C. IR (KBr, cm⁻¹): 3477, 3403 (NH/NH₂), 1681 (C═N), 1482 (C═S). ¹H NMR (400 MHz, DMSO-d₆) δ (ppm): 2 (s, 3H, CH₃), 2.2 (s, 3H, CH₃), 6.17 (s, 2H, NH₂), 6.34 (s, 1H, pyrrolyl), 7.38–7.53 (m, 5H, ArH), 7.73 (d, J = 8 Hz, 2H, ArH), 7.80 (d, J = 8 Hz, 2H, ArH), 8.57 (s, 1H, CH═N), 9.73 (s, 1H, –NH). ¹³C NMR (100 MHz, DMSO-d₆) δ (ppm): 11.2, 13.1 (CH₃s), 104.86, 116.3, 127.2, 128, 128.3, 128.9, 129.5, 130, 133.3, 136.7, 139.5, 140.46 (Ar-Cs), 157.4 (C═NH), 199.2 (C═S). Anal. Calcd C₂₀H₂₀N₄S (348.14): C, 68.93; H, 5.79; N, 16.08; found: C, 69.15; H, 5.86; N, 16.27%

• 2-(1-(2,5-Dimethyl-1-phenyl-1H-pyrrol-3-yl)ethylidene)hydrazinecarbothioamide (VId). Buff solid (0.4 g, 84%); MP = 230–233 °C. IR (KBr, cm⁻¹): 3466, 3342 (NH/NH₂), 1684 (C═N), 1521 (C═S). ¹H NMR (400 MHz, DMSO-d₆) δ (ppm): 1.94 (s, 3H, CH₃), 2.13 (s, 3H, CH₃), 2.3 (s, 3H, CH₃–C═N), 6.14 (s, 2H, NH₂), 6.38 (s, 1H, pyrrolyl), 7.24 (d, J = 8 Hz, 2H, ArH), 7.47 (t, J = 8 Hz, 1H, ArH), 7.54 (t, J = 8 Hz, 2H, ArH), 9.05 (s, 1H, NH).¹³C NMR (100 MHz, DMSO-d₆) δ (ppm): 13, 13.6, 16.1 (CH₃s), 107, 119.2, 127.2, 128.6, 129.8, 130, 138.1, 144.9 (Ar-Cs), 158.1 (C═NH), 194.2 (C═N). HRMS (ESI) m/z: calcd for C₁₅H₁₈N₄S [M + H]⁺ 287.1334, found 287.1499. Anal. Calcd C₁₅H₁₈N₄S (286.13): C, 62.91; H, 6.33; N, 19.56; found: C, 63.15; H, 6.45; N, 19.75%.

• 2-(1-(1-(4-Chlorophenyl)-2,5-dimethyl-1H-pyrrol-3-yl)ethylidene)hydrazinecarbothioamide (VIe). Yellow solid (0.3 g, 78%); MP = 200–203 °C. IR (KBr, cm⁻¹): 3463, 3382 (NH/NH₂), 1695 (C═N), 1521 (C═S). ¹H NMR (400 MHz, DMSO-d₆) δ (ppm): 1.95 (s, 3H, CH₃), 2.14 (s, 3H, CH₃), 2.30 (s, 3H, CH₃–C═N), 6.15 (s, 2H, NH₂), 6.40 (s, 1H, pyrrolyl), 7.32 (d, J = 8 Hz, 2H, ArH), 7.58 (d, J = 8 Hz, 2H, ArH), 8.98 (s, 1H, NH). ¹³C NMR (100 MHz, DMSO-d₆) δ (ppm): 12.9, 13.5, 16.1 (CH₃s), 108.8, 119.4, 127.3, 129.8, 130.4, 133.1, 137, 144.8 (Ar-Cs), 158 (C═NH), 194.3 (C═S). Anal. Calcd C₁₅H₁₈ClN₄S (321.86): C, 55.98; H, 5.63; N, 17.41; found: C, 56.09; H, 5.41; N, 17.63%.

General procedure for the synthesis of hydrazinecarboxamide derivatives (VIIa–e): A solution of the key intermediates IVa–e (1 mmol) in absolute ethanol (20 mL) was added slowly dropwise to a solution of semicarbazide hydrochloride (0.2 mmol, 0.25 g) and sodium acetate (0.2 mmol, 0.32 g) dissolved in 15–20 mL of distilled water. The reaction mixture was heated under reflux for 12 h, and the progress of the reaction was monitored by TLC (6:3:1, hexane:ethyl acetate:ethanol). The reaction mixture was rinsed with ice-cold water, and the solid formed was filtered and crystallised from hot ethanol, giving the pure compounds VIIa–e.

• 2-((2,5-dimethyl-1-phenyl-1H-pyrrol-3-yl)methylene)hydrazinecarboxamide (VIIa)⁸². Yellow solid (0.23 g, 95%); MP = 250–255 °C (reported = 294 °C⁸²) ¹H NMR (400 MHz, DMSO-d₆) δ (ppm): 1.94 (s, 3H, CH₃), 2.01 (s, 3H, CH₃), 6.17 (s, 2H, NH₂), 6.28 (s, 1H, pyrrolyl), 7.27 (d, J = 8 Hz, 2H, ArH), 7.47 (t, J = 8 Hz, 1H, ArH), 7.54 (t, J = 8 Hz, 2H, ArH), 7.84 (s, 1H, CH═N), 9.72 (s, 1H, –NH). Anal. Calcd C₁₄H₁₆N₄O (256.30): C, 65.61; H, 6.29; N, 21.86; found: C, 65.83; H, 6.45; N, 21.98%.

• 2-((1-(4-Chlorophenyl)-2,5-dimethyl-1H-pyrrol-3-yl)methylene)hydrazinecarboxamide (VIIb). Yellow solid (0.26 g, 90%); MP = 248–251 °C. IR (KBr, cm⁻¹): 3466 (NH/NH₂), 1645 (C═N), 1681 (C═O). ¹H NMR (400 MHz, DMSO-d₆) δ (ppm): 1.95 (s, 3H, CH₃), 2.02 (s, 3H, CH₃), 6.17 (s, 2H, NH₂), 6.29 (s, 1H, pyrrolyl), 7.33 (d, J = 8 Hz, 2H, ArH), 7.58 (d, J = 8 Hz, 2H, ArH), 7.82 (s, 1H, CH═N), 9.72 (s, 1H, –NH). ¹³C NMR (100 MHz, DMSO-d₆) δ (ppm): 11.1, 13 (CH3s), 105, 116.5, 129.4, 129.9, 130.3, 133.2, 136.4, 136.7 (Ar-Cs), 146.4 (C═N), 157.3 (C═O). HRMS (ESI) m/z: calcd for C₁₄H₁₇ClN₄O [M + H]⁺ 291.1015, found 291.1016. Anal. Calcd C₁₄H₁₇ClN₄O (290.75): C, 59.11; H, 5.62; N, 18.38; found: C, 59.34; H, 3.78; N, 18.60%.

• 2-((1-([1,1′-Biphenyl]-4-yl)-2,5-dimethyl-1H-pyrrol-3-yl)methylene)hydrazinecarboxamide (VIIc). Brown solid (0.27 g, 84%); MP = 230–233 °C. IR (KBr, cm⁻¹): 3466 (NH/NH₂), 1613 (C═N), 1684 (C═O). ¹H NMR (400 MHz, DMSO-d₆) δ (ppm): 2.03 (s, 3H, CH₃), 2.2 (s, 3H, CH₃), 6.17 (s, 2H, NH₂), 6.32 (s, 1H, pyrrolyl), 7.34–7.53 (m, 5H, ArH), 7.73 (d, J = 8 Hz, 2H, ArH), 7.8 (d, J = 8 Hz, 2H, ArH), 8.57 (s, 1H, CH═N), 9.72 (s, 1H, –NH). ¹³C NMR (100 MHz, DMSO-d₆) δ (ppm): 11.2, 13.1(CH₃s), 104.8, 116.3, 125.8, 127.2, 128, 128.3, 128.9, 129.5, 130.1, 134, 136.8, 139.5 (Ar-Cs), 140.2 (C═N), 157.4 (C═O). HRMS (ESI) m/z: calcd for C₂₀H₂₀N₄O [M + H]⁺ 333.1717, found 333.1718. Anal. Calcd C₂₀H₂₀N₄O (332.16): C, 72.04; H, 6.06; N, 16.86; found: C, 72.04; H, 6.31; N, 16.9%.

• 2-(1-(2,5-Dimethyl-1-phenyl-1H-pyrrol-3-yl)ethylidene)hydrazine carboxamide (VIId). Buff solid (0.21 g, 78%); MP = 238–240 °C. IR (KBr, cm⁻¹): 3463 (NH/NH₂), 1585 (C═N), 1687 (C═O). ¹H NMR (400 MHz, DMSO-d₆) δ (ppm): 1.93 (s, 3H, CH₃), 2.13 (s, 3H, CH₃), 2.32 (s, 3H,CH₃–C═N), 6.14 (s, 2H, NH₂), 6.39 (s, 1H, pyrrolyl), 7.24 (d, J = 8 Hz, 2H, ArH), 7.44 (t, J = 8 Hz, 1H, ArH), 7.51 (t, J = 8 Hz, 2H, ArH), 9.04(s, 1H, NH). ¹³C NMR (100 MHz, DMSO-d₆) δ (ppm): 12.9, 13.6, 16.1 (CH₃s), 107, 119.2, 127.1, 127.4, 128.6, 129.8, 130, 138.1 (Ar-Cs), 144.9 (C═NH), 158.1 (C═O). HRMS (ESI) m/z: calcd for C₁₅H₁₈N₄O [M + H]⁺ 271.1559, found 271.1560. Anal. Calcd C₁₅H₁₈N₄O (270.15): C, 66.64; H, 6.71; N, 20.73; found: C, 66.81; H, 6.85; N, 20.97%.

• 2-(1-(1-(4-Chlorophenyl)-2,5-dimethyl-1H-pyrrol-3-yl)ethylidene)hydrazinecarboxamide (VIIe). Yellow solid (0.19 g, 65%); MP = 244–247 °C. IR (KBr, cm⁻¹): 3466 (NH/NH₂), 1578 (C═N), 1687 (C═O). ¹H NMR (400 MHz, DMSO-d₆) δ (ppm): 1.95 (s, 3H, CH₃), 2.14 (s, 3H, CH₃), 2.5 (s, 3H, CH₃–C═N), 6.15 (s, 2H, NH₂), 6.40 (s, 1H, pyrrolyl), 7.32 (d, J = 8 Hz, 2H, ArH), 7.58 (d, J = 8 Hz, 2H, ArH), 8.98 (s, 1H, NH). ¹³C NMR (100 MHz, DMSO-d₆) δ (ppm): 12.8, 13.5, 16.1 (CH₃s), 108.7, 120.7, 127.5, 128.5, 130, 130.4, 133.8, 134.7 (Ar-Cs), 136 (C═NH), 158 (C═S). HRMS (ESI) m/z: calcd for C₁₅H₁₇ClN₄O [M + H]⁺ 305.1171, found 305.1173. Anal. Calcd C₁₅H₁₇ClN₄O (304.77): C, 59.11; H, 5.62; N, 18.38; found: C, 59.35; H, 5.75; N, 18.62%.

Biological evaluation

Antimicrobial activity

The minimum inhibitory concentrations (MICs) of the synthesised final compounds and control antibiotics were determined using the broth microdilution method according to the guidelines outlined by the Clinical and Laboratory Standards Institute (CLSI)⁸³ or as described in previous reports^84,85.

In vitro cytotoxic assay

Compounds (Vb, Vc, Vd, and Ve) were assayed against normal mammalian cell line VERO (African Green Monkey Kidney cells) to determine the potential toxic effect on mammalian cells in vitro, as described in previous reports^86,87.

Antibiofilm assay

The biofilm formation was quantified as previously described^88–90. Briefly, bacterial overnight cultures in tryptone soya broth (TSB) were prepared and diluted with TSB to an optical density OD600 nm of 0.2 in 96 wells microtiter plate (150 µL per well). 1/4th MIC dose (dissolved in DMSO) of each drug was added to each well of the tested bacterial species. One well representing negative control (bacteria free) and another well representing positive control (with vehicle) were included. The plates were incubated for 24 h at 37 °C. The planktonic cells were aspirated, and the plates were washed softly with deionised water and then left to dry. The adherent bacterial cells were fixed for 25 min with methanol, followed by staining with crystal violet (1% w/v) for another 20 min. The excessive dye was washed, dried, then stained biofilms were imaged using a fully motorised Leica AF7000 microscope with a 20× objective lens. Finally, the adhered dye was dissolved in glacial acetic acid (33% v/v), and the absorbance was measured at 590 nm using a Biotek Spectrofluorometer (Biotek, Winooski, VT). The test was made in triplicate and the absorbance was shown as mean ± standard error of percentage changes from untreated controls. The percentage inhibition and reduction of biofilm growth was calculated by using the following equation:

$$\text{\% Inhibition}=(\mathrm{\text{OD}}590\_\mathrm{\text{Nc}}-\mathrm{\text{OD}}590\_\mathrm{E})/\mathrm{\text{OD}}590\_\mathrm{\text{Nc}}\times 100$$

In the above equation, Nc stands for negative control, and E for experimental samples.

Antiprotease activity

To assess the inhibitory effect of the tested drugs on bacterial protease activity, the skim milk agar method was employed according to previous descriptions^88–90. Suspensions from overnight cultures in 96-well plate of the indicated bacteria in TSB, along with each drug (1/4th of the MIC) were centrifuged at 10 000 × g for 20 min to pellet the planktonic cells. Subsequently, aliquots (100 µL) of supernatants were added to wells made in 5% skim milk agar plates, followed by incubation at 37 °C overnight. Supernatants from untreated cultures, and 1% SDS were used as positive control for protease activity (clear zone around the well). Negative control of protease activity was bacteria free media.

In silico studies

Docking studies

All the molecular docking studies done were based on the crystal structure of undecaprenyl diphosphate phosphatase (UPPP) (PDB code: 6CB2)⁶⁴. Molecular docking was done via the CDOCKER protocol DS. The CDOCKER protocol allows us to simulate the docking of a ligand into the target’s binding site and utilises several scoring functions to assess the docked poses⁹¹. It has been demonstrated that the CHARMM-based CDOCKER protocol yields highly accurate docked poses⁹². The protein preparation tool was used to correct common problems in the protein structure by adding missing loops and hydrogens and excluding alternate conformers. All water molecules that did not contribute to the binding interactions were removed. The binding site was then identified using the define and edit binding site tool, resulting in a sphere of 14 Å, and the binding site atomic co-ordinates were X: 0.499, Y: 4.206, and Z: 6.943.

The reference lead compound (E)-2-(1-(2-(4-butylphenyl)-4-methylthiazol-5-yl)ethylidene)hydrazinecarboximidamide and our N-arylpyrrole designed derivatives were sketched using DS and then prepared by the ligand preparation tool to fix any incorrect valences and generate 3D conformers. Both the reference ligand and our compounds were docked into the UPPP binding site, and the reference ligand interactions with UPPP were used to validate and justify the interactions of our newly designed compounds.

ADMET studies

In silico studies for the prediction of physicochemical/pharmacokinetic features and toxicity of the most potent compounds (Vb, Vc, Vd, and Ve) have been performed using the ADMET descriptors algorithm in Discovery Studio 2024 software package^29,67 to investigate their ADMET profiles.

Funding Statement

The researchers did not obtain any funding for the work they submitted. The publishing fee is covered under the CSIC agreement with Taylor & Francis. GY was supported by AYUDAS JUAN DE LA CIERVA-INCORPORACIÓN (IJC2020-044166-I). This work is also supported by the Ministerio de Ciencia, Innovación y Universidades Spain.

Disclosure statement

The authors report no conflicts of interest.