N-Trifluoromethylthiolated Sulfonimidamides and Sulfoximines: Anti-microbial, Anti-mycobacterial, and Cytotoxic Activity

Abstract

Herein we demonstrate the expanded utility of a recently described N-trifluoromethylthiolation protocol to sulfonimidamide containing substances. The novel N-trifluoromethylthio sulfonimidamide derivatives thus obtained were evaluated for antibacterial activity against Mycobacterium tuberculosis (M. tb.) and Mycobacterium abscessus and Gram + Ve (Streptococcus aureus, Bacillus subtilis), and Gram – Ve (Escherichia coli, Pseudomonas aeruginosa) bacteria. Two compounds, 13 and 15 showed high antimycobacterial activity with MIC value of 4–8 μg/mL; i.e. comparable to WHO recommended first line antibiotic for TB infection ethambutol. The same compounds were also found to be cytotoxic in HepG2 cells (compound 13 IC₅₀ = 15 μg/mL; compound 15 IC₅₀ = 65 μg/mL). A structure activity relationship, using matched pair analysis, gave the unexpected conclusion that the trifluoromethylthio moiety was responsible for the cellular and bacterial toxicity. Given the increasing use of the trifluoromethylthio group in contemporary medicinal chemistry, this observation calls for considerations before implementation of the functionality in drug design.

Drug resistant microorganisms are becoming a major threat to human health and civilization. If we fail to overcome this menace, we will re-enter an era where minor infections again become lethal.^1,2 Furthermore, the limited arsenal of new antibiotics and the declining global efforts in identification of new chemical entities (NCEs) with antibiotic activity enhance the alarming spread of strains that are resistant to known drugs.^3,4Mycobacterium tuberculosis (M. tb.) is the causative agent of tuberculosis (TB), one of the world’s leading airborne contagious diseases. According to the WHO report, it is estimated that TB remained as one of the top infectious diseases in 2017 with 10 million new infections and 1.6 million deaths! Drug resistant strains of M. tb. threaten the health of millions of people. Hence, urgent efforts are required to identify NCEs with novel mechanisms of action and significant therapeutic activities against both Gram-positive and Gram-negative bacteria as well as M. tb..

Sulfur containing molecules have historically played a vital role in pharmaceutical and agrochemical applications.⁵ Among these, the sulfonamide scaffold has been a central functional group in molecules with a wide spectrum of biological activities over a long time.⁶ Analogs of the sulfonamide functional group, e.g. their mono aza-analogue sulfonimidamides (SIAs), have recently gained attention as a sulfonamide bioisostere.⁷

Analogously, contemporary medicinal-chemistry and agrochemistry make extensive use of fluorine and fluorinated functional groups to optimize the pharmacological activity, not least the metabolic stability, of small molecule drugs.⁸⁻¹² The trifluoromethylthio (-SCF₃) functional group has rapidly emerged as a potential substitute for the more common trifluoromethyl (-CF₃) group. Some representative examples of biologically active molecules containing a -SCF₃ group are depicted in Figure 1.

Figure 1.

Active pharmaceutical ingredients containing a C-SCF₃ group.

Notably, all compounds in Figure 1 contain C-SCF₃ linkages; to the best of our knowledge no biological activity of any molecule with an N-SCF₃ bond has been reported until now.^12,13 Although the electron-withdrawing properties of -CF₃ (σ_m: 0.44) and -SCF₃ (σ_m: 0.40) are similar, the lipophilicity of the SCF₃ moiety (clogP = 1.95 for Me-SCF₃) is much higher than that of the -CF₃ group (clogP = 1.11 for Me-CF₃). The increased lipophilicity has been shown to increase membrane permeability, and in some cases, improve the metabolic stability of the final molecules.¹⁴ In this manuscript, we explore the interesting combination of the SIA and the -SCF₃ functional groups and report the unexpected antimycobacterial and cytotoxic activity that was linked to the trifluoromethylthio functional group.

Methods for selective introduction of -CF₃ or -SCF₃ groups typically make use of toxic reagents and harsh experimental conditions; still contemporary advances such as fluorine exchange reactions and gaseous trifluoromethylthiolating agents are noteworthy.¹⁵⁻¹⁷ Recently, much research has focused on using advanced electrophilic reagents such as N-trifluoromethylthio phthalimide, N-trifluoromethylthio saccharin, and trifluoromethanesulfanylanilides for introducing −CF₃ and -SCF₃ groups under mild conditions.¹⁸⁻²² To date there is only one example of N-trifluoromethylthiolation of a S=N functional group; Bolm and co-workers reported the preparation of N-trifluoromethylthio-sulfoximines through a two-step “umpolung” protocol in which the sulfoximine (SI) was first brominated to the corresponding N-bromo sulfoximine, which underwent nucleophilic attack by the ^–SCF₃ anion from silver trifluoromethanethiolate (AgSCF₃).²³ We decided to explore this methodology for trifluoromethylation of the related SIA functionality.

The starting materials for these studies, i.e. SIAs 6a–e,^24,2511a–d²⁶ and SIs 21a and 22a were prepared according to published procedures. Detailed descriptions of the synthetic sequences are reported in the Supporting Information, Schemes S1 and S2. Next, we employed Bolm’s procedure for trifluoromethylthiolation of SIs to these SIAs. First, silver trifluoromethanethiolate (AgSCF₃) was prepared by refluxing a suspension of AgF in CS₂.²⁷ Initially, we used substrate 6a as a model for optimizing the protocol. SIA 6a was treated with N-bromosuccinimide (NBS) to get its corresponding bromo derivative, which was treated with AgSCF₃ reagent (1.2 equiv) at room temperature. Indeed product 12 was formed but the reaction was incomplete. Compound 12 was identified by ¹⁹F, ¹H, ¹³C NMR and HRMS spectroscopy. The signals for the SCF₃ group in SIA 12 were found at δ −50.36 in the ¹⁹F, which is very similar to δ −50.69 and δ −50.39 observed for the similar N-SCF₃ sulfoximines 21 and 22. In contrast, the ¹⁹F chemical shifts of CF₃ substituted sulfoximines 23 and 24 have been observed at δ −41.83 and δ −40.97, respectively. Our efforts to expand Bolm’s N-trifluoromethylthiolation protocol to SIA derivatives proved successful without much additional optimization, see the SI for our attempt to optimize the reaction time. Using this methodology, we could prepare the novel N-trifluoromethylthio SIAs 12–20 in Scheme 1.

Scheme 1. Synthesis of N-SCF₃ Substituted SIAs.

The normally nucleophilic imine-N of SIA is subjected to “umpolung” through bromination and undergoes nucleophilic replacement with the CF₃S^– anion.

In order to compare structure activity (SAR) relationships, we also prepared a set of known -SCF₃ and -CF₃ functionalized sulfoximines. N-Trifluoromethylthio analogs 21 and 22(23) and the N-trifluoromethyl derivatives 23 and 24(28) were prepared using Bolm’s procedure as shown in Scheme 2.

Scheme 2. Synthesis of Known N-Trifluoromethylthio- and N-Trifluoromethyl-sulfoximines for SAR Analysis.

Up to now, N-trifluoromethylthiolated SIAs have never been synthesized; consequently, any possible biological activities of these compounds are yet to be explored. In this direction, we subjected several of the novel SIAs (Schemes S1 and S2 Supporting Information) and SCF₃–SIAs prepared in Schemes 1 and 2 for our routine screen of antibacterial activity vs the Gram-positive bacteria S. aureus and B. subtilis and the Gram-negative bacteria E. coli and P. aeruginosa. None of the SIAs with free imine-NH (e.g., 6 or 11) displayed antibacterial activities. Among the tested SCF₃–SIAs, only compound 13 had noteworthy activity vs B. subtilis.

Despite the poor antibacterial activity observed, a few compounds were evaluated for efficacy against M. tb. and M. abscessus (which are neither Gram-positive nor Gram-negative). To our surprise, SCF₃–SIAs 13 and 15 showed potent activity vs M. tb. with MIC as low as 4 μg/mL, which is comparable to WHO recommended drugs for TB such as ethambutol (1–4 μg/mL)²⁹ or the carbapenem class of antibiotics (Meropenem 5–10 μg/mL).³⁰ This promising discovery made us prepare and evaluate related SCF₃–SIs 21 and 22. Indeed, these N-SCF₃ functionalized analogs also proved to be potent inhibitors of M. tb. with MICs 4–8 μg/mL and with selectivity over Gram-positive and Gram-negative bacteria, Table 1. In addition to activity against M. tb., moderate activity against M. abscessus suggests that these compounds likely exhibit selectivity toward mycobacteria.

Table 1. Summary of Biological Activities Measured for -SCF₃ Sulfonimidamide and Sulfoximines and -CF₃ Sulfoximines Derivatives.

	MIC (μg/mL)
	Gram + Ve		Gram – Ve		MIC (μg/mL)		Cytotoxicity
Entry	S. aureus	B. subtilis	E. coli	P. aeruginosa	M. tb H37Rv	M. abscessus	HEPG2 (μg/mL)	LDHa (μg/mL)
6a	-	-	-	-	>64	>64	-	-
11c	-	-	-	-	>64	>64	-	-
12	32	32	128	256	-	-	-	-
13	32	8	128	512	4–8	>64	15	0.1
14	-	-	-	-	8–16	16–32	-	-
15	-	-	-	-	4–8	>64	65	0.1
16	64	32	>512	>512			-	-
17	128/256	128/256	128/256	256	16–32	32–64	-	-
18	32	64	-	-	>64	32–64	-	-
19	32	64	>512	>512	16–32	16–32	-	-
20	32/64	16/32	-	>512	16–32	16–32	-	-
21a	>512	>512	>512	>512	-	-	>100	100
21	16	16	16	128	4–8	>64	15	0.1
22	32	16	256	256	4–8	>64	90	1
23	>512	>512	>512	>512	>64	-	>200	-
24	>512	>512	>512	>512	>64	-	200	200

- = Not analyzed.

^aLowest concentration where LDH release was observed.

The activity of N-SCF₃ functionalized SIAs 13 and 15 and SIs 21 and 22 versus M. tb., but not vs Gram-positive or Gram-negative bacteria, suggested that we had identified a novel class of selective antitubercular agents. We therefore profiled the substance’s cytotoxic effects on a human-derived liver carcinoma cell line (HepG2). The MTT assay, a colorimetric measure of cellular metabolic activity, measured cell viability, and extracellular lactate dehydrogenase (LDH) was used as an indicator of necrotic cell death. Both compounds 13 (15 μg/mL; 20 μM) and 15 (65 μg/mL; 50 μM) reduced cell viability in HEPG2 cells as evidenced by the MTT assay, Table 1. Similarly, the SCF₃ SIs 21 and 22 showed cytotoxicity versus HepG2 cells, although at higher concentration for compound 22 (90 μg/mL; 300 μM) than 21 (14 μg/mL; 50 μM). Interestingly, the SI analog 21a, which lacks the SCF₃ group, only displayed toxicity above 650 μM.

It appears from the above results, that the SCF₃ group imposed toxicity to both M. tb. and cells of human origin when linked to the imine Nitrogen atom of sulfonimidamides and sulfoximines. In order to better understand the SAR, we prepared two known N-trifluoromethyl (CF₃) sulfoximines, i.e. 23 and 24 and evaluated their activity on M. tb. Intriguingly, none of these compounds demonstrated any anti-M. tb. activity. Likewise, 23 and 24 lacked cytotoxicity vs HepG2 cells. This matched-pair analysis further supports the hypothesis that the SCF₃ functional group is responsible for the observed cell toxicity. However, the -SCF₃ moiety is present in many advanced stage drug substances without known toxicity, e.g. Figure 1. Notably, in all these advanced molecules, the -SCF₃ group is situated on an aromatic ring, or in some cases on an aliphatic carbon, i.e. with a C-SCF₃ bond.

Although a full understanding of the toxicity induced by the -SCF₃ functionality is beyond the scope of this article, it is tempting to speculate that the N–SCF₃ bond in SIAs and SIs is labile and generates trifluoromercaptomethanol, or some other derivative, that is responsible for the cell killing effect. We observed no chemical instability of the N–SCF₃ bond in SIAs or SIs. Similarly, we could not detect a difference in Cystein-adduct formation for the SCF₃ SI 22 or the matched CF₃ SI 24, since no adducts were found for any of the compounds (only attempted in the absence of metabolic activation; see SI for details). We therefore hypothesize that a cellular metabolic process is involved in the liberation of toxic species. This assumption could also rationalize the difference in toxicity seen for the Gram-positive and Gram-negative bacterial strains tested and the M. tb and human cells. The compounds found to be active on M. tb. and HEPG2, i.e. 13, 15, 21, and 22, also displayed the highest antibacterial activity vs the investigated strains, although far less potent than their activity on M. tb. The difference in cellular toxicity can be rationalized by a difference in metabolizing enzymes or susceptibility to toxicity by the liberated toxicophore. Yet another observation supporting the involvement of N-SCF₃ bond stability in the liberation of the toxicophore relates to the electronic nature of the molecules. The M. tb. activity of the matched SIA pairs 14 vs 17, 15 vs 18, and 13 vs 19 demonstrates that the strongly electron withdrawing nitro-substituent always decreases the compound́s antibacterial effect.

In summary, we have demonstrated that the trifluoromethylthiolation protocol initially developed for sulfoximines can be expanded to sulfonimidamide substrates. We also report an unexpectedly high activity of N-trifluoromethylthiolated sulfonimidamides (SIAs) and sulfoximines (SIs) versus M. tb.—the causative pathogen of TB—the most widespread infectious disease that is responsible for millions of deaths every year. Further exploration of this finding demonstrated that both N-SCF₃ substituted SIAs and SIs display cytotoxicity versus cells of human origin, but more modestly vs Gram-positive and Gram-negative bacteria. Comparison with the corresponding N-CF₃ substituted SIs suggests that the SCF₃ functionality is responsible for the toxicity. As far as we know, the present study is the first report of biological activity of N-SCF₃ derivatives and the results obtained are of particular importance given the high utilization of the -SCF₃ group in contemporary drug design!

Supporting Information Available

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acsmedchemlett.9b00285.

• Experimental details for synthetic procedures, and associated chemical data for all final compounds (PDF)

Author Present Address

^∇ (NT) Department of Drug Design and Pharmacology, Faculty of Health and Medical Sciences, University of Copenhagen, Universitetsparken 2, 2100 Copenhagen, Denmark.

Author Present Address

^○ (PM) Chemical Science Research Group, Division of Research and Development, Lovely Professional University, Phagwara, Punjab 144411, India.

Author Contributions

NT, PM, KKR, and TN performed synthesis and compounds characterization; SN, NSA, and AAC performed and supervised cytotoxicity measurements; AK and GL performed TB testing; AMS, HGK, and TG performed and supervised antibacterial testing; PIA initiated the project and wrote the manuscript together with NT, PM, and TN. All authors contributed to the final version.

The authors declare no competing financial interest.