Synthesis and in vitro antiprotozoal evaluation of novel Knoevenagel hydroxychloroquine derivatives

Abstract

Leishmaniasis and Chagas diseases affect millions of people, particularly in developing countries, with conventional treatments proving unsatisfactory due to increasing drug resistance and high toxicity. Therefore, there is an urgent need for new drugs to combat neglected tropical diseases (NTDs). In this study, we synthesized 15 new Knoevenagel adducts derived from hydroxychloroquine and evaluated their antiprotozoal activity against Leishmania infantum, L. amazonensis, and Trypanosoma cruzi. The new adducts exhibited low toxicity in epithelial LLC-MK2 cells and J774A.1 macrophages. The Knoevenagel adducts derived from meta- and para-chloro benzaldehyde demonstrated antiprotozoal activity against T. cruzi epimastigotes, though with a lower selective index (SI) compared to the standard drug benznidazole. However, the adducts derived from isovaleraldehyde and ortho-, meta-, and para-chloro benzaldehyde showed SI values ranging from 10.97 to 8.11 against L. amazonensis, similar to amphotericin B (AmpB, SI = 9.37), with no statistically significant difference (p > 0.05). These same compounds inhibited L. infantum promastigotes, but with less activity compared to AmpB. These results suggest that Knoevenagel adducts derived from hydroxychloroquine may serve as selective antileishmanial agents.

Antiprotozoal evaluation of novel Knoevenagel hydroxychloroquine derivatives.

Introduction

Neglected tropical diseases (NTDs) affect underdeveloped countries, exacerbating social inequalities, with over 1 billion people suffering from them. Among the 20 diseases¹ classified as NTDs, leishmaniasis and Chagas disease are included,^1–7 both caused by parasites^2,3,7,8 from the Trypanosomatidae family.⁷ Leishmaniasis is classified into visceral, mucosal, and cutaneous forms,^7–9 with transmission occurring through sandflies.^3,7,8 In the case of Chagas disease, the Triatoma insect transmits the Trypanosoma cruzi parasite. The disease is predominantly found in rural areas of South America and is considered endemic to the region.^7,8

Currently, the treatment for leishmaniasis includes pentavalent antimonials, amphotericin B (AmpB), miltefosine, and paromomycin,^{4,7,8,10–12} depending on the availability of resources in the region.⁸ Benznidazole and nifurtimox have been used to treat Chagas disease.^3,4,7,8,13 However, the effectiveness of these drugs for Chagas disease is not guaranteed due to the complexities associated with the different stages of the infection.⁸ Additionally, the drugs prescribed for both diseases often exhibit unsatisfactory efficacy and high toxicity, highlighting the urgent need for the development of new therapeutic structures with improved activity.^7,8,11,14

Natural and synthetic quinoline molecules have been explored as potential antiprotozoal agents for the treatment of leishmaniasis and Chagas disease.⁷ Additionally, promising results have been observed in the evaluation of antimalarial commercial drugs, which contain the quinoline core, against these diseases.^15,16

The chloroquine derivative 1 demonstrated antiprotozoal activity in vivo against Leishmania panamensis.¹⁴ Additionally, 7-chloroaminoquinoline 2, also known as GF1061, exhibited superior antiprotozoal activity compared to AmpB against promastigotes and amastigotes of L. infantum and L. amazonensis (Fig. 1).¹⁷ Furthermore, derivatives of 7-chloroaminoquinoline showed activity against both leishmaniasis^{2,6,10–12,14,17–21} and T. cruzi.^2,5,13,22 Hybrid aminoquinoline-chalcone compounds also displayed promising in vitro activity against malaria²³ and leishmaniasis.²⁰

Fig. 1. Quinoline derivatives used against Leishmania sp. and malaria.

Aminoquinolines are a well-known class of compounds with versatile applications in the treatment of various diseases. Commercially available quinoline drugs, such as primaquine (3) and hydroxychloroquine (4) (Fig. 1), have been widely used in the treatment of malaria and other infectious diseases caused by Plasmodium parasites.²⁴ Additionally, modifications of primaquine have been reported as an effective strategy for developing novel antiplasmodial agents.²⁵

Although hydroxychloroquine (HCQ, 4) is primarily classified as an antimalarial, it is also prescribed for a variety of other conditions, including rheumatic and skin diseases. These include systemic lupus erythematosus, palindromic rheumatism, rheumatoid arthritis, eosinophilic fasciitis, dermatomyositis, Sjögren's syndrome, porphyria cutanea tarda, polymorphous light eruption, granuloma annulare, lichen planus, lupus panniculitis, discoid lupus, and other immunological and infectious diseases.^24,26

Indeed, HCQ (4) gained significant attention during the COVID-19 pandemic due to initial in vitro studies that suggested it might have potential antiviral properties. The studies indicated that HCQ could limit the entry of SARS-CoV-2 into human cells by inhibiting the glycosylation of viral receptors on the cell surface, specifically the ACE2 receptor, which is the primary target for the virus. This led to early optimism and wide public endorsements, including from political figures, who suggested that HCQ could play a role in treating COVID-19.

However, despite the initial interest, clinical trials and randomized controlled studies failed to confirm HCQ's efficacy in treating COVID-19 patients. The FDA (U.S. Food and Drug Administration) reviewed the available evidence and, after rigorous evaluation, determined that hydroxychloroquine did not provide significant benefits for COVID-19 treatment.²⁷ As a result, the FDA revoked its emergency use authorization for HCQ in COVID-19 treatment.

Recently, hydroxychloroquine is currently being repurposed for cancer treatment. Preclinical studies have shown its potential in anti-cancer therapy, particularly when combined with conventional cancer treatments, as it sensitizes tumor cells to drugs, thereby enhancing therapeutic activity.²⁸

As mentioned above, given the repurposing strategy and the synthetic modifications made to produce many biologically active aminoquinolines for the development of new medicines, we hypothesize that structural modifications to the terminal hydroxyl group of hydroxychloroquine (4) could result in promising biological activity against leishmaniasis and Chagas disease.

Herein, we synthesized 15 new Knoevenagel adducts of hydroxychloroquine and evaluated their antiprotozoal activity against L. infantum and L. amazonensis promastigotes, which cause visceral and cutaneous leishmaniasis, respectively, as well as T. cruzi epimastigotes, the causative agents of Chagas disease (Fig. 2).

Fig. 2. Modification of the terminal hydroxy group of hydroxychloroquine to yield new alkylidene derivatives.

Results and discussion

Synthesis of Knoevenagel adducts

The synthesis of Knoevenagel adducts of hydroxychloroquine was carried out as outlined in Scheme 1. First, hydroxychloroquine sulfate (4·H₂SO₄) was desulfated through an acid–base reaction²⁹ (Scheme 1, step i), yielding 97% of free hydroxychloroquine (4, HCQ). The ¹H and ¹³C NMR analyses were consistent with the literature.³⁰

Scheme 1. Synthesis of Knoevenagel alkylidenes 7a–o of hydroxychloroquine.

Next, the synthesis of the key intermediate hydroxychloroquine acetoacetate (5, HCQAA) was investigated. The transesterification reaction was carried out by reacting 4 with methyl acetoacetate in methanol, using sulfamic acid (NH₂SO₃H) as a catalyst (30% mol).³¹ The progress of the reaction was monitored by thin-layer chromatography (TLC).

However, after 22 hours under reflux, only a trace of the corresponding 1,3-dicarbonyl compound 5 was detected by ¹H NMR analysis. The same protocol was also tested using iodine as a catalyst (30% mol), but again, only traces of the product were observed.³²

These results prompted us to investigate the use of 2,2,6-trimethyl-4H-1,3-dioxin-4-one (TMD).³³ TMD is a direct precursor to 1,3-dicarbonyl compounds, and when 4 and TMD were mixed with anhydrous sodium acetate in THF for 24 hours at 85 °C, compound 5 was obtained in a 77% yield (Scheme 1, step ii) after purification by column chromatography on silica gel.³⁴

The ¹H NMR spectrum of 5 showed two characteristics singlets (δ 2.23 and 3.43 ppm) to hydrogens alpha bis-carbonylic and methyl group, respectively, present in the 1,3-dicarbonilic system. In addition, the spectrum showed the O–CH₂ group in δ 4.20 (t, J = 6.0 Hz). This signal indicates the presence of the β-ketoester derivative formed from hydroxychloroquine (4) upon modification. In the ¹³C NMR spectrum of hydroxychloroquine acetoacetate (5), the presence of two distinct carbonyl carbon signals further supports the structure of the 1,3-dicarbonyl system (Fig. S3†).

The synthesis of the alkylidenes 7a–o was successfully achieved via the Knoevenagel condensation between hydroxychloroquine acetoacetate (5) and various aldehydes (Scheme 1, step iii).^35–37 The reaction was catalyzed by piperidinium acetate ([Pip][CH₃COO]), and the products 7a–o were obtained with yields ranging from 47% to 89%. The synthesized compounds were characterized by various analytical techniques, including ¹H NMR, ¹³C NMR, IR, and high-resolution mass spectrometry (HRMS).

The ¹H and ¹³C NMR spectra of the synthesized 7a–o compounds revealed signals corresponding to the presence of a mixture of E/Z isomers in all cases. Given the spectral complexity, the E/Z isomerism was further confirmed using HSQC (Heteronuclear Single Quantum Coherence) experiments.

The HSQC experiment provided valuable information by showing heteronuclear correlations between the olefinic hydrogen and the respective carbon of the alkylidene and benzylidene moieties, confirming the E/Z isomerism and providing further structural insight into the alkylidene and benzylidene components of the molecules.

Afterward, a study using the Heteronuclear Single Quantum Multiple Bond Correlation (HSQMBC) pulse sequence was developed to quantify the E/Z isomers of the Knoevenagel adducts, following the procedure extensively described by Saurí et al.³⁸

The ¹H–¹³C HMBC NMR spectrum was initially acquired to identify the olefinic hydrogens, as these are the only protons exhibiting long-range correlations with both carbonyl carbons in the structure. Then, the power and duration of the selective ¹H NMR pulse at the frequency of the olefinic hydrogen were calibrated and used to obtain the dsel-HSQMBC-IPAP spectra (Fig. 3). This allowed for precise measurement of the coupling constants (J) between the olefinic hydrogen and the ester and ketone carbons in the stereoisomers. The coupling constants provided important information about the spatial arrangement of the atoms in the E/Z isomeric forms, further confirming their stereochemistry and contributing to a comprehensive structural understanding of the Knoevenagel alkylidenes.

Fig. 3. dsel-HSQMBC-IPAP spectra for the mixture of isomers (E/Z)-7i.

According to the HSQMBC analysis, the E/Z isomer ratio of the Knoevenagel adducts remained consistent across all compounds, with the Z isomers predominating in the major compounds, exhibiting ratios ranging from 35 : 65 to 40 : 60 (Fig. S67–S81†). This provides insight into the isomer distribution within the synthesized compounds. The presence of Z isomers was confirmed through the correlation of the olefinic hydrogen, which exhibited a higher coupling constant with the ester carbonyl carbon, indicating an antiperiplanar relationship. The high coupling constant with the ketone's carbonyl carbon further confirmed the presence of the E isomer.

Biological tests

The antiparasitic activity of Knoevenagel adducts 7a–o, HCQ·H₂SO₄ (4·H₂SO₄), HCQ (4), and HCQAA (5) against T. cruzi, L. infantum, and L. amazonensis was evaluated, and the results were compiled in Tables 1 and 2.

Table 1. In vitro antiprotozoal activity against epimastigotes of T. cruzi (IC₅₀, μM) and cytotoxicity (CC₅₀, μM) on LLC-MK₂ cells.

Comp.	IC50a (μM)	CC50a (μM)	SIb (CC50/IC50)
	Epimastigotes	LLC-MK2 cells	T. cruzi
	T. cruzi
4·H2SO4	>100	137.69 ± 14.32	c
4, HCQ	>100	164.33 ± 16.05	c
5, HCQAA	>100	221.51 ± 18.61	c
7a	40.72 ± 1.64	256.21 ± 19.02	6.29
7b	60.01 ± 15.05	230.12 ± 24.78	3.84
7c	>100	199.89 ± 11.90	c
7d	>100	242.61 ± 24.54	c
7e	>100	235.46 ± 30.30	c
7f	>100	209.64 ± 14.91	c
7g	46.54 ± 1.28	176.35 ± 14.28	3.79
7h	37.34 ± 1.95	252.44 ± 38.79	6.76
7i	44.33 ± 2.19	249.09 ± 34.77	5.62
7j	>100	91.19 ± 2.67	c
7k	73.50 ± 8.03	190.86 ± 13.09	2.60
7l	40.26 ± 1.61	195.28 ± 29.12	4.85
7m	64.36 ± 2.66	118.93 ± 13.46	1.85
7n	>100	73.97 ± 6.71	c
7o	73.00 ± 12.93	221.80 ± 30.02	3.04
Benznidazole	11.84 ± 1.04	504.68 ± 25.71	42.63

^aIC₅₀ and CC₅₀ values were calculated based on the average from triplicate of each point.

^bSI: selectivity index (CC₅₀/IC₅₀).

^cDoes not show antiparasitic activity against the tested strain.

Table 2. In vitro antiprotozoal activity (IC₅₀, μM) in L. infantum and L. amazonensis, cytotoxicity (CC₅₀, μM) on J774A.1 macrophages, and selectivity index (CC₅₀/IC₅₀).

Comp.	IC50a (μM)	IC50a (μM)	CC50a (μM)	SIb (CC50/IC50)	SIb (CC50/IC50)	clog P
	Promastigotes	Promastigotes	MΦJ774A.1	L. infantum	L. amazonensis
	L. infantum	L. amazonensis
4·H2SO4	>100	>100	348.88 ± 63.87	d	d	—
4, HCQ	>100	>100	409.69 ± 24.19	d	d	4.12
5, HCQAA	>100	>100	417.37 ± 30.95	d	d	4.63
7a	36.11 ± 5.28	33.34 ± 1.15	355.22 ± 47.79	9.83	10.65	6.79
7b	78.38 ± 1.60	72.24 ± 6.40	399.41 ± 55.36	5.10	5.53c	6.64
7c	92.82 ± 12.05	>100	396.16 ± 29.49	4.27	d	6.81
7d	93.31 ± 1.90	>100	401.80 ± 28.58	4.31	d	6.56
7e	>100	>100	191.04 ± 30.49	d	d	6.30
7f	37.01 ± 0.27	>100	206.28 ± 20.53	5.57	d	6.79
7g	40.34 ± 5.13	56.42 ± 4.02	419.97 ± 26.88	10.41	8.17c	7.36
7h	28.67 ± 2.63	23.63 ± 1.60	259.43 ± 67.82	9.04	10.97c	7.36
7i	38.25 ± 5.89	36.65 ± 5.08	297.36 ± 36.29	7.77	8.11c	7.36
7j	>100	44.11 ± 3.73	254.29 ± 51.50	d	5.77	6.39
7k	81.07 ± 5.65	26.50 ± 1.31	<10	d	d	6.39
7l	74.50 ± 5.84	65.34 ± 11.72	<10	d	d	6.39
7m	50.11 ± 7.21	38.76 ± 5.30	230.06 ± 12.58	4.59	5.94c	7.25
7n	>100	>100	81.60 ± 7.69	d	d	5.82
7o	>100	>100	238.45 ± 10.58	d	d	6.29
AmpB	1.05 ± 0.11	9.89 ± 4.09	92.70 ± 10.27	88.29	9.37	—

^aIC₅₀ and CC₅₀ values were calculated based on the average from triplicate of each point.

^bSI: selectivity index (CC₅₀/IC₅₀).

^cTreatments compared with the AmpB (amphotericin B) standard using the one-tailed t-test.

^dDoes not show antiparasitic activity against the tested strain.

The cytotoxicity evaluation of compounds 4·H₂SO₄, 4, 5, and the Knoevenagel adducts 7a–o was performed against epithelial LLC-MK2 cells and J774A.1 macrophages using the MTT assay (Tables 1 and 2).

The selectivity index (SI) is an important measure of the potential therapeutic efficacy of a compound, particularly in evaluating its antiparasitic activity while minimizing toxicity to host cells. The SI is determined by the ratio of the 50% cytotoxic concentration (CC₅₀) to the 50% inhibitory concentration (IC₅₀) for both the epimastigote (T. cruzi) and promastigote (L. infantum and L. amazonensis) forms.

The data from your study indicating that compounds 4·H₂SO₄, 4 and 5 did not inhibit the growth of the parasites, even at a high concentration of 100 μM, suggests that these precursors are ineffective against T. cruzi, L. infantum, and L. amazonensis in the context of your experiment (Tables 1 and 2). Similar results have been reported in the literature for hydroxychloroquine and chloroquine, two well-known antimalarial drugs. These studies indicated that both hydroxychloroquine and chloroquine were not effective in inhibiting the promastigote form of L. amazonensis within 72 hours at doses as high as 50 μM.¹⁵ The results (Table 1) showed that the Knoevenagel adducts 7a–b (IC₅₀ = 40.72 and 60.01 μM), 7g–i (IC₅₀ = 37.34–46.54 μM), 7k–m (IC₅₀ = 40.26–73.50 μM), and 7o (IC₅₀ = 73.00 μM), derived from HCQ, were active against epimastigotes of T. cruzi, with a SI > 1. On the other hand, other Knoevenagel adducts did not show significant activity against the epimastigotes, even at the maximum concentration of 100 μM.

Among the derivatives, the adducts 7a, 7h, and 7i, which were synthesized from isovaleraldehyde, meta-chloro, and para-chloro benzaldehyde, respectively, exhibited particularly promising antiprotozoal activity against epimastigotes. These compounds demonstrated SI > 5, with values of 6.29, 6.76, and 5.62, respectively, indicating that they were not only effective against the parasite but also had a relatively low cytotoxicity toward host cells (Table 1). However, compared to the standard drug benznidazole (SI = 42.63), the SI values for these molecules are approximately 7 times lower. Although the literature has reported that thiosemicarbazones with a chlorine group in the aromatic ring exhibit activity against T. cruzi,³⁹ in our study, the Knoevenagel adducts showed lower activity than the standard benznidazole.

The Knoevenagel adducts 7a–o were also submitted to antiparasitic assay against L. infantum and L. amazonensis, and the evaluation of cytotoxicity against macrophages in J774A.1 (Table 2). According to Table 2, most alkylidenes were active against both Leishmania species. In addition, the cytotoxicity of most was lower (range CC₅₀ = 191.04–419.57 μM) and only compounds 7n (CC₅₀ = 81.60 ± 7.69 μM), derived from furfuraldehyde, 7k and 7l (CC₅₀ < 10), derived from nitrobenzaldehyde, were more cytotoxic than AmpB (CC₅₀ = 92.7 μM). Among the Knoevenagel adducts investigated in vitro, the alkylidenes 7a, 7b, 7g, 7h, 7i and 7m derived from isovaraldehyde, benzaldehyde, ortho-chloro, meta-chloro and para-chloro benzaldehyde, and cinnamaldehyde, respectively, provided interesting activity in both Leishmania species. The compounds 7a, 7g, 7h, 7i, 7m showed IC₅₀ ≤ 56.42 ± 4.02 μM, while 7b had a IC₅₀ = 78.38 ± 1.60 μM and 72.24 ± 6.40 μM against L. infantum and L. amazonensis, respectively. The results showed that the inhibition concentrations obtained to adducts 7a, 7g–i, 7m were like those observed by Antinarelli et al.¹⁸ for the hybrid 7-chloro-4-quinolinylhydrazone derived from para-hydroxy benzaldehyde, exhibiting antiparasitic activity in L. amazonensis against promastigotes (IC₅₀ = 52.5 μM) and amastigotes (IC₅₀ = 8.1 μM).

In addition, the cytotoxicity indexes were low (CC₅₀ ≥ 230.06 ± 12.58 μM) in J774A.1 macrophage resulting in a SI (CC₅₀/IC₅₀) between 4.27 to 10.41 for L. infantum and 5.53 to 10.97 for L. amazonensis.

According to the SI (Table 2), the alkylidenes with higher biological activity against L. amazonensis where 7a, 7g, 7h, and 7i, derived from isovaraldehyde, ortho-, meta- and para-chloro benzaldehydes, respectively (Fig. 4), showing the SI range between 8.11 and 10.97, similar to the AmpB (SI = 9.37). The Student t-test was applied to compare those SI values of the compounds with the of AmpB. The results indicated no statistical differences (p > 0.05) between each Knoevenagel adduct mentioned and the AmpB.

Fig. 4. Compounds that demonstrated higher biological activity and low cytotoxicity against L. amazonensis promastigotes.

However, despite the low cytotoxicity and higher biological activity of compounds 7a, 7g, 7h and 7i against L. amazonensis, their selectivity index (SI) was still nine times lower than that of AmpB (SI = 88.29) against L. infantum, with SI values ranging between 7.77 and 10.41 (Table 2). These results indicate that the same Knoevenagel adducts inhibited promastigotes of both L. infantum and L. amazonensis, but the compounds exhibited greater activity against L. amazonensis. The dose–response curves and the determination of the half-maximal inhibitory concentration (IC₅₀, μM) for L. infantum and L. amazonensis promastigotes, as well as cytotoxicity (CC₅₀, μM) on J774A.1 macrophages for alkylidenes 7a, 7g, 7h, are shown in Fig. 5.

Fig. 5. Graphical dose–response analysis and determination of IC₅₀ (μM) against L. infantum and L. amazonensis promastigotes, along with CC₅₀ (μM) on J774A.1 macrophages.

Considering that molecules 7a and 7g–i exhibited SI values ranging from 8.11 to 10.97 and demonstrated the lowest IC₅₀ values for L. amazonensis, their cytotoxicity was further assessed in L-929 fibroblasts. The SI values of these compounds were found to range between 5.13 and 6.02, which is comparable to AmpB (SI = 7.07) when evaluated for cytotoxicity in L-929 fibroblasts. In the case of L. amazonensis, the SI values of the respective structures ranged between 5.35 and 7.23, which is 7 to 9 times higher than the standard drug AmpB (SI = 0.75). Similar to the findings in J774A.1 macrophages, the compounds showed greater selectivity against the species responsible for cutaneous leishmaniasis, reinforcing their potential for further investigation as promising antiprotozoal agents.

Our results demonstrated that the selectivity index of the novel Knoevenagel adducts was similar to that of other quinoline derivatives described in the literature for Leishmania and Chagas disease. Coa and coworkers³ synthesized quinoline–hydrazone hybrids and evaluated their antiparasitic activity against L. (V) panamensis and T. cruzi, where the SI values ranged from 4.7 to 2.6 and 11.6 to 0.5, respectively. Although the quinoline–hydrazone compounds exhibited moderate SI values, these molecules were considered potential candidates for further disease-related studies due to their favorable EC₅₀ values.

As shown in Table 2, the three structures containing a chlorine atom in the ortho-, meta-, and para-positions of the aromatic ring – benzylidenes 7g, 7h, and 7i – demonstrated promising results against L. amazonensis and L. infantum. These findings suggest that the presence of the chlorine group in these positions of the ring plays a crucial role in inhibiting the activity of the Leishmania parasite.⁴⁰

Luczywo and coworkers⁴¹ also observed similar behavior with 2-styrylquinoline-4-carboxylic acid while investigating the antiparasitic activity of both the promastigote and amastigote forms of L. amazonensis. The best results were observed when the 2-styryl portion contained a chloro group at the ortho or para position, as indicated by their SAR study. Generally, substitutions on the phenyl moiety with electron-donating groups (EDGs), such as chloro, can enhance the activity against Leishmania.¹⁰ According Table 2 the Knoevenagel adducts 7c–e, 7j–l, and 7n–o demonstrated low cytotoxicity, as indicated by the CC₅₀ values in J774A.1 macrophages. However, these compounds demonstrated low selectivity (SI < 4.5) or were unselective for Leishmania promastigotes. Additionally, some Knoevenagel adducts showed isolated antiparasitic activity against Leishmania species. For L. infantum compound 7f exhibited an SI of 5.57 with IC₅₀ 37.01 ± 0.27 while for L. amazonensis, compound 7j showed an SI of 5.77 and an IC₅₀ 44.11 ± 3.73.

Additionally, the partition coefficient (clog P) values of the alkylidenes 7a–o were calculated to estimate their lipophilicity (Table 2). The results show that compounds 7a–o exhibited greater lipophilic character, with clog P values ranging from 5.82 to 7.36, compared to their precursors (4, clog P = 4.12, and 5, clog P = 4.63). The Knoevenagel adducts containing a chloro group on the phenyl ring (7g, 7h, and 7i), which also exhibited the best SI for L. amazonensis, demonstrated the highest lipophilicity within this group (clog P = 7.36). The adduct 7a, with a lipophilic isopropyl chain (clog P = 6.79), also showed promising SI values, similar to those of the chlorine derivatives. These findings suggest that the more lipophilic molecules 7a, 7g, 7h, and 7i may have enhanced permeability across the plasma and mitochondrial membranes of the protozoan in the promastigote form. Increased lipophilicity logically leads to enhanced transport across biological barriers, although the activity itself may not be solely due to lipophilicity. However, further studies are required to verify the mechanism of action.

Experimental

Apparatus

The hydroxychloroquine sulfate (purity ≥99%) was supplied by Dideu Medichem (China) and 2,2,6-trimethyl-4H-1,3-dioxin-4-one (TMD) purchased from Aldrich Chemical Co. and aromatic aldehydes from Alfa Aesar were used without further purification.

All organic solvents used for synthesis were of analytical grade. Solvents were dried and freshly distilled. Column chromatography was performed with silica gel (0.063–0.2 mm/70–230 mesh ASTM, Macherey-Nagel™). Reactions were monitored using thin-layer chromatography (TLC) on silica gel plates (Merck 60GF245) with CH₂Cl₂ : MeOH as eluent. Fourier-transform infrared spectroscopy (FT-IR) spectrums were acquired using Bruker model Vertex 70 spectrometer (Massachusetts, USA) in the transmittance mode (range between 4000–400 cm⁻¹, 16 scans), through a KBr cell. NMR spectra of the ¹H, ¹³C, HSQC and HMBCQ NMR were recorded on a Bruker 400 MHz AVANCE III NMR (¹H at 400 MHz and ¹³C at 100 MHz) and Bruker 600 MHz ASCEND (¹H at 600 MHz and ¹³C at 150 MHz) spectrometers. Each sample was diluted in 0.6 μL of deuterated chloroform (CDCl₃) using as internal standard tetramethylsilane (TMS). The acquired spectrums were processed and interpreted using the software TopSpin version 4.1.1. High-resolution mass spectrometry (HRMS) experiments were performed on a MicroTOF-Q II hybrid quadrupole time-of-flight mass spectrometer (Bruker Daltonics®, Fremont, CA, USA) equipped with an electrospray ion source in positive ion mode. A Harvard Model 22 Dual syringe pump (Harvard Apparatus, South Natick, USA) with a flow rate of 180 μL h⁻¹ was used to infuse the samples directly into the mass spectrometer. High-purity nitrogen produced using a nitrogen generator (PEAK Scientific Instruments, Chicago, USA) was used as nebulizer gas at 4.0 bar and as drying gas with a flow of 6.0 L min⁻¹ at 180 °C. The capillary voltage was set at 4.5 kV. Direct infusion mass spectrometry acquisition parameters were set at 200 to 900 m/z mass range. Data was acquired using an MS Workstation with oToFControl software (Bruker Daltonics®, Fremont, CA, USA). The samples diluent consisted of methanol with 0.1% formic acid.

Synthetic procedures

Obtaining of free hydroxychloroquine (4) (2-((4-((7-chloroquinolin-4-yl)amino)pentyl)(ethyl)amino)ethan-1-ol)²⁹

Hydroxychloroquine sulfate (4·H₂SO₄, 23 mmol, 10.0 g) and 10 mL of methanol were added in a round-bottom under magnetic stirring at 25 °C until the solubilization. Next, an aqueous solution of 50 mL of.

NaOH (69 mmol, 2.76 g) was added and the system was stirring at 25 °C for 1 h. Afterward, the mixture was extracted with CHCl₃ (3 × 100 mL) and the combined organic phases were dried over MgSO₄, filtered, and evaporated under reduced pressure to provide the free hydroxychloroquine (4). The compound was dried in a high vacuum pump and stored in the freezer at −20 °C.

Synthesis of hydroxychloroquine acetoacetate 5 (2-((4-((7-cloroquinolin-4-il)amino)pentil)(etil)amino)etil 3-oxobutanoato)

The free hydroxychloroquine (4) (2 mmol, 0.6758 g) was solubilized in 2 mL of anhydrous THF in a Falcon tube. Then, 2,2,6-trimethyl-4H-1,3-dioxin-4-one (TMD, 2.6 mmol, 0.345 mL) and anhydrous sodium acetate (2 mmol, 0.1641 g) were added, the reaction was kept under stirring for 24 h at 85 °C. Afterwards, was added 40 mL of CH₂Cl₂ and the mixture was extracted with H₂O (3 × 20 mL). The organic phase was dried over MgSO₄, filtered, and evaporated under reduced pressure. The crude mixture was purified through column chromatography with silica gel (0.063–0.2 mm/70–230 mesh ASTM, Macherey-Nagel™) using a gradient mixture of CH₂Cl₂ until CH₂Cl₂ : MeOH, 97 : 3 ratio, to obtain the pure hydroxychloroquine acetoacetate 5. The compound was dried in a high vacuum pump and stored in the freezer at −20 °C.

General procedure for the synthesis of hydroxychloroquine Knoevenagel adducts 7a–o³⁵

Hydroxychloroquine acetoacetate 5 (0.5 mmol, 0.2105 g) was solubilized in 1.5 mL of anhydrous CH₂Cl₂ in a Falcon tube. The piperidine acetate (0.05 mmol, 0.0073 g) was added, and the system was kept under magnetic stirred for 5 minutes at 20 °C. Then, the aldehydes 6a–o (0.65 mmol) was added, stirring for more 45 minutes at the same temperature. After that time, the previously activated 4Å molecular sieve was placed in the system, remaining under reflux at 50 °C for 21 h. The reaction was monitored through TLC (thin-layer chromatography) using silica plates with eluent CH₂Cl₂ : MeOH 5%, revealed in an alcoholic solution of acidic vanillin under hot air. After, the reaction was solubilized in 20 mL of CH₂Cl₂ and washed with distilled H₂O (2 × 10 mL). The organic phase was recovered, dried in MgSO₄, filtered, and evaporated under reduced pressure. The crude was purified through a chromatographic column using silica (0.063–0.2 mm/70–230 mesh ASTM, Macherey-Nagel™) as the stationary phase and eluent with gradient elution (CH₂Cl₂ until CH₂Cl₂ : MeOH, 97 : 3 ratio). The products 7a–o were dried in a high vacuum pump and stored in a freezer at −20 °C.

Spectroscopic analyses

Experiment for identification and quantification of E/Z isomers

The compounds were characterized according to Saurí, Nolis, Parella³⁸ on a 9.4 Tesla Bruker AVANCE III HD spectrometer operating at 400 MHz at the core frequency of ¹H. The characterization consisted of obtaining and assigning the one-dimensional spectra of ¹H and ¹³C{¹H} and two-dimensional spectra of HSQC and HMBC. The double bond configuration was determined by measuring the ³JCH couplings of the olefinic hydrogen using the sel-HSQMBC-IPAP NMR pulse sequence. The spatial relationship between olefinic hydrogen and carbonyl carbons was determined using the ³JCH couplings, which agree with the Karplus relationship. As the structure has two carbonyl carbons three bonds away from the olefinic hydrogen, it is possible to use the Karplus relation to attribute the spatial relation between the H and C atoms involved in the ³JCH coupling and unequivocally assign the configuration of the double bond. According to the Karplus relation, the higher value of ³JCH coupling occurs between atoms in antiperiplanar relation, while the lower value of ³JCH coupling occurs between atoms in synperiplanar relation.

Biological assays

Parasites

Epimastigotes of T. cruzi (strain Y) were cultured in liver infusion tryptose (LIT) medium at pH 7.4, supplemented with 10% fetal bovine serum (FBS), inactivated at 56 °C, 5000 U mL⁻¹ penicillin, and 5 mg mL⁻¹ streptomycin. Parasites were maintained in a biochemical oxygen demand (B.O.D.) incubator at 28 °C and utilized during the stationary phase of growth. Promastigotes of L. amazonensis (WHOM/BR/75/Josefa) and L. infantum MON 1 (MCAN/BR/97/P142) were cultivated in Warren medium supplemented with 10% FBS, 5000 U mL⁻¹ penicillin, and 5 mg mL⁻¹ streptomycin. Cultures were maintained in a B.O.D. incubator at 25 °C, with passages conducted every 48 hours.

LLC-MK₂ cells and J774A.1 macrophages

LLC-MK₂ cells (Cell Bank Rio de Janeiro, RJ, 25250-020, Brazil) were maintained in an incubator at 37 °C with a 5% CO₂ atmosphere in Dulbecco's modified Eagle's medium (DMEM), supplemented with 10% FBS, 5000 U mL⁻¹ penicillin, and 5 mg mL⁻¹ streptomycin. As for J774A.1 macrophages, it was maintained in RPMI-1640 medium supplemented with 10% FBS and 5000 U mL⁻¹ penicillin and 5 mg mL⁻¹ streptomycin in incubator at 37 °C with 5% of CO₂ atmosphere.

Both lineages were used with the formation of the confluent monolayer.

Antiproliferative assay

Promastigotes and epimastigotes at a density of 1 × 10⁶ parasites per milliliter, along with 10% FBS, were seeded into a 96-well plate. Subsequently, varying concentrations of the compounds (1, 5, 10, 50, and 100 μM) or culture medium were added. The plate was then incubated in a B.O.D. incubator for 72 hours or 96 hours, respectively. The assay was evaluated by the colorimetric salt XTT (0.5 mg mL⁻¹), when added it is enzymatically reduced by viable cell's mitochondrial dehydrogenases forming an orange-colored formazan compound. After the addition of XTT, the plate was put in a BOD incubator in the dark for 4 h. Then a reading was made at 450 nm in light spectrometer (Power WaveXS, BIO-TEK, EUA), and the IC₅₀ determined by non-linear regression analysis of the data.

Cytotoxicity assay

LLC-MK₂ cells (2.5 × 10⁵ cells per mL) or J774A.1 macrophages (5 × 10⁵ cells per mL), supplemented with 10% FBS, were seeded into a 96-well plate and incubated at 37 °C in a 5% CO₂ atmosphere for 24 hours. Following this initial incubation, the medium was removed, and varying concentrations of the compounds (10, 50, 100, 500, and 1000 μM) were added. The plate was then incubated for an additional 96 hours for LLC-MK₂ cells and 48 hours for macrophages.

After the incubation time, the assay was evaluated by the colorimetric salt MTT (2 mg mL⁻¹). MTT is reduced by viable cell's mitochondrial enzymes, originating a purple formazan crystal. After adding MTT, the plate was incubated in the dark in incubator at 37 °C with 5% CO₂ atmosphere for 4 h. Then, the MTT was removed, the formazan crystals diluted in DMSO, and reading was made at 570 nm in light spectrometer (Power WaveXS, BIO-TEK, EUA). The CC₅₀ was determined by non-linear regression analysis of the data.

Statistical analysis

Statistical analysis was performed in Microsoft Excel® version 2402, using triplicate values of the standard and selectivity index (SI) in each treatment with the different selected molecules. Initially, the Shapiro–Wilk test corroborated the normal distribution within the measurements in the same group treated with a given compound. Afterward, the statistical difference between the SI value of each treatment concerning the standard was verified using the one-tailed t-test (p < 0.05).

Lipophilicity calculations

clog P values (logarithm of n-octanol/water partition coefficient P) were obtained by the software PerkinElmer ChemDraw®, Level: Professional, Version: 16.0.1.4 (77).

Conclusions

A new series of Knoevenagel alkylidenes 7a–o derived from hydroxychloroquine were synthesized in good yields. Subsequently, compounds 7a–o, along with their precursors HCQ·H₂SO₄ (4·H₂SO₄), HCQ (4), and HCQAA (5), were subjected to antiparasitic assays against T. cruzi, L. infantum, and L. amazonensis.

The results indicated that, even at the highest concentration (100 μM), the Knoevenagel precursors 4·H₂SO₄, 4, and 5 did not inhibit the growth of any of the parasites.

For T. cruzi epimastigotes, the Knoevenagel alkylidenes 7a, 7h, and 7i, derived from isovaleraldehyde and para- and meta-chlorinated benzaldehydes (SI = 6.29, 6.79, 5.62, respectively), exhibited antiprotozoal activity. However, compared to the standard drug benznidazole (SI = 42.63), the selectivity index was approximately seven times lower.

Against L. amazonensis promastigotes, alkylidenes 7a, 7g, 7h, and 7i, which contain an isovaleraldehyde chain and chlorine at the ortho-, meta-, and para-positions of the phenyl group, demonstrated higher biological activity, with SI values ranging from 8.11 to 10.97, comparable to AmpB (SI = 9.37). Moreover, these compounds exhibited lower toxicity (CC₅₀ between 244.68 ± 42.51 and 419.97 ± 26.88 μM) in J774A.1 macrophages, compared to the standard drug (AmpB, CC₅₀ = 92.70 ± 10.27 μM). The same alkylidenes (7a, 7g, 7h, 7i) showed SI values between 7.77 and 10.41 for L. infantum promastigotes, although these values were nine times lower than AmpB (SI = 88.29).

Based on these results, we conclude that the Knoevenagel adducts are more active against L. amazonensis. These findings underscore the importance of the isobutyl and chlorine substituents in the aromatic ring of the Knoevenagel adducts derived from hydroxychloroquine in enhancing their biological activity.

Further studies are underway to evaluate their effects on L. amazonensis and to explore the potential mechanisms of action of the Knoevenagel adduct structures derived from hydroxychloroquine, with the aim of contributing to the development of new antileishmanial drugs.

Data availability

The data supporting this article have been included as part of the ESI.†

Author contributions

Conceptualization: Priscila P. Dario, Luis H. D. Yamashita, Marcelo G. M. D'Oca, Danielle Lazarin-Bidoia, Celso Vataru Nakamura. Synthesis and characterization of molecules: Priscila P. Dario, Gabriel L. Kosinski, Guilherme A. Justen. Experiment for identification and quantification of E/Z isomers: Kahlil S. Salome. Biological assays: Luis H. D. Yamashita, Danielle Lazarin-Bidoia, Celso V. Nakamura. Statistical analysis: Priscila P. Dario. Formal analysis: Priscila P. Dario, Luis H. D. Yamashita. Resources: Marcelo G. M. D'Oca, Daniel da S. Rampon, Danielle Lazarin-Bidoia, Celso V. Nakamura, Fernanda A. Rosa. Data curation: Priscila P. Dario, Luis H. D. Yamashita, Kahlil S. Salome. Writing – original draft: Priscila P. Dario, Luis H. D. Yamashita. Writing – review & editing: Priscila P. Dario, Luis H. D. Yamashita, Marcelo G. M. D'Oca, Daniel da S. Rampon, Danielle Lazarin-Bidoia, Celso V. Nakamura. Supervision: Marcelo G. M. D'Oca, Daniel da S. Rampon, Danielle Lazarin-Bidoia, Celso V. Nakamura. Project administration: Marcelo G. M. D'Oca, Danielle Lazarin-Bidoia, Celso V. Nakamura. Funding acquisition: Marcelo G. M. D'Oca, Daniel da S. Rampon, Danielle Lazarin-Bidoia, Celso V. Nakamura, Fernanda A. Rosa.

Conflicts of interest

There are no conflicts to declare.