Design, synthesis and mechanistic exploration of anti-plasmodial Indolo[2,3-b]quinoxaline-7-chloroquinoline hybrids

Abstract

Aim: The aim of this study is to synthesize indolo[2,3-b]quinoxaline-4-aminoquinoline-based hybrids and evaluate their effectiveness against chloroquine-susceptible (3D7) and resistant (W2) Plasmodium falciparum strains, with expected inhibition of P. falciparum chloroquine resistance transporter (PfCRT) and heme.

Methods: The hybrids were synthesized and in vitro evaluated against both susceptible and resistant strains. Molecular docking and studies were conducted to assess the binding affinities for the PfCRT protein. Additionally, heme-inhibition studies using hemin chloride provided valuable insights into the interaction between the ligand and heme. The binding constant (logK) was calculated, providing quantitative details about the strength of this interaction.

Conclusion: The synthesized hybrids showed reasonable potency against both P. falciparum strains. The most potent hybrid 10d, with fluorine-substitution exhibited good activity. Molecular docking studies indicated strong binding affinities for the PfCRT protein. Heme inhibition studies further supported the potential of 10d as an effective anti-plasmodial agent.

Graphical Abstract

Plain language summary

Article highlights.

• A series of indolo[2,3-b]quinoxaline-7-chloroquinoline based hybrids were designed and synthesized and evaluated for their anti-plasmodial activity.

• The hybrids were tested against both chloroquine-sensitive (3D7) and chloroquine-resistant (W2) strains of Plasmodium falciparum. The results demonstrated that the compounds exhibited reasonable potency, suggesting that these hybrids could serve as potential candidates for overcoming chloroquine resistance in malaria treatment.

• Among the series, compound 10d, featuring a fluorine substitution, displayed notably superior activity against the chloroquine-resistant W2 strain.

• Molecular docking studies were conducted to investigate the binding interactions between the synthesized compounds and the PfCRT protein, a key target in chloroquine resistance. The results indicated that 10d, in particular, exhibited strong binding affinities for the PfCRT protein.

• Further validation of the anti-plasmodial potential of compound 10d was obtained through heme inhibition studies. These studies confirmed that compound 10d effectively inhibits heme detoxification within the parasite, thereby enhancing its activity. This dual mode of action underscores the compound's promise as an effective treatment option.

• Molecular modeling shows that both 10d and 10f bind strongly to PfCRT, with 10d exhibiting better interaction retention and 10f showing promising binding despite lower stability, highlighting their potential as effective PfCRT inhibitors.

1. Introduction

Malaria, a widespread infectious disease transmitted by female Anopheles mosquitoes, continues to be a major global health concern. In 2022, global malaria cases exceeded pre-COVID-19 levels, with around 249 million reported cases, up by five million from 2021. This resulted in about 608,000 deaths, as per the WHO's 2023 World Malaria Report [1]. Children under five and pregnant women are particularly vulnerable. The majority of human malaria cases are caused by P. falciparum, P. vivax and P. ovale [2].

For many years, compounds containing quinoline, such as chloroquine (CQ), amodiaquine and quinine, were the primary treatments for malaria. However, in the late 1950s, resistance to CQ emerged in P. falciparum, first in Southeast Asia and South America, before spreading to India and Africa [3,4]. In response, the World Health Organization (WHO) now advises using artemisinin combination therapy (ACT), which pairs artemisinin with quinoline-derived medications [5]. Nevertheless, resistance to artemisinin has also emerged in recent years [6].

Despite resistance, CQ retains its appeal for chemical modification due to its effective clinical outcomes, minimal harm to the host and simple, cost-efficient synthesis [7,8]. Its core, the 7-chloro-4-aminoquinoline structure, inhibits the formation of hematin, a defence mechanism of the parasite, thus exerting its anti-malarial effect [9]. Chloroquine resistance in malaria is intricately connected to specific mutations within the Plasmodium falciparum Chloroquine Resistance Transporter (PfCRT) gene. The resistant form of PfCRT, known as PfCRTCQR, plays a pivotal role in CQ resistance by actively pumping the drug out of the parasite's acidic digestive vacuole, where CQ exerts its antiplasmodial action. Importantly, PfCRTCQR can also reduce the parasite's susceptibility to other quinoline drugs. To counter this resistance mechanism, researchers have ingeniously modified the quinoline nucleus of these drugs. These modifications disrupt the expulsion of drugs from the digestive vacuole of parasite by interacting with PfCRT, offering a potential solution to combat CQ resistance [10]. Additionally, resistance-reversers further inhibit this expulsion, presenting a potential breakthrough [11]. Effective resistance-reversal properties, as proposed by Bhattacharjee et al., entail specific structural features, including two hydrophobic aromatic rings fused to another ring containing a nitrogen atom. This nitrogen-bearing ring should terminate with a three-carbon chain ending in an H-bond acceptor group, preferably nitrogen. Many other resistance-reversers share similar features, typically featuring two aromatic rings and a chain ending with a basic nitrogen atom [12]. Quinoxaline, fitting this description, also exhibits antiplasmodial properties [13].

In recent years, our research group has focused on the development and synthesis of new anti-plasmodial derivatives. This resulted in hybrids with good anti-plasmodial activity. For instance, the 4-aminoquinoline-naphthalimide hybrids (Figure 1(I)) displayed an IC₅₀ of 0.07 μM against the CQ-resistant W2 strain [14]. Similarly, fusing triazolopyrimidine with the 4-aminoquinoline core (Figure 1(II)) afforded potent compound with an IC₅₀ of 0.20 μM against the resistant strain [15]. Furthermore, aminoquinoline-oxindole hybrid (Figure 1(III)) exhibited an IC₅₀ of 0.03 μM against the W2 strain [16]. Importantly, all these hybrids demonstrated a dual mechanism of action, targeting both hemazoin inhibition and PfCRT function.

Figure 1.

Design Strategy for triazole-tethered indolo[2,3-b]quinoxaline-7-chloroquinoline hybrids.

Considering these factors, we initiated our work using indolo-quinoxaline derivatives as an initial step in this field, as there are no reports of indolo-quinoxalines as anti-plasmodial agents to the best of our knowledge. So, we designed and synthesized a series of hybrids by linking CQ and indolo-quinoxaline using an appropriate linker and a triazole ring. The triazole functionality was chosen due to its favorable properties, such as moderate dipole nature, hydrogen bonding capacity, structural stability and in vivo resilience, all contributing to its diverse biological activities [17]. Recent studies have highlighted the antimalarial potential of hybridized 1,2,3-triazoles [18]. In one investigation, a chloroquine-triazole derivative (Figure 1(IV)) showed an IC₅₀ of 0.16 μM against the K1 strain [19], while another study found that a 4-aminoquinoline-based 1,2,3-triazole (Figure 1(V)) displayed potent antimalarial activity with an IC₅₀ of 0.04 μM against the 3D7 P. falciparum strain [20].

Therefore, as step forward in our search for new antimalaria compounds [14–16], we decided to combine indolo[2,3-b]quinoxaline and 7-chloroquinoline fragments to make hybrids with dual functionality. In this work, we describe the synthesis of indolo[2,3-b]quinoxaline-7-chloroquinoline hybrids and evaluate their in vitro antiplasmodial efficacy against both CQ-susceptible and CQ-resistant P. falciparum strains. Additionally, we explore their potential modes of action, including inhibiting heme polymerization and binding to PfCRT.

2. Materials & methods

2.1. Experimental section

NMR spectra were obtained using dimethylsulfoxide (DMSO-d₆) with Jeol 400 (400 MHz), Bruker 500 (500 MHz) and Bruker 600 (600 MHz) spectrometers and TMS as an internal standard. Chemical shifts are reported in parts per million relative to TMS, and coupling constants (J) are given in hertz. Splitting patterns are indicated as follows: s (singlet), d (doublet), t (triplet), q (quartet), quin (quintet), m (multiplet), dd (doublet of doublets), td (triplet of doublets), ddd (doublet of doublet of doublets) and br (broad peak). High resolution mass spectra were performed on a Bruker-micrOTOF-Q II spectrometer. Microwave-assisted reactions were conducted using AntonPaar Monowave 200, employing sealed 10 ml process vials.

2.1.1. General procedure for synthesis of 1H-1,2,3-triazole-tethered indolo[2,3-b]quinoxaline-7-chloroquinoline hybrids (9a-e & 10a-j)

In a microwave vial, 2 ml dimethylsulfoxide (DMSO), catalytic amount of CuI (0.05 mmol) was added along with the substituted propargylated precursor 3a-e (1.0 mmol) and azido-precursors (5 and 8a,b) (1.0 mmol). The mixture was heated to 80°C for 10 min in the presence of N,N’-diisopropylethylamine (0.2 ml). After confirming the completion of the reaction via thin-layer chromatography (TLC), ice-cold water (25 ml) was added in the mixture. The resulting precipitates were filtered, followed by washing with saturated solution of ethylenediaminetetraacetic acid (EDTA) (10 × 2 ml) to ensure the removal of residual copper content. The precipitates were then air-dried to obtain the indolo[2,3-b]quinoxaline-7-chloroquinoline click adducts in good yield (71–89%).

No safety hazards were encountered during the synthesis if target molecules.

2.1.2. Spectral data of click adducts

2.1.2.1. 6-((1-(7-chloroquinolin-4-yl)-1H-1,2,3-triazol-4-yl)methyl)-6H-indolo[2,3-b]quinoxaline (9a)

Yellow solid (82% yield); decompose above 228°C; ¹H NMR (500 MHz, DMSO-d₆) δ 8.20 (d, J = 4.0 Hz, 1H, Ar-H), 8.01 (s, 1H, triazole-H), 7.65 (d, J = 8.8 Hz, 1H, Ar-H), 7.51–7.42 (m, 1H, Ar-H), 7.37 (d, J = 9.0 Hz, 1H, Ar-H), 7.19–7.11 (m, 2H, Ar-H), 7.09–7.02 (m, 1H, Ar-H), 6.85 (d, J = 8.8 Hz, 1H, Ar-H), 6.72–6.68 (m, 2H, Ar-H), 6.67 (d, J = 3.8 Hz, 1H, Ar-H), 6.62–6.57 (m, 1H, Ar-H), 6.51–6.46 (m, 1H, Ar-H), 5.09 (s, 2H, -N-CH₂-CH-). ¹³C NMR (125 MHz, DMSO-d₆) δ 152.55, 150.05, 148.67, 147.46, 146.64, 144.04, 142.47, 142.18, 140.68, 134.85, 133.48, 132.87, 132.75, 130.78, 129.87, 128.04, 127.85, 127.06 126.85, 125.41, 125.20, 123.67, 121.84, 112.95, 112.31, 108.64, 54.53. HRMS calcd for C₂₆H₁₆ClN₇ [M]⁺ 461.1156 and [M + 2] 463.1126 and found 461.1168 and 463.1139.

2.1.2.2. 6-((1-(7-chloroquinolin-4-yl)-1H-1,2,3-triazol-4-yl)methyl)-9-fluoro-6H-indolo[2,3-b]quinoxaline (9b)

Yellow solid (85% yield); decompose above 110°C; ¹H NMR (500 MHz, DMSO-d₆) δ 8.42 (d, J = 7.6 Hz, 1H, Ar-H), 8.30 (dd, J = 8.3, 1.1 Hz, 1H, Ar-H), 8.24 (s, 1H, triazole-H), 8.17 (dd, J = 8.4, 1.0 Hz, 1H, Ar-H), 7.93 (dd, J = 13.4, 8.7 Hz, 2H, Ar-H), 7.85 (ddd, J = 8.3, 6.9, 1.4 Hz, 1H, Ar-H), 7.83–7.77 (m, J = 11.9, 4.7 Hz, 2H, Ar-H), 7.77–7.74 (m, 2H, Ar-H), 7.71 (dd, J = 9.1, 1.9 Hz, 1H, Ar-H), 7.45 (t, J = 7.3 Hz, 1H, Ar-H), 5.98 (s, 2H, -N-CH₂-CH-). ¹³C NMR (150 MHz, DMSO-d₆) δ 150.65, 149.56, 147.81, 146.90, 145.51, 142.96, 142.62, 141.88, 134.43, 133.48, 133.32, 132.18, 130.60, 129.39, 128.56, 127.47, 127.08, 126.56, 125.98, 125.27, 124.37, 121.79, 113.86, 112.55, 82.21, 55.16. HRMS calcd for C₂₆H₁₅ClFN₇ [M]⁺ 479.1061 and [M + 2] 481.1032 and found 479.1075 and 481.1044.

2.1.2.3. 4-(6-((1-(7-chloroquinolin-4-yl)-1H-1,2,3-triazol-4-yl)methyl)-6H-indolo[2,3-b]quinoxalin-9-yl)morpholine (9c)

Yellow solid (79% yield); decompose above 214°C; ¹H NMR (600 MHz, DMSO-d₆) δ 8.27 (d, J = 4.6 Hz, 1H, Ar-H), 8.18 (s, 1H, triazole-H), 7.86 (d, J = 8.9 Hz, 1H, Ar-H), 7.64–7.58 (m, 1H, Ar-H), 7.45 (d, J = 9.0 Hz, 1H, Ar-H), 7.37–7.32 (m, 1H, Ar-H), 7.22–7.18 (m, 1H, Ar-H), 6.99–6.94 (m, 1H, Ar-H), 6.89–6.84 (m, 2H, Ar-H), 6.78 (d, J = 3.7 Hz, 1H, Ar-H), 6.72–6.68 (m, 1H, Ar-H), 6.54–6.49 (m, 1H, Ar-H), 5.57 (s, 2H, -N-CH₂-CH-), 3.97–3.94 (m, 4H, -CH₂-O-CH₂-), 3.69–3.66 (m, 4H, -CH₂-N-CH₂-). ¹³C NMR (150 MHz, DMSO-d₆) δ 152.85, 150.84, 148.95, 147.59, 147.31, 142.54, 142.53, 140.28, 133.84, 133.82, 133.47, 133.45, 130.36, 128.94, 128.57, 127.11, 127.05, 125.96, 125.58, 125.27, 123.37, 120.95, 112.66, 112.42, 105.61, 55.16. 53.28, 46.69, 46. HRMS calcd for C₃₀H₂₃ClN₈O [M]⁺ 546.1683 and [M + 2] 547.1717 and found 546.1691 and 547.1729.

2.1.2.4. 6-((1-(7-chloroquinolin-4-yl)-1H-1,2,3-triazol-4-yl)methyl)-9-(piperidin-1-yl)-6H-indolo[2,3-b]quinoxaline (9d)

Olive green solid (87% yield); decompose above 127°C; ¹H NMR (500 MHz, DMSO-d₆) δ 8.51 (s, 1H, Ar-H), 8.28 (d, J = 7.7 Hz, 1H, Ar-H), 8.14 (d, J = 8.1 Hz, 1H, Ar-H), 8.1 (s, 1H, triazole-H), 7.87 (dd, J = 13.3, 7.7 Hz, 3H, Ar-H), 7.77 (t, J = 7.3 Hz, 3H, Ar-H), 7.62 (d, J = 8.8 Hz, 1H, Ar-H), 7.43–7.25 (m, 2H, Ar-H), 5.78 (s, 2H, -N-CH₂-CH-), 4.46–4.34 (m, 4H, -CH₂-N-CH₂-), 3.06–2.94 (m, 4H, -CH₂-CH₂-CH₂-), 2.19–2.08 (m, 2H, -CH₂-CH₂-CH₂-). ¹³C NMR (125 MHz, DMSO-d₆) δ 149.89, 149.54, 148.64, 147.55, 146.51, 141.45, 141.34, 140.64, 132.63, 131.35, 130.47, 129.63, 129.38, 128.53, 128.42, 126.64, 126.15, 125.55, 124.54, 124.19, 123.35, 122.89, 114.46, 113.34, 109.45, 57.23. 51.16, 50.95, 25.18, 25.07, 23.42. HRMS calcd for C₃₁H₂₅ClN₈ [M]⁺ 544.1891 and [M + 2] 546.1861 and found 544.1900 and 546.1873.

2.1.2.5. 6-((1-(7-chloroquinolin-4-yl)-1H-1,2,3-triazol-4-yl)methyl)-N,N-diethyl-6H-indolo[2,3-b]quinoxalin-9-amine (9e)

Dark brown solid (75% yield); decompose above 130°C; ¹H NMR (500 MHz, DMSO-d₆) δ 8.35 (s, 1H, Ar-H), 8.29 (d, J = 7.5 Hz, 1H, Ar-H), 8.17 (d, J = 7.7 Hz, 2H, Ar-H), 8.09 (s, 1H, triazole-H), 7.76–7.68 (m, 3H, Ar-H), 7.66 (t, J = 8.8 Hz, 3H, Ar-H), 7.25 (s, 2H, Ar-H), 5.27 (s, 2H, -N-CH₂-CH-), 2.67–2.55 (m, 4H, -CH₂-N-CH₂-), 1.27–1.20 (m, 6H, 2 × CH₃). ¹³C NMR (150 MHz, DMSO-d₆) δ 155.35, 148.51, 146.44, 143.41, 143.27, 142.22, 138.41, 137.84, 132.65, 132.27, 132.00, 131.94, 131.17, 131.09, 130.74, 129.63, 129.27, 128.93, 128.47, 125.84, 124.95, 123.26, 122.07, 120.08, 111.36, 48.39, 39.49, 36.75, 32.09, 31.77. HRMS calcd for C₃₀H₂₅ClN₈ [M]⁺ 532.1891 and [M + 2] 534.1861 and found 532.1898 and 534.1875.

2.1.2.6. N-(2-(4-((6H-indolo[2,3-b]quinoxalin-6-yl)methyl)-1H-1,2,3-triazol-1-yl)ethyl)-7-chloroquinolin-4-amine (10a)

Brown solid (89% yield); decompose above 116°C; ¹H NMR (400 MHz, DMSO-d₆) δ 8.32 (d, J = 7.5 Hz, 1H, Ar-H), 8.23 (d, J = 8.2 Hz, 1H, Ar-H), 8.04–8.00 (m, J = 6.7 Hz, 2H, Ar-H + triazole-H), 7.80–7.74 (m, 2H, Ar-H), 7.74–7.61 (m, 4H, Ar-H), 7.43–7.25 (m, 3H, Ar-H), 7.22 (d, J = 9.1 Hz, 1H, Ar-H), 5.69 (s, 2H, -N-CH₂-CH-), 4.51 (t, J = 5.2 Hz, 2H, -N-CH₂-), 3.65–3.57 (m, 2H, -NH-CH₂-). ¹³C NMR (150 MHz, DMSO-d₆) δ 147.81, 146.90, 145.51, 142.96, 142.62, 141.88, 134.43, 132.18, 130.60, 129.39, 127.47, 127.08, 125.27, 124.37, 121.79, 113.86, 82.21, 50.96, 45.37, 39.38. HRMS calcd for C₂₈H₂₁ClN₈ [M]⁺ 504.1578 and [M + 2] 506.1548 and found 504.1589 and 506.1560.

2.1.2.7. N-(3-(4-((6H-indolo[2,3-b]quinoxalin-6-yl)methyl)-1H-1,2,3-triazol-1-yl)propyl)-7-chloroquinolin-4-amine (10b)

Light yellow solid (84% yield); decompose above 106°C; ¹H NMR (400 MHz, DMSO-d₆) δ 8.35 (d, J = 7.7 Hz, 1H, Ar-H), 8.25 (d, J = 7.5 Hz, 1H, Ar-H), 8.11 (s, 1H, Ar-H), 8.09 (s, 1H, triazole-H), 7.79 (t, J = 7.1 Hz, 2H, Ar-H), 7.75–7.66 (m, 4H, Ar-H), 7.38 (t, J = 6.9 Hz, 3H, Ar-H), 7.29 (s, 1H, Ar-H), 5.75 (s, 2H, -N-CH₂-CH-), 4.37 (t, J = 6.9 Hz, 2H, -N-CH₂-), 3.16–3.06 (m, 2H, -NH-CH₂-), 2.15–2.00 (m, 2H, -CH₂-CH₂-CH₂-). ¹³C NMR (100 MHz, DMSO-d₆) δ 147.44, 146.46, 145.99, 145.30, 144.38, 140.43, 140.10, 139.35, 134.29, 133.72, 131.93, 129.67, 129.63, 128.09, 126.83, 125.78, 124.87, 124.55, 124.05, 122.76, 121.85, 119.27, 118.86, 111.38, 47.91, 36.90, 28.80. HRMS calcd for C₂₉H₂₃ClN₈ [M]⁺ 518.1734 and [M + 2] 520.1705 and found 518.1741 and 520.1718.

2.1.2.8. 7-chloro-N-(2-(4-((9-fluoro-6H-indolo[2,3-b]quinoxalin-6-yl)methyl)-1H-1,2,3-triazol-1-yl)ethyl)quinolin-4-amine (10c)

Yellow solid (71% yield); decompose above 120°C; ¹H NMR (600 MHz, DMSO-d₆) δ 8.26 (s, 1H, Ar-H), 8.19 (d, J = 8.4 Hz, 1H, Ar-H), 8.10 (s, 1H, triazole-H), 7.96 (d, J = 8.6 Hz, 2H, Ar-H), 7.85 (dd, J = 14.0, 8.2 Hz, 2H, Ar-H), 7.76 (dd, J = 13.7, 8.0 Hz, 2H, Ar-H), 7.52 (d, J = 9.1 Hz, 1H, Ar-H), 7.36–7.28 (m, 3H, Ar-H) 5.76 (s, 2H, -N-CH₂-CH-), 4.37 (t, J = 6.8 Hz, 2H, -N-CH₂-), 3.35–3.26 (m, 2H, -NH-CH₂-). ¹³C NMR (150 MHz, DMSO-d₆) δ 151.88, 148.56, 145.52, 143.84, 142.46, 141.61, 139.98, 139.61, 137.56, 135.24, 130.45, 130.08, 128.82, 128.25, 127.55, 126.24, 126.05, 125.92, 124.44, 122.25, 121.91, 121.52, 118.47, 118.08, 117.75, 111.74, 36.35, 28.85, 28.83. HRMS calcd for C₂₈H₂₀ClFN₈ [M]⁺ 522.1483 and [M + 2] 524.1454 and found 522.1491 and 524.1468.

2.1.2.9. 7-chloro-N-(3-(4-((9-fluoro-6H-indolo[2,3-b]quinoxalin-6-yl)methyl)-1H-1,2,3-triazol-1-yl)propyl)quinolin-4-amine (10d)

Yellow solid (78% yield); decompose above 107°C; ¹H NMR (600 MHz, DMSO-d₆) δ 8.58 (s, 1H, Ar-H), 8.35 (d, J = 9.1 Hz, 1H, Ar-H), 8.21 (d, J = 9.1 Hz, 1H, Ar-H), 8.17 (s, 1H, triazole-H), 7.94 (dd, J = 14.2, 8.5 Hz, 3H, Ar-H), 7.84 (dd, J = 13.8, 8.6 Hz, 2H, Ar-H), 7.69 (d, J = 9.5 Hz, 1H, Ar-H), 7.56–7.33 (m, 3H, Ar-H) 5.84 (s, 2H, -N-CH₂-CH-), 4.45 (t, J = 7.0 Hz, 2H, -N-CH₂-), 3.20–3.08 (m, 2H, -NH-CH₂-), 2.24–2.12 (m, 2H, -CH₂-CH₂-CH₂-). ¹³C NMR (150 MHz, DMSO-d₆) δ 152.18, 149.25, 144.87, 143.86, 143.37, 140.13, 139.86, 139.63, 138.83, 135.24, 131.42, 129.08, 129.02, 128.84, 128.01, 127.53, 126.27, 125.78, 125.31, 122.22, 122.19, 121.35, 118.83, 117.08, 117.00, 110.84, 36.19, 28.90, 28.59. HRMS calcd for C₂₉H₂₂ClFN₈ [M]⁺ 536.1640 and [M + 2] 538.1610 and found 536.1657 and 538.1621.

2.1.2.10. 7-chloro-N-(2-(4-((9-morpholino-6H-indolo[2,3-b]quinoxalin-6-yl)methyl)-1H-1,2,3-triazol-1-yl)ethyl)quinolin-4-amine (10e)

Green solid (85% yield); decompose above 217°C; ¹H NMR (600 MHz, DMSO-d₆) δ 8.46 (s, 1H, Ar-H), 8.24 (d, J = 8.0 Hz, 1H, Ar-H), 8.03 (d, J = 5.1 Hz, 2H, Ar-H), 7.96 (s, 1H, triazole-H), 7.82 (m, 2H, Ar-H), 7.80–7.75 (m, 2H, Ar-H), 7.66–7.55 (m, 2H, Ar-H), 7.48–7.36 (m, 1H, Ar-H), 7.21–7.11 (m, 1H, Ar-H), 5.66 (s, 2H, -N-CH₂-CH-), 4.59–4.50 (m, 6H, -N-CH₂- + -CH₂-O-CH₂-), 3.35–3.31 (m, 4H, -CH₂-N-CH₂-), 2.27–2.23 (m, 2H, -NH-CH₂-). ¹³C NMR (150 MHz, DMSO-d₆) δ 151.18, 148.82, 147.81, 145.28, 144.22, 143.81, 142.81, 141.28, 139.81, 138.65, 135.25, 143.56, 128.81, 127.81, 126.24, 126.52, 125.62, 125.18, 124.21, 123.52, 121.48, 120.48, 119.53, 117.53, 116.25, 114.45, 111.52, 52.17, 36.41, 36.01, 27.15, 26.96. HRMS calcd for C₃₂H₂₈ClN₉O [M]⁺ 589.2105 and [M + 2] 591.2076 and found 589.2117 and 591.2088.

2.1.2.11. 7-chloro-N-(3-(4-((9-morpholino-6H-indolo[2,3-b]quinoxalin-6-yl)methyl)-1H-1,2,3-triazol-1-yl)propyl)quinolin-4-amine (10f)

Yellowish green solid (83% yield); decompose above 214°C; ¹H NMR (500 MHz, DMSO-d₆) δ 8.28 (d, J = 6.9 Hz, 1H, Ar-H), 8.18 (s, 1H, Ar-H), 8.14 (d, J = 6.5 Hz, 1H, Ar-H), 8.10 (s, 1H, triazole-H), 7.87–7.73 (m, 4H, Ar-H), 7.68–7.55 (m, 2H, Ar-H), 7.43–7.30 (m, 3H, Ar-H), 5.78 (s, 2H, -N-CH₂-CH-), 4.49–4.31 (m, 2H, -N-CH₂-), 3.11–2.95 (m, 4H, -CH₂-O-CH₂-), 2.53 (s, 4H, -CH₂-N-CH₂-), 2.22–1.98 (m, 4H, -NH-CH₂- + CH₂-CH₂-CH₂-). ¹³C NMR (150 MHz, DMSO-d₆) δ 150.26, 147.95, 147.38, 145.91, 144.86, 142.51, 141.64, 140.58, 139.55, 138.56, 137.85, 134.96, 128.81, 127.63, 127.14, 126.52, 125.88, 125.45, 124.51, 122.56, 121.64, 120.56, 119.54, 118.38, 115.65, 113.59, 109.68, 51.51, 38.93, 28.22, 26.28. HRMS calcd for C₃₃H₃₀ClN₉O [M]⁺ 603.2262 and [M + 2] 605.2232 and found 603.2271 and 605.2245.

2.1.2.12. 7-chloro-N-(2-(4-((9-(piperidin-1-yl)-6H-indolo[2,3-b]quinoxalin-6-yl)methyl)-1H-1,2,3-triazol-1-yl)ethyl)quinolin-4-amine (10g)

Olive green solid (79% yield); decompose above 130°C; ¹H NMR (500 MHz, DMSO-d₆) δ 8.32 (s, 1H, Ar-H), 8.20 (s, 1H, Ar-H), 8.16 (d, J = 8.8 Hz, 2H, Ar-H), 8.02 (s, 1H, triazole-H), 7.78 (dd, J = 13.8, 7.6 Hz, 2H, Ar-H), 7.65 (dd, J = 13.5, 7.4 Hz, 2H, Ar-H), 7.60–7.54 (m, 2H, Ar-H), 7.46 (s, 1H, Ar-H), 7.27 (m, 1H, Ar-H), 5.47 (s, 2H, -N-CH₂-CH-), 4.37–4.19 (m, 2H, -N-CH₂-), 2.64–2.60 (m, 4H, -CH₂-N-CH₂-), 2.12–2.07 (m, 2H, -NH-CH₂-), 1.30–1.17 (m, 6H, 2×-CH₂). ¹³C NMR (125 MHz, DMSO-d₆) δ 149.18, 147.84, 146.91, 146.37, 144.85, 143.74, 142.75, 142.28, 140.30, 137.64, 136.21, 133.76, 128.46, 127.61, 127.24, 127.16, 126.87, 125.51, 124.45, 123.58, 122.45, 121.45, 118.53, 117.64, 116.61, 113.45, 110.84, 50.49, 37.57, 37.48, 26.77, 26.16, 25.25. HRMS calcd for C₃₄H₃₂ClN₉ [M]⁺ 601.2469 and found 601.2480. HRMS calcd for C₃₃H₃₀ClN₉ [M]⁺ 587.2313 and [M + 2] 289.2283 and found 587.2325 and 589.2294.

2.1.2.13. 7-chloro-N-(3-(4-((9-(piperidin-1-yl)-6H-indolo[2,3-b]quinoxalin-6-yl)methyl)-1H-1,2,3-triazol-1-yl)propyl)quinolin-4-amine (10h)

Yellowish green solid (83% yield); decompose above 170°C; ¹H NMR (500 MHz, DMSO-d₆) δ 8.52 (s, 1H, Ar-H), 8.29 (s, 2H, Ar-H), 8.20–8.13 (m, 2H, Ar-H), 8.12 (s, 1H, triazole-H), 7.88 (s, 2H, Ar-H), 7.77 (s, 2H, Ar-H), 7.68–7.52 (m, 2H, Ar-H), 7.37 (s, 1H, Ar-H), 5.78 (s, 2H, -N-CH₂-CH-), 4.43–4.30 (m, 2H, -N-CH₂-), 2.59–2.53 (m, 6H, -CH₂-N-CH₂- + -NH-CH₂-), 2.09–2.07 (m, 2H, -CH₂-CH₂-CH₂-), 1.27–1.11 (m, 6H, 2×-CH₂). ¹³C NMR (125 MHz, DMSO-d₆) δ 154.18, 150.27, 148.00, 145.17, 144.86, 144.47, 142.64, 141.16, 140.33, 139.82, 136.84, 132.85, 129.78, 129.09, 128.67, 128.27, 127.94, 126.44, 125.86, 124.41, 122.57, 122.34, 121.57, 119.13, 117.87, 117.08, 111.55, 109.94, 49.52, 37.75, 36.20, 27.83, 26.16, 25.25. HRMS calcd for C₃₄H₃₂ClN₉ [M]⁺ 601.2469 and [M + 2] 603.2440 and found 601.2480 and 603.2449.

2.1.2.14. 6-((1-(2-((7-chloroquinolin-4-yl)amino)ethyl)-1H-1,2,3-triazol-4-yl)methyl)-N,N-diethyl-6H-indolo[2,3-b]quinoxalin-9-amine (10i)

Olive green solid (77% yield); decompose above 190°C; ¹H NMR (500 MHz, DMSO-d₆) δ 8.50 (s, 1H, Ar-H), 8.30 (d, J = 6 Hz, 1H, Ar-H), 8.09 (d, J = 6.1 Hz, 2H, Ar-H), 8.04 (s, 1H, triazole-H), 7.89–7.82 (m, 2H, Ar-H), 7.82–7.75 (m, 3H, Ar-H), 7.69 (s, 1H, Ar-H), 7.41 (s, 1H, Ar-H), 7.27 (s, 1H, Ar-H), 5.74 (s, 2H, -N-CH₂-CH-), 4.59–4.52 (m, 2H,-N-CH₂-), 2.56–2.55 (m, 6H, -CH₂-N-CH₂- + -NH-CH₂-), 1.27–1.19 (m, 6H, 2 × CH₃). ¹³C NMR (125 MHz, DMSO-d₆) δ 152.65, 149.54, 147.14, 144.51, 143.81, 140.48, 139.18, 138.89, 133.55, 133.29, 132.52, 131.45, 131.15, 130.59, 130.41, 129.44, 128.25, 127.93, 127.17, 125.61, 124.15, 123.14, 121.87, 120.51, 110.54, 48.51, 45.49, 39.95, 37.64, 37.60, 21.89, 21.85. HRMS calcd for C₃₂H₃₀ClN₉ [M]⁺ 575.2313 and [M + 2] 577.2283 and found 575.2327 and 577.2296.

2.1.2.15. 6-((1-(3-((7-chloroquinolin-4-yl)amino)propyl)-1H-1,2,3-triazol-4-yl)methyl)-N,N-diethyl-6H-indolo[2,3-b]quinoxalin-9-amine (10j)

Dark green solid (72% yield); decompose above 207°C; ¹H NMR (400 MHz, DMSO-d₆) δ 8.48 (s, 1H, Ar-H), 8.24 (d, J = 7.9 Hz, 1H, Ar-H), 8.11 (d, J = 7.6 Hz, 2H, Ar-H), 8.07 (s, 1H, triazole-H), 7.87–7.78 (m, 3H, Ar-H), 7.74 (t, J = 8.9 Hz, 3H, Ar-H), 7.35 (s, 2H, Ar-H), 5.74 (s, 2H, -N-CH₂-CH-), 4.41–4.32 (m, 2H,-N-CH₂-), 3.55–3.53 (m, 2H, -NH-CH₂-), 2.51–2.49 (m, 6H, -CH₂-N-CH₂- + -CH₂-CH₂-CH₂-), 1.20–1.16 (m, 6H, 2 × CH₃). -). ¹³C NMR (100 MHz, DMSO-d₆) δ 150.85, 148.55, 146.15, 145.65, 144.45, 141.54, 141.15, 140.41, 139.54, 137.58, 136.15, 133.26, 129.21, 128.52, 127.18, 126.41, 125.68, 124.64, 123.58, 122.41, 121.51, 120.15, 118.51, 117.41, 112.12, 107.44, 51.85, 49.15, 38.95, 37.91, 26.51, 26.25, 24.61. HRMS calcd for C₃₃H₃₂ClN₉ [M]⁺ 589.2469 and [M + 2] 591.2440 and found 589.2477 and 591.2449.

2.2. Molecular docking protocols

Molecular docking was conducted with Schrödinger's suite (Schrödinger Release 2022-1: Maestro, Schrödinger, LLC, New York, NY). The single-particle structure of PfCRT (PDB ID:6UKJ) was obtained from the Protein Data Bank (www.rcsb.org). Initial steps included assigning bond orders, adding hydrogens and removing water molecules. To account for protein permeability and drug solubility, heteroatoms were ionized at biological pH using the EPIC tool and hydrogen bonds were optimized to minimize steric clashes, focusing on histidine, aspartate, glutamate and hydroxyl-containing amino acids. The protein structure was then minimized using the OPLS 2005 force field.

Ligand preparation involved generating the lowest energy state for the ligand, which included converting 1D/2D structures to 3D using LigPrep. The ligand was also minimized with the OPLS 2005 force field. For docking, grids were created using a grid-based energy descriptor with default settings and a Van der Waals radius of 1.0. The docking process utilized the Glide program in extra precision (XP) mode with the previously prepared receptor grid and ligand molecules. Interactions between the ligand and receptor, such as hydrogen bonding, hydrophobic interactions and π-π stacking, were evaluated and scored using the Glide ligand docking program [21–25].

2.3. Molecular dynamics protocols

Molecular dynamics (MD) simulations of both complexes (PfCRT-10d and PfCRT-10f) were conducted using the Desmond module [26]. The protein was then embedded in a pre-equilibrated 1-palmitoyl-2-oleoyl-glycero-3-phosphocholine (POPC) membrane model at 300 K. The systems were solvated with a simple point charge (SPC) water model in an orthorhombic box, maintaining periodic boundary conditions with a 10 Å buffer in all dimensions. To neutralize the system, the appropriate number of chloride ions was added and experimental conditions were simulated by introducing 0.15 M NaCl. MD simulations were then carried out for a period of 0–100 ns, with trajectories recorded at 100 ps intervals. The simulations used an NPγT ensemble, maintaining a constant temperature of 300 K and pressure of 1.01325 bar, employing the Nose-Hoover chain thermostat method. Before each simulation, the systems were relaxed and the final results were analyzed using the simulation interaction diagram tool.

2.4. Biological evaluation

The Plasmodium falciparum strains, the chloroquine-resistant strain W2 (from Indochina;) and the chloroquine-susceptible 3D7 (from West Africa) were obtained from The Malaria Research and Reference Reagent Resource Center (MR4) (Manassas, VA, USA). They were cultured in RPMI 1640 (Invitrogen, Paisley, UK) supplemented with 10% human serum (Abcys S.A., Paris, France) and buffered with 25 mM 2-[4-(2-hydroxyethyl)piperazin-1-yl]ethanesulfonic acid (HEPES) and 25 mM NaHCO₃. Parasites were cultured in A-positive human blood (Etablissement Français du Sang, Marseille, France) under controlled conditions (10% O₂, 5% CO₂ and 85% N₂) at 37°C with 95% humidity. Before use, both Plasmodium falciparum strains were synchronized twice using sorbitol [27]. Clonality was confirmed by PCR genotyping of msp1, msp2 and microsatellite loci in both our lab and an independent lab Worldwide Antimalarial Resistance Network (WWARN) [28,29]. Compounds were dissolved and diluted in RPMI-DMSO 5% (v/v) to achieve 11 concentrations, ranging from 0.5 nM to 12,000 nM (0.5% DMSO final concentration). All the standard antimalarial drugs were provided from Sigma-Aldrich (Saint Quentin Fallavier, France). Stock solution of CQ diphosphate was prepared in water and then diluted in RPMI in order to have 11 final concentrations ranging from 6.3 nM to 3231 nM. Standard antimalarial drugs were sourced from Sigma-Aldrich (Saint Quentin Fallavier, France). Chloroquine stock was diluted in RPMI to obtain final concentrations from 6.3 nM to 3231 nM. Stock solutions of quinine and mefloquine were prepared in methanol and subsequently diluted in RPMI, resulting in 11 final concentrations ranging from 6.2 to 3149 nM and 1.5 to 392 nM, respectively. These standard antimalarial drugs were used as benchmarks for assessing Plasmodium falciparum activity.

For the in vitro analysis, 25 μl of each synthesized hybrid concentration was combined with 200 μl of parasitized red blood cells (final parasitemia, 0.1%; final haematocrit, 1.5%) in 96-well plates. These plates were then incubated for 72 h at 37°C under controlled atmospheric conditions (5% CO₂, 10% O₂ and 75% N₂). Following incubation, the plates were thawed and the hemolyzed cultures were mixed by vortexing to ensure homogenization. The drug susceptibility assay was conducted using the HRP2 ELISA-based assay Malaria Ag Celisa kit (ref KM2159, Cellabs PTY LDT, Brookvale, Australia) following the previously described method [30].

The concentration at which the drugs inhibited 50% of parasite growth (IC₅₀) was determined using the inhibitory sigmoid E_max model, with the IC₅₀ values calculated via non-linear regression in R software (IC Estimator version 1.2). The antiplasmodial in vitro assay was performed five-times for each compound, with the IC₅₀ values representing the average from five independent experiments.

2.5. Cytotoxicity assay on Vero cell lines

Vero kidney epithelial cells (ATCC^® CCL-81™) cultured in RPMI 1640 medium with 10% fetal calf serum at 37°C and 5% CO₂. Upon reaching the appropriate confluency, the cells were detached from the flask surface using trypsin, centrifuged at 220 × g for 5 min and the resulting pellet was resuspended in 1 ml of culture medium. Cells were counted using a Malassez chamber and seeded at approximately 2 × 10⁴ cells per well in 96-well plates, allowing them to adhere overnight. For cytotoxicity assessment, serial dilutions of synthetic compounds (100 μg/ml to 3.1 μg/ml) were prepared and the cultures were incubated for 24 h at 37°C and 5% CO₂. After incubation, 10% (v/v) resazurin was added to each well and the plates were returned to the incubator at 37°C with 5% CO₂. Fluorescence was measured using a plate reader (excitation wavelength: 540 nm and emission wavelength: 590 nm). DMSO was used as a negative control and 20% SDS as a positive control at predetermined concentrations. The half-maximal inhibitory concentration (IC₅₀) was calculated from a non-linear regression dose-response curve using the GraphPad Software (GraphPad Prism 5). The Selectivity Index (SI) was determined by dividing the IC₅₀ by the minimum inhibitory concentration (MIC).

2.6. Binding studies with monomeric heme

7.8 mg of hemin chloride was dissolved in 10 ml of DMSO to prepare a stock solution of hemin chloride with final concentration of 1.2 mM. To make a working solution (12 μM), 00 μl of this stock solution was diluted with 1 ml of 0.02 M HEPES buffer (pH 7.4) and 4 ml of DMSO and the final volume was adjusted to 10 ml using HPLC-grade water. The resulting 40% aqueous DMSO solution maintain the hemin solutions in monomeric state at concentrations used. Similarly, a 10 mM stock solution of compound 10d was prepared in DMSO. Then, 100 μl of this stock solution was diluted with 1 ml of 0.02 M HEPES buffer (pH 7.4) and 4 ml of DMSO, with the final concentration adjusted to make a working solution of 100 μM. All working solutions were stored in the dark to prevent photo-degradation. Diluent solutions were prepared by mixing 1 ml of 0.02 M HEPES buffer with 4 ml of spectroscopic-grade DMSO and bringing the volume up to 10 ml by adding 5 ml of HPLC-grade water. For the titration, 2 ml of the hemin chloride working solution was titrated with increasing concentrations of 10d. After each addition of the 10d, the absorbance was recorded. A similar titration was performed for the hemin chloride and 10d solutions at pH 5.6, using 2-(N-morpholino)ethanesulfonic acid (MES) buffer (pH 5.4).

3. Results

3.1. Synthetic chemistry

The synthesis of triazole-bearing indolo[2,3-b]quinoxaline-7-chloroquinoline hybrids involved a series of steps. Initially, o-phenylenediamine was subjected to a reaction with substituted isatins 1 (synthesized using reported protocols [31]) in the presence of acetic acid, leading to the formation of functionalized 6H-indolo[2,3-b]quinoxalines 2. The subsequent reaction of 2 with propargyl bromide facilitated the synthesis of alkyne precursors 3 (Figure 2A).

Figure 2.

(A) Synthesis of precursor 3a-e; (B) synthesis of precursor 5 and 8a,b; (C) synthesis of indolo[2,3-b]quinoxaline-7-chloroquinoline click adducts 9a-e and 10a-j.

Another important aspect of the synthesis involved the reaction of 4,7-dichloroquinoline (4) with sodium azide, yielding the azido precursor 5. Additionally, other azide-substituted 4-aminoquinoline derivatives (8a,b) were synthesized as described in previous reports [32]. This was achieved by heating ethanolamine/propanolamine with 4,7-dichloroquinoline in ethanol, followed by mesylation to introduce a mesyl group. The final step involved a subsequent azidation reaction in DMF to synthesize of the desired azide-substituted derivatives (Figure 2B).

The synthesis of the desired isatin-indolo[2,3-b]quinoxaline hybrids (9a-e and 10a-j) was carried out by employing precursors 3, 5 and 8 in a Cu-promoted azide-alkyne cycloaddition (CuAAC) reaction. The reaction was conducted in a microwave reactor using DMSO as a solvent, at 80°C for 10 min, as outlined in Figure 2C. DIPEA (N,N-diisopropylethylamine) was employed as a base to stabilize Cu(I) in solution, preventing its oxidation to Cu(II), which is less effective in catalyzing the click reaction. After the reaction, the final product was washed with an EDTA solution to ensure the complete removal of residual copper content.

The structural elucidation of the synthesized hybrids were accomplished through a comprehensive array of analytical and spectroscopic techniques. For instance, 10b, showed a peak at m/z 518.1746 and 520.1719 [M]⁺ and [M + 2]⁺ in its HRMS-ESI. Its ¹H NMR spectrum exhibited a distinctive singlet at δ 8.09, attributed to the triazole ring proton and a multiplet at δ 5.75, representing the methylene protons. The presence of a triplet at 4.3 (J = 6.9 Hz) indicated the alkyl chain linked to the quinoline core. The presence of a signal at δ 47.91, which corresponds to the methylene carbon attached to the indolo[2,3-b]quinoxaline core, in addition to the aliphatic carbons at δ 36.93 and 29.80 within the ¹³C NMR spectrum, provided additional confirmation of the assigned structure.

4. Discussion

4.1. In vitro anti-plasmodial activity

The antiplasmodial activities of the synthesized compounds were evaluated against both CQ-susceptible (3D7) and CQ-resistant (W2) P. falciparum strains. The results, as summarized in Table 1, suggested that the synthesized hybrids exhibited decent antiplasmodial activities, with IC₅₀ values ranging from 0.47 to 5.84 μM on the CQ-susceptible strain and 0.93 to 5.40 μM on the CQ-resistant strain.

Table 1.

In vitro antiplasmodial efficacy of synthesized hybrids against CQ-Sensitive (3D7) and CQ-Resistant (W2) P. falciparum strains.

Entry	n	R	mean IC50a ± SDb in μM		IC50e in μM	SIf		RIg
			3D7c	W2d		3D7c	W2d
9a	–	H	5.85 ± 0.515	5.27 ± 0.594	–	–	–	0.90
9b	–	F	2.95 ± 0.321	4.27 ± 0.384	–	–	–	1.44
9c	–	Morpholine	4.51 ± 0.310	4.60 ± 0.346	–	–	–	1.01
9d	–	Piperidine	4.43 ± 0.811	4.90 ± 0.382	–	–	–	1.10
9e	–	Diethylamine	5.46 ± 0.715	5.35 ± 0.279	–	–	–	0.98
10a	2	H	2.51 ± 0.468	3.55 ± 0.329	–	–	–	1.41
10b	3	H	0.91 ± 0.120	3.69 ± 0.354	210	230.52	56.96	4.04
10c	2	F	2.18 ± 0.118	5.40 ± 0.386	–	–	–	2.47
10d	3	F	0.54 ± 0.047	0.93 ± 0.163	300	552.49	321.89	1.71
10e	2	Morpholine	1.93 ± 0.379	4.09 ± 0.331	–	–	–	2.12
10f	3	Morpholine	0.47 ± 0.067	1.90 ± 0.692	320	685.47	168.17	4.07
10g	2	Piperidine	2.37 ± 0.146	2.53 ± 0.328	–	–	–	1.06
10 h	3	Piperidine	1.05 ± 0.085	2.64 ± 0.571	270	257.72	102.11	2.52
10i	2	Diethylamine	2.87 ± 0.384	2.65 ± 0.455	–	–	–	0.92
10j	3	Diethylamine	2.17 ± 0.276	2.53 ± 0.422	220	115.39	98.87	1.16
CQ			0.03 ± 0.004	0.58 ± 0.132	–	–	–	20.84
QN (quinine)			0.14 ± 0.036	0.44 ± 0.071	–	–	–	3.22
MQ (mefloquine)			0.07 ± 0.014	0.02 ± 0.004	–	–	–	0.29

^aMean inhibitory concentration of 6 independent experiments on strains of P. falciparum.

^bStandard deviation.

^cP. falciparum CQ-Susceptible strain.

^dP. falciparum CQ-Resistant strain.

^eCytotoxicity on Vero cells.

^fSelectivity index: SI = IC₅₀ (cytotoxicity)/IC₅₀ (antiplasmodial activity).

^gResistance index: RI = IC₅₀(W2)/IC₅₀(3D7).

In the structure-activity-relationship (SAR) investigations, compounds (9a-e) displayed no significant antiplasmodial activities when the triazole ring was directly attached to quinoline core. Their IC_50s ranged from 2.95 to 5.85 μM against the susceptible (3D7) and from 4.27 to 5.35 μM against the resistant (W2) strain. Notably, these hybrids showed similar potency against both strains, indicating that the introduction of fluorine and different secondary amines (morpholine, piperidine and diethylamine) did not result in substantial improvement in their effectiveness as anti-plasmodial.

In contrast, compounds featuring ethyl and propyl spacers, i.e., 10a-j, exhibited reasonable activity, underscoring the preference for longer linker lengths in fostering anti-plasmodial activity. Interestingly, these hybrid compounds demonstrated slightly better efficacy against the 3D7 strain when compared with the W2 strain. Notably, fluoro-substituted compound 10d, featuring a propyl chain, showed better activity, i.e., IC₅₀ = 0.54 μM and 0.93 μM than their unsubstituted counterpart 10b, with IC₅₀ value of 0.91 μM and 3.69 μM against susceptible and resistant strains, respectively.

The introduction of secondary amines showed no improvement in the activity profiles, except in the case of morpholine. Compound 10f, with a morpholine substituent, exhibited IC₅₀ = 0.47 μM and 1.90 μM on 3D7 and W2 strains, respectively. The protonated forms of the synthesized compounds is also analyzed for their partition coefficient values (ClogP) because it provides a better depiction of how the compounds might perform in vivo (Supplementary Figure S1). The ClogP values of CQ was calculated to be 4.15. Under physiological conditions, the protonated form of CQ exhibited a ClogP value of 3.01. The ClogP values of potent molecules in the series i.e., 10d and 10f, in their neutral forms were higher (5.10 and 4.37) compared with the protonated forms 10d²⁺ and 10f²⁺ (3.88 and 3.09). This change suggests that the protonated forms align better with Lipinski's Rule, particularly concerning ClogP values.

The synthesized compounds exhibited a resistance index (RI) ranging from 0.90 to 4.07, indicating minimal cross-resistance with CQ (RI = 20.84). Furthermore, the cytotoxicity assessment of the most promising hybrids on Vero kidney cell lines, detailed in Table 1, revealed that these compounds show no cytotoxicity against Vero cells. These active hybrids displayed a selective index (SI) ranging from 115.39 to 685.47 for the 3D7 strain and from 56.96 to 321.89 for the W2 strain. A graphical representation of the structure-activity-relationship (SAR) is provided in Figure 3.

Figure 3.

Graphical representation of structure-activity-relationship (SAR) of synthesized hybrids.

4.2. Mechanistic studies

4.2.1. Heme-inhibition study

Malaria parasites sustain their growth by breaking down hemoglobin, yielding a surplus of free heme (ferriprotoporphyrin-IX or Fp-IX). Fp-IX is toxic to the parasite, but the parasite employs a defense mechanism to transform free heme into a non-toxic crystalline substance known as hemozoin [33]. 4-aminoquinolines are believed to obstruct hemozoin formation by forming a complex with Fp-IX via π-π stacking, between the quinoline ring and the porphyrin ring of free heme [34]. The interaction between Fp-IX and CQ can be analyzed by examining the UV-Vis spectrum of aqueous hematin [35,36]. To delve into this key mechanism, we conducted heme-binding studies focusing on the most potent scaffold (10d). Titrations of 10d with free heme were performed in solutions of hemin (12 μM) containing 40% DMSO and water with pH conditions simulating both physiological (pH 7.4) and the acidic environment found in the digestive vacuole of parasites (pH 5.6). During these experiments, a spectral band at 401 nm corresponding to monomeric heme was initially observed. As 4 μM aliquots of 10d were added, a noticeable decrease in the absorption band's intensity was detected, indicating the binding of the drug to heme (Figure 4). Notably, new peaks corresponding to the drug began to appear prominently as the complexation between 10d and heme reached completion. For comparative purposes, a UV-Vis experiment under identical conditions was also conducted for 10d (refer to Supplementary Figure S2). Utilizing HypSpec, a nonlinear least-squares fitting software for data analysis, we quantified the binding affinity of 10d to heme, represented by logK values under both physiological and acidic pH conditions. As shown in Table 2, the binding affinity of 10d for heme was slightly stronger than that of CQ under both tested pH conditions.

Figure 4.

UV-absorption titration spectra of monomeric heme with an increasing concentration of 10d at: (A) pH 7.4 in HEPES buffer; and (B) pH 5.6 in MES buffer, in an aqueous DMSO solution, with the inset showing the linear relationship of absorption at 401 nm.

Table 2.

Calculated binding constant values (logK) for 10d and CQ at pH 5.6 and 7.4.

Entry	pH 5.6 (HEPES Buffer)	pH 7.4 (MES Buffer)
10d	5.98	6.07
CQ	5.10	5.18

4.2.2. Molecular docking study

We conducted molecular docking studies using PfCRT as the target enzyme to better understand the antiplasmodial effects of compounds, given its crucial role in expelling anti-malarial drugs like chloroquine from the parasite's digestive vacuole. As a central player in chloroquine resistance in P. falciparum, understanding the mechanisms of PfCRT's action is critical for unraveling drug resistance and developing innovative strategies against malaria [37]. Our research aims to uncover the intricate molecular processes behind malaria drug resistance, with the goal of identifying potential inhibitors capable of effectively blocking PfCRT's drug transport function. We obtained the crystal structures of PfCRT from the Protein Data Bank with PDB ID: 6UKJ. The computational analyzes were conducted using the Schrödinger molecular modeling suite (release 2022-1).

In comparison to CQ, which possesses a docking score of -3.8 [15], both 10d and 10f exhibit comparably higher docking scores, indicating stronger binding affinities with the target PfCRT protein (Table 3). Biological evaluations underscore the better potency of both 10d and 10f against the CQ-resistant W2 strain of P. falciparum when compared with other synthesized hybrids. Although the docking scores for 10d and 10f are quite similar, at -4.976 and -5.033, respectively, suggesting no substantial difference in their binding affinity. However, when it comes to in vitro antimalarial activity, 10d outperforms 10f due to its specific binding interactions with amino acid residues of the PfCRT protein, as shown in Figure 5A. In particular, 10d forms two important π-π stacking interactions, one with TYR 62 and another with TYR 384 amino acid residues. In contrast, 10f lacks such pronounced interactions within the target's active site (Figure 5B), although its morpholine component fits into a polar cavity in that site.

Table 3.

Calculated docking score and glide e-model energy for 10d, 10f and CQ.

Target	Ligand	IC50 (μM) against W2 strain	Docking score	Glide e-model
PfCRT	10d	0.93	-4.979	-74.659
	10f	1.90	-5.033	-83.309
	CQ	0.58	-3.771	-40.737

Figure 5.

3D representation of molecular docking simulation of (A) 10d, (B) 10f into the active site of PfCRT protein and (C) molecular dynamics simulation RMSD profiles illustrating the stability of the protein-ligand complex within the binding site.

4.2.3. Molecular dynamics simulations

Molecular dynamics (MD) simulations of the PfCRT-10d and PfCRT-10f complexes were performed to investigate the time evolution of ligand stability and protein conformational changes within the binding cavity. Root-mean-square deviation (RMSD) values were analyzed to assess the stability of these ligand-protein complexes over the 100 ns simulation period.

The RMSD plots indicate that both PfCRT-10d and PfCRT-10f induced minor conformational changes in the protein upon binding but ultimately maintained stable interactions (Figure 5C). For PfCRT-10d, the RMSD for ligand starts around 1.0 Å and shows an early increasing trend, peaking at approximately 3.5 Å between 10–20 ns. After this initial rise, PfCRT-10d exhibits some stabilization. During the 50–60 ns interval, the RMSD fluctuates between 1.8 and 3.5 Å. These RMSD values suggest minor conformational changes or flexibility within the binding site, without significant instability or displacement and PfCRT-10d remains stable up to 100 ns.

In contrast, PfCRT-10f exhibited significant instability, with RMSD fluctuations greater than 1.5 Å during the 10–70 ns interval and a sharp increase above 5 Å after 90 ns. This behavior suggests that PfCRT-10f is less stable than PfCRT-10d but slightly more stable than CQ²⁺ (Supplementary Figure S3) [38]. Despite the structural similarities between PfCRT-10d and PfCRT-10f, the differences in RMSD may be due to variations in their binding modes, interaction strengths with key residues, or subtle conformational changes in the protein.

Protein-ligand interaction fraction plots (Supplementary Figure S4A & B) further corroborate these observations. PfCRT-10d demonstrated strong and consistent interactions with key residues like SER 65, TYR 62 and ARG 392, maintaining hydrogen bonds and hydrophobic contacts for most of the simulation time. In contrast, PfCRT-10f exhibited a strong hydrophobic interaction with TYR 62 but weaker and more transient interactions otherwise. This behavior is similar to CQ²⁺, which struggles to maintain stable binding. Specifically, PfCRT-10d forms a hydrogen bond with SER 65 (68% of the time), π-π stacking with TYR 62 (59%) and Van der Waals interactions with PHE 340 (37%). PfCRT-10f, while still interacting with SER 388 (23%) and TYR 62 (45%), does not maintain the same level of stability in its contacts.

The molecular modeling results highlight the strong binding profiles and stability of both 10d and 10f in targeting PfCRT. 10d, in particular, shows better retention of key protein-ligand interactions, while 10f, despite being less stable, still exhibits promising binding characteristics. These findings suggest that both molecules hold significant potential as effective inhibitors of PfCRT.

4.2.4. Physicochemical properties & ADME prediction

Computational analysis of the ADME (Absorption, Distribution, Metabolism and Excretion) profile and drug-likeness for the hybrid compounds is summarized in Supplementary Table S1. Notably, none of the molecules fully comply with Lipinski's Rule of Five. Specifically, they exceed the MW criterion and possess logP values exceeding 5, indicating a need for further refinement to strike a balance between membrane permeability and solubility for optimal distribution. Additionally, they exhibit a higher number of hydrogen bond acceptors (HBA) than the recommended limit, which could raise concerns regarding their oral bioavailability.

Furthermore, the forecasted blood-brain barrier permeability (log BBB) values for 10b, 10d, 10f, 10h and 10j suggest limited capability to traverse the blood-brain barrier, with values ranging from -1.374 to -1.697. This implies potential challenges in reaching the central nervous system (CNS) for exerting therapeutic effects within the brain. Similarly, CNS permeability (log PS) values, ranging from -2.053 to -3.479, also indicate constrained penetration into the CNS. Interestingly, all compounds exhibited high predicted gastrointestinal absorption in this study.

5. Conclusion

The hybrid compounds synthesized in this study demonstrated reasonable potency against both CQ-susceptible and CQ-resistant strains of Plasmodium falciparum. In-depth structure-activity-relationship (SAR) investigations unveiled that the compounds with propyl linker structures displayed better activity than the rest of the series. Compound 10d with a fluorine substitution displayed an IC₅₀ of 0.93 μM against the CQ-resistant W2 strain. Upon replacing the fluorine with a morpholine group in 10f, its activity decreased, yielding an IC₅₀ of 1.90 μM against the same strain. However, this trend reversed when assessing their performance against the CQ-susceptible 3D7 strain, where 10d and 10f displayed IC₅₀ values of 0.54 μM and 0.46 μM, respectively. Molecular docking studies suggested that 10d and 10f have strong binding affinities for the target PfCRT protein, with 10d demonstrating specific interactions leading to its superior in vitro antimalarial activity. Molecular dynamics simulations further confirmed the stability and interaction retention of both PfCRT-10d and PfCRT-10f complexes, suggesting their significant potential as effective PfCRT inhibitors.

With the emergence of chloroquine-resistant and multidrug-resistant strains, these hybrids show potential for development as antimalarial agents. Future research will focus not only on optimizing the structure of these hybrids but also on designing new analogues with varied mechanisms of action. Specifically, targeting both PfCRT and heme pathways offers a dual approach to combat resistance and enhance antimalarial efficacy. Additionally, the synthetic accessibility of the indolo[2,3-b]quinoxaline and 4-aminoquinoline makes it possible to modify them with various substituents, enhancing their properties and efficacy.

Funding Statement

UGC (NFSC) has been acknowledged for providing fellowship to S Chowdhary (Ref No. 1594/CSIR-UGC NET DEC 2018). V Kumar thanks Council of Scientific and Industrial Research (CSIR) for providing financial support (grant no. 02(0400)/21/EMR-II). B Pradines thanks the French Institute for Public Health Surveillance (Santé Publique France) (grant CNR Paludisme) and the Direction Générale de l'Armement (grant no NBC-2-B-2120) for financial support.

Supplemental material

Supplemental data for this article can be accessed at https://doi.org/10.1080/17568919.2024.2419354

Author contributions

S Chowdhary: synthesized, conducted docking study and analyzed the data; S Arora: synthesized the compounds; A Anand: analyzed computational data; I Fonta and J Mosnier: conducted the biological studies; B Pradines; analyzed the biological data and finalized the draft V Kumar: conceptualized the work and finalized the draft. All authors have given approval to the final version of the manuscript.

Financial disclosure

UGC (NFSC) has been acknowledged for providing fellowship to S Chowdhary (Ref No. 1594/CSIR-UGC NET DEC 2018). V Kumar thanks Council of Scientific and Industrial Research (CSIR) for providing financial support (grant no. 02(0400)/21/EMR-II). B Pradines thanks the French Institute for Public Health Surveillance (Santé Publique France) (grant CNR Paludisme) and the Direction Générale de l'Armement (grant no NBC-2-B-2120) for financial support.

Competing interests disclosure

The authors have no relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript. This includes employment, consultancies, honoraria, stock ownership or options, expert testimony, grants or patents received or pending, or royalties.