Ce(SO₄)₂-catalysed the highly diastereoselective synthesis of tetrahydroquinolines via an imino Diels Alder ABB′ type reaction and their in vivo toxicity and imaging in zebrafish embryos

Abstract

An efficient and practical approach has been developed for the synthesis of N-(tetrahydroquinolinyl-4) amides 3a–l in good yield with high diastereoselectivity. The strategy comprises the domino type ABB′ imino Diels Alder reaction catalysed by a cerium(IV) salt between anilines and N-vinyl amides for the preparation of a 12-membered library of tetrahydroquinolines that were tested for their in vivo toxicity against zebrafish embryos. Upon determining their LC₅₀ values, N-(8-methoxy-2-methyl-tetrahydroquinolinyl-4) acetamide 3k was identified as the most toxic derivative with an LC₅₀ below 95 μM (24 mg L⁻¹). Finally, the phenotypes induced, at concentrations below their LC₅₀, were analyzed at 48, 72 and 96 hours post fertilization, wherein the treated embryos manifested diverse visual phenotypes, such as big yolk sacs (3b, 3h, 3j), pericaldial edemas (3a, 3i) and red blood cells in the liver region (3b, 3l), in comparison to the morphology of the control embryos, the phenotypes could be associated with specific biological targets.

---

Introduction

The 1,2,3,4-tetrahydroquinoline core (THQ) is a privileged class of N-heterocycle present in a number of natural and synthetic products with biological activity, which play an important role in medicinal chemistry.^1,2 In particular, substituted 2-methyl-THQs derivatives have found many applications in pharmaceutical industries such as antibiotics,^3,4 inhibitors of the P-glycoprotein in multidrug resistant cancer cells⁵ and antagonists of the prostaglandin D2 receptors.⁶ They have also shown control in the expression of the Ecdysone Receptor (EcR) in a Aedes aegypti model for agrochemical purposes⁷ and have been employed as chiral ligands for transition-metal catalysts in asymmetric organic synthesis.⁸

Due to their biological and chemical properties, the development of newer and efficient methodologies for the heterocyclic ring construction of 2-methyl-THQs derivatives is of current interest and is still in demand. During the last decade, one-pot and multicomponent reactions have allowed the direct synthesis of these molecules from simple substrates in a highly efficient manner using the Povarov reaction.⁹ Several Lewis acids (BF₃·OEt₃, InCl₃, Ln(OTf)₃) and Brønsted acids (HSO₃Cl, TFA, TsOH) have been found to be excellent catalysts for the three-component Povarov reaction (ABC) in the construction of highly substituted libraries of THQs.¹⁰ However, these conditions have not been very efficient for the synthesis of 2-methyl-THQs derivatives.¹¹

Since 2011, our group has explored an extended version of multicomponent reactions, designated as ABB′ type reactions, in which the component B is incorporated in two distinct manners (B and B′) into the component A, ensuring the complexity and functional diversity of the final product.¹² As a result, we developed a novel approach used for the synthesis of 2-methyl-THQs via the Povarov reaction catalysed by phthalic acid.¹³ Although the good yields in which the new THQs derivatives were obtained, this method was recently improved by the replacement of the solvent medium from acetonitrile to SDS as a micellar aqueous medium, generating a new library of amidyl-2-methyl-THQs in high yield and diastereoselectivity.¹⁴

Nowadays, preliminary in vivo toxicological tests of organic small molecules (SMs) are considered as one of the main and necessary steps during the discovery and development of future drugs. Among the different models for in vivo bioprospection of novel SMs, the zebrafish embryo model provides an inexpensive, reliable and efficient first-level screening model for testing the toxicity, efficacy, and tissue-targeting of a large number of these SMs because of the close homology between the zebrafish and human genome.¹⁵

Furthermore, huge amounts of zebrafish embryos can be generated, developed quite rapidly and synchronously with well-defined developmental stages, wherein its transparency at embryonic and larval stages facilitate the direct visual observation of the toxic and phenotypic effects of SMs in vivo.^16,17

According to the statements described above and with the current need to conduct studies related with the synthesis of 2-methyl-THQs that could be tested in the zebrafish model to reveal their toxicological and phenotypic profiles, our research was focused on: (i) previous reports wherein cerium salts, specially cerium ammonium nitrate (CAN), have catalysed the formation of THQs and quinolines from vinyl ethers,¹⁸ establish the optimal conditions for the reaction of anilines and N-vinyl amides catalysed by Ce(IV) salts according to the variables: solvent, reaction times, temperature, catalyst loading and the yield in which the desired product was obtained; (ii) using the standardized conditions, prepare a 12-membered library of 2-methyl-THQs using the imino Diels Alder (Povarov) ABB′ type reaction; (iii) determine the toxicity (LC₅₀) of the prepared 2-methyl-THQs in zebrafish embryos and (iv) analyse the phenotypes induced by these SMs at concentrations below their LC₅₀, in zebrafish embryos at 24, 48, and 72 hours post chemical exposure. All this, to contribute to future SAR studies of this interesting class of molecules.

Results and discussion

We initiated our study by testing CAN, a well-known promoter of single-electron transfer reactions (SETR) via a cation radical-mediated chain mechanism,¹⁹ as catalyst in the model reaction between 4-aminoacetofenone 1a and N-vinyl acetamide 2a. The first experiments were carried out in two different solvents (MeOH and MeCN) using CAN (10 mol%) as the catalyst, resulting in the formation of the respective 6-acetyl-2-methyl-4-acetamido-1,2,3,4-tetrahydroquinoline 3a after 30 minutes in 92% yield when MeCN was used as the solvent at room temperature (Table 1, entry 2).

Table 1 Synthesis of 2-methyl-THQ 3a catalysed by Ce(IV) salts. The screening of Ce(IV) salts, solvent systems and temperature range^a

Entry	Ce(IV) salt (mol%)	Solvent	Time (h)	Yield^b (%)	d.r.^c (cis/trans)
a The reaction was performed on a 1 mmol scale using 1a (2 mmol), 2a (4.2 mmol), catalyst (10–30 mol%) and in the respective solvent (20 mL) at room temperature.b The isolated yield after chromatographic purification.c The diastereomeric ratio of 3a was determined by ^1H NMR analysis.d The reaction was performed at 60 °C.
1	CAN (10)	MeOH	2	56	87/13
2	CAN (10)	MeCN	0.5	92	87/13
3	Ce(SO4)2 (30)	MeCN	5	78	97/3
4	Ce(SO4)2 (15)	MeCN	6	89	97/3
5	Ce(SO4)2 (15)	MeCN	6	67^d	95/5

---

Encouraged by the promising results obtained with CAN, but being aware of the tedious work-up operations during the extraction and purification process and knowing the oxidizing properties of CAN that may cause the oxidation of the THQ core to the corresponding quinoline,²⁰ we were forced to find and evaluate alternative cerium(IV) salts as a possible catalyst in this ABB′ type reaction. With that knowledge in mind, our attention was drawn to explore the use of Ce(SO₄)₂ due to its low toxicity, solubility in many organic solvents, low cost, stability to air moisture and commercial availability in comparison to CAN. In addition, this catalyst has been used as a mild and efficient oxidant catalyst for various organic transformations.^21–23 In that order, we performed the model reaction in the presence of Ce(SO₄)₂, studying the effect of reducing the catalyst loading from 30 to 15 mol% in MeCN at room temperature, obtaining the desired compound 3a in good to excellent yield, 78–89% (Table 1, entries 3–5).

Although the use of CAN as a catalyst provided the 2-methyl-THQs 3a in shorter reaction times (0.5–2 h) and in good yield (56–92%) than the experiments performed with Ce(SO₄)₂, the selection of better reaction conditions were carried out based on the best yield and diastereomeric ratio in which 3a was isolated (Table 1). Thus, we found that at a higher catalyst loading, 30 mol%, the product 3a was obtained in 78% yield and with a cis-diastereoselectivity (cis/trans 97 : 3) when the reaction was performed at room temperature (entry 3).

Interestingly, when the catalyst loading was decreased to 15 mol%, the best results in terms of yield and diastereoselectivity were obtained for 3a (89%, cis/trans = 97/3) after 6 hours (entry 4). Finally, increasing the reaction temperature to 60 °C significantly decreased the yield and the diastereoisomeric ratio of the Povarov reaction, suggesting a negative influence on the chemical transformation associated with the equilibrium of the reacting species (entry 5).

Further studies using ¹H NMR and COSY experiments led to establishing the complete stereochemistry of cis-3a through the trans-diaxial relationship between protons H_4ax and H_3ax (J = 11.5 Hz) and H_2ax and H_3ax (J = 11.5 Hz), which means that cis-THQ 3a was formed selectively. This finding shows the high diasteroselectivity of the imino Diels Alder ABB′ type reaction catalysed by Ce(SO₄)₂, improving the results previously reported by Menéndez et. al.¹⁸ wherein under the catalysis of CAN the respective THQs were obtained in a diastereoisomeric ratio of 90 : 10 (cis/trans). A possible explanation for the high diastereoselectivity displayed by this reaction catalysed by Ce(SO₄)₂, could be related with the molecular orbital (MO) interactions, wherein an endo approach is often preferred between the diene AB and the dienophile 2a′,²⁴ and with the nature of ligands bonded to the cerium(IV) ion. In the case of Ce(SO₄)₂, the SO₄²⁻ ligands adopt a tetrahedrical disposition that promote the coordination between the metal and the nitrogen present in the diene AB (N⋯Ce(IV)), as well with the oxygen of the dienophile 2a′ (Ce(IV)⋯O). Thus, the proximity and strength of these interactions increase the diastereoselectivity of the reaction in contrast with CAN, wherein the NO₃⁻ ligands form a bulky complex with a cuboctahedron geometry that reduces the force of these interactions and lowers the diastereoselectivity.²⁵

Having the optimized reaction conditions in hand, the versatility, substrate scope and limitations of our protocol for the Povarov reaction were broadened to other anilines and diverse enamides with different functional groups, wherein the reaction proceeded smoothly and afforded the corresponding 2-methyl-THQs 3a–l in good to excellent yield (63–89%) with a diastereoselectivity >97% in all cases (Table 2).

Table 2 The substrate scope of the Povarov ABB′ type reaction catalyzed by Ce(SO₄)₂. The synthesis of 2-methyl-THQs under the optimized reaction conditions^a

a The reaction was performed on a 2 mmol scale using 1a (2 mmol), 2a (4.2 mmol), catalyst (15 mol%) in MeCN (20 mL) at room temperature.b The isolated yield after SiO₂ column chromatography. The diastereomeric ratio of 3a–l was determined by ¹H NMR analysis.

---

The substituents influence at the aniline ring was first investigated and showed that the reaction could proceed well using diverse para-substituted anilines 1a–d (H, OMe, Ac, COOH) giving the corresponding products 3 without affecting the yield or diastereoselectivity of the THQ product.

However, the chemical structure of enamides 2a–c influenced negatively the yields in which the product 3 was isolated, suggesting that the reactivity of the enamides follows an apparent trend of 2a ≥ 2c > 2b. The structure of the 2-methyl-THQs 3a–l was elucidated through ¹H, ¹³C and 2D NMR experiments, wherein the HMBC and NOESY experiments were relevant to prove the relative stereochemistry at C-2 and C-4 and the diastereoselectivity of the Povarov ABB′ type reaction (Fig. 1).

Fig. 1 Selected (a) HMBC (----) and (b) NOESY (↔) correlations for compound 3i.

For the case of 3i, the stereochemistry of the methyl group and the stereochemistry at C-2, wherein the methyl group is located in an equatorial position. Finally, the key NOESY correlation of protons H_ax at C-2 and H_ax′ at C-4 indicates that they are on the same plane, establishing the stereochemistry at C-4, confirming the cis-configuration of groups bonded to C-2 and C-4.

Previous debates and discussions have suggested that the reaction mechanism for the formation of THQs under the catalysis of Ce(IV) salts such as CAN occur through concerted asynchronous transition states through species 4a–b generated via a SETR (Scheme 1a).^26–28 To prove this hypothesis, our first attempt was involved the addition of some radical trapping agents (RTAs) to the reaction mixture with the idea that these will react with the intermediates 4a–b, diverting in that way the course of the reaction.²⁹ However, when 1,1-diphenylethylene 5 or 3,4-dihydroxybenzoic acid 6, two well-known RTAs, were introduced in excess to our model reaction between 1a and 2a, using Ce(SO₄)₂ or CAN, the reaction proceeded normally and afforded the corresponding 2-methyl-THQ 3a in 89% yield, indicating that the expected radical intermediates 4a–b were not formed (Scheme 1b).

Scheme 1 (a) The proposed intermediates 4a–b in a radical mechanism. (b) The control experiment performed for the model reaction with the addition of radical traps (RTAs).

Another debated mechanistic proposal has suggested a stepwise mechanism via ionic intermediates and with the purpose of confirming this hypothesis we used two nucleophiles, such as ethanol and n-butanol, to trap the iminium cation I3 in the model reaction in the presence of Ce(SO₄)₂ (Scheme 2).³⁰ We observed that during the course of the reaction the GC/MS analysis revealed that the addition products, which would incorporate a nucleophile molecule, were not observed under equimolar amounts with respect to 1a. According to these results and the reported background on mechanistic aspects of imino DA reactions,³¹ a possible mechanism can now be proposed. First, the key intermediate I2 is formed in situ when aniline 1a reacts with the enamide 2a obtaining intermediate I1, which is stabilized by Ce(IV) that also promotes the liberation of an amide as a by-product BP. The second role of the Ce(IV) salt is to stabilize imine I2 and make it act as a more electron-poor diene to undergo the Povarov reaction catalysed by Ce(SO₄)₂ with the other enamide molecule in a stepwise manner. This leads to the formation of intermediate I3, which during the last cyclization step will adopt a chair-like transition state, explaining the preference of an equatorial arrangement of the methyl (C-2) and amido (C-4) groups and the high cis stereochemistry observed for intermediate I4 that finally undergoes a deprotonation step to obtain 3a–l.

Scheme 2 The proposed mechanism for the reaction of arylamines and N-vinyl amides under Ce(SO₄)₂ catalysis.

Finally, having this chemical toolbox available, we focussed our efforts in exploring their biological value to search for functional SMs in two zebrafish screens. These are (i) embryo toxicity testing and (ii) phenotypic analysis of the images of zebrafish embryos treated with the 2-methyl-THQs 3a–l, divided into three groups, 2-methyl-4-acetamido-THQs 3a–d, 2-methyl-4-formyl-THQ 3e–h and 2-methyl-4-pyrrolidin-2-one-THQs 3i–l.¹⁷

Determination of the zebrafish embryo LC₅₀

After the zebrafish embryos were established as a potential alternative for the acute toxicity test of SMs, Ali et al. described the standard protocol for the rapid toxicity assessment³² that we adapted to determine the toxicity of the 2-methyl-THQs 3a–l library in zebrafish embryos.

Zebrafish embryos were exposed to 2-methyl-THQs 3a–l to determine their level of toxicity. The SMs were introduced into the E3 medium after dechorionation at 24 hours post fertilization (hpf) and the exposed embryos were incubated until 96 hpf, the point that corresponds to the final stage of embryogenesis. The exposed embryos were incubated at 28 ± 2 °C and were examined at 24, 48 and 72 hours of chemical exposure (96 hpf) using a light-dissecting stereomicroscope. Once the embryos were classified as dead according to the established endpoints,³³ the data collected from three independent exposures were analysed statistically and the determined LC₅₀ values, at 96 hpf expressed in μmol L⁻¹, are shown in Table 3.

Table 3 Zebrafish embryo LC₅₀ values found for the 2-methyl-tetrahydroquinoline-based small molecule library

Comp.	Zebrafish LC50^a		Aquatic animal acute toxicity^b
	μmol L^−1 ± SEM	mg L^−1 ± SEM
a LC50 values are the mean ± SEM of three different experiments in triplicate.b Toxicity scale (mg L^−1) = highly toxic 0.1–1 (HT), moderately toxic 1–10 (MT), slightly toxic 10–100 (ST), practically non-toxic 100–1000 (PN) and relatively harmless >1000 (RH).
3a	351.2 ± 5.7	86.5 ± 1.4	ST
3b	223.8 ± 7.1	45.7 ± 1.5	ST
3c	400.4 ± 7.2	93.8 ± 1.7	ST
3d	429.7 ± 0.3	106.7 ± 0.1	PN
3e	429.3 ± 0.4	99.7 ± 0.1	ST
3f	216.0 ± 7.1	41.1 ± 1.4	ST
3g	435.9 ± 3.7	95.9 ± 0.8	ST
3h	449.2 ± 10.2	105.2 ± 2.4	PN
3i	426.3 ± 7.0	116.1 ± 1.9	PN
3j	330.7 ± 13.7	76.2 ± 3.2	ST
3k	94.9 ± 3.7	24.7 ± 1.0	ST
3l	358.4 ± 9.0	98.3 ± 2.5	ST

---

Comparing the results depicted in Table 3, wherein the LC₅₀ is expressed in mg L⁻¹, with the acute toxicity rating scale established by the Fish and Wildlife service (FSW), we found that almost all the 2-methyl-THQs 3a–l can be classified as a practically non-toxic or slightly toxic agents possessing a LC₅₀ between 24 and 116 mg L⁻¹.³⁴

We also found that the THQ bearing a methoxy group in the C-6 position and the pyrrolidin-2-one core at C-4 3k was found to be the most lethal compound (LC₅₀ = 25 mg L⁻¹). This may be due to the synergism of these two organic functions because compounds 3c and 3g, wherein the methoxy group is present with the acetamido and formyl moiety, and compounds 3i, 3j and 3l, wherein the pyrrolidin-2-one core is present without the methoxyl group, exhibit a LC₅₀ above 77 mg L⁻¹.

Regarding the less toxic compounds, we observed that the compounds substituted with a carboxyl group at C-6 3d, 3h and 3l were found to be practically non-toxic (LC₅₀ = 98.3–105.2 mg L⁻¹), similar to the toxicity of aspirin (LC₅₀ 101 mg L⁻¹) in zebrafish.³⁵

In general, taking into account that the toxicity of unsubstituted THQs at C-6 (3b, 3f, 3j) was found to be independent of the group present at C-4 (pyrrolidin-2-one, acetamide or formyl moieties), the toxicity of the 2-methyl-THQs 3a–l is strongly correlated with the nature of the substituent in C-6, a fact that needs to be proved in further studies with di- and tri-substituted THQs to select novel non-toxic agents that could be used in advance biological studies.

In vivo zebrafish phenotyping

Once the LC₅₀ for compounds 3a–l was determined, a range of concentrations were established to screen our molecules using zebrafish embryos to study how the activity and normal expression of endogenous genes and proteins were subtly manipulated by the exogenous 2-methyl-THQs 3a–l molecules.³⁶

Table 4 shows a summary of the morphological defects observed from the Prim-5 stage (24 hpf) to 96 hpf after chemical exposure to the twelve selected 2-methyl-THQs 3a–l at concentrations below their LC₅₀.

Table 4 A summary of the effects of 2-methyl-THQs 3a–l on zebrafish embryos after 96 hpf

Comp. (μM)	Morphological defects^a
	Curved bodies	Delayed hatching	Yolk sac edema	Mild intestine	Pericardial edema
a ++++ = very severe effect (75–100%); +++ = severe effect (50–75%); ++ = moderate effect (25–50%); + = minimal effect (5–25%); +/− = either minimal or no effect (0–5%); — = no effect (0%).b E3 medium + DMSO (2%).c E3 medium without DMSO.
3a (100)	+	—	++	++	+++
3b (150)	++	—	+++	+++	+
3c (200)	—	++++	—		—
3d (200)	+	—	++		+++
3e (125)	—	++++	—		—
3f (150)	+/−	—	+		+/−
3g (250)	—	++++	—		—
3h (200)	+/−	—	++++	+++	++
3i (100)	+	—	+/−		++++
3j (200)	+++	—	+++	+	+
3k (25)	++	—	+/−		+/−
3l (200)	++	—	++		+
Control^b	—	—	—		—
Blank^c	—	—	—		—

---

Compounds 3c, 3e and 3g induced a developmental delay in the treated embryos, exhibiting a delayed hatching after 96 hpf that did not allow the observation of phenotypes around the spinal cord, yolk sac or heart cavity at these concentrations and two or three concentrations below the LC₅₀ value.

In general, all the embryos treated with compounds 3a–l, at three or four concentrations below the LC₅₀ of each derivative, did not manifest any visual phenotype and those embryos reached their corresponding development stage after 96 hpf, without any visual evidence that might indicate that the morphology of these treated embryos differed from the control embryo morphology (see ESI†).

However, at one or two concentrations below the LC₅₀ of each compound, the treated embryos exhibit big yolk sac edemas (YS), mild intestine (MI), curved bodies (CB), and pericardial edemas (PE) around the heart cavity compared to the control embryo morphology, phenotypes that were identified as the major visual morphological defects observed for rest of the series of 2-methyl-THQs 3a–l (Table 3).

The CB abnormalities exhibited by almost all the THQs 3a–l, in low or high degrees were concentration-dependent of the administered compound and cannot be associated with any gene, enzyme or protein at first sight because the additional observation of the yolk sac and pericardial edemas in these treated embryos could come together with the CB. If the larvae is unable to absorb nutrients, minerals, phospholipids, triaglycerols and vitamins from the intestinal tract, edemas will be formed in the heart cavity resulting in heart failure and slow the heart rate. The distribution of those endogenous reserves from the YS and oxygen into the bloodstream will eventually delay the development of the treated embryos, resulting in CB.³⁷

Compounds 3d, 3h and 3j showed the most severe edemas around the yolk sac, accompanied by disorders or abnormalities in the middle intestine, at concentrations below their LC₅₀, but the PE shown by these embryos were not very severe. This suggests that these compounds will not reach higher concentrations in the cardiovascular system, maybe because they have affected the liver during their transformation, damage that disturbs the subsequent metabolism of the lipids contained in the yolk sac during the early stages of development, exhibiting wider YS than those of the control (Fig. 2).³⁸

Fig. 2 Microscope images at 96 hpf of the treated embryos with compounds 3b (150 μM), 3h (200 μM) and 3j (200 μM); main phenotypes identified: severe yolk sac edema (YS); abnormalities across the middle intestine (MI) and mild pericardial edemas (PE).

We confirmed that these phenotypes were not dependent on the LC₅₀ of each derivative, because the YS, MI or PE phenotypes were not lethal to the treated embryos; for example, compound 3h, one of the less lethal compounds, displayed several YS and MI and mild MI phenotypes at concentrations much lower than the LC₅₀. In addition, the degree in which the phenotypes YS, MI and PE were expressed for compounds 3b and 3j can be correlated with the substituent in C-4, where we found that the N-acetyl group was found to be more toxic than the pyrrolidin-2-one moiety for the unsubstituted THQs 3b and 3j, specially towards the gut, mid intestine and heart cavity.

Pericardial inflammation of the heart cavity was observed for compounds 3a, 3d, 3i and 3l, accompanied with CB, YS and MI, malformations that were identified as the main phenotypes observed in these embryos when they were compared to the control embryos. Although the embryos treated with compound 3e did not break their membrane, showing a delayed hatching after 96 hpf that did not allow the identification of other phenotypes. The other compounds substituted with the acetyl group at C-6 (3a and 3i) exhibited the most large PE among the series of compounds 3a–l. In comparison with the control, wherein the looping process places the ventricle (V) and atrium (A) side by side, so that the two chambers largely overlap each other and have a small size that make them indistinguishable by the lateral view. The embryos treated with compounds 3a and 3i showed looping defects characterized by the abnormal morphology of the heart chambers, resulting in a linear heart with a stretched string-like atrium, positioning the V anterior to the A wherein they can be easily distinguished without overlap into the prominent bulbs around the heart cavity (Fig. 3).

Fig. 3 Microscope images at 96 hpf of the treated embryos with compounds 3a (100 μM) and 3i (100 μM); heart looping defects of the treated embryos, arrowhead indicates a large pericardial edemas (PE) and yolk sac edemas (YS). Blue and red indicate the ventricle (V) and atrium (A), respectively.

We also noted that the morphology of V and A in the treated embryos became abnormal after 72 hpf when the observed looping disturbance creates an alteration such that the ventricle was positioned anterior to the atrium. These phenotypes, which clearly involve a severe cardiac insufficiency, can be correlated with the leucine-rich repeat containing protein 10 (Lrrc10), a cardiac-specific factor that could be perturbed by the exogenous agents 3a and 3i, disturbing the normal cardiac development that promotes the heart's looping and compaction within the pericardium.³⁹

In addition to defective hearts and PE phenotypes, the treatment of embryos with THQs 3a–l, especially the molecules 3d and 3l substituted with the carboxylic acid group at C-6, allowed the visualization of an accumulation of red blood cells (RBCs) in the liver region that suggests some type of liver damage after THQs exposure. This agglomeration perhaps will be responsible for the circulation defects and consequently leads to malformations during the heart development, because it has been established the functional relationship between the cardiovascular and digestive system in zebrafish (Fig. 4).⁴⁰

Fig. 4 Microscope images at 96 hpf of the treated embryos with compounds 3d (200 μM) and 3l (200 μM); embryos showed mil pericardial edemas (PE) and an accumulation of red blood cells (RBCs) inside the yolk sac in the liver region.

With the evidence of the PE, YS, MI and RBCs phenotypes, our findings indicate that THQs exposure (24–96 hpf) causes liver and heart failure that may lead to embryonic death. A possible mechanism that explains its toxicological process is defined herein (Scheme 3).

Scheme 3 The proposed toxicological and phenotypic profile of the THQs 3a–l.

First, the absorption of the THQs by the digestive system induces a liver injury when these molecules are trying to be metabolized by this organ. Once the liver is damaged, blood cells accumulate around this tissue and the larvae is unable to absorb the nutrients, minerals and vitamins from the intestinal tract into the bloodstream, resulting in developmental delays and edemas in the YS and the MI. Because the blood circulation has been perturbed, insufficient blood flow that contains low to middle concentrations of THQs and their metabolites will cause defects during the heart development after the Lrrc10 protein is disturbed, resulting in the abnormal morphology of the heart chambers and in PEs.

Therefore, if the liver and heart are damaged, the cessation of blood circulation and the lack of nutrient distribution will finally induce the loss of renal function and the eventual death of the embryo. However, additional studies are needed to prove this hypothesis.

Conclusions

We have developed a mild and efficient protocol for the diastereoselective synthesis of 4-amidyl-2-methyl-THQ based on the imino Diels Alder (Povarov) ABB′ type reaction catalysed by Ce(SO₄)₂, in which the yields and the diastereoselectivity of the obtained products were improved in comparison to previous reports. This strategy, which can be adapted to the synthesis of a range of other natural and synthetic tetrahydroquinolines, lead to the synthesis of a diverse library of THQs 3a–l from readily available starting materials and tolerance of a variety of substituents under the reaction conditions, providing the desired tetrahydroquinoline derivatives in modest to high yields and with excellent diastereoselectivity.

Finally, when this library was tested on the zebrafish screen, we found that molecule 3k was the most toxic derivative with an LC₅₀ below 95 μM, wherein the pyrrolidine ring at position C-4 and the methoxy group at position C-6 of the THQ core were identified as the main toxicophores. Notwithstanding, to understand the precise mode of action of these compounds that were active in this phenotypic screen, further study is needed. The selection of molecules that induced the discussed visual phenotypes (3a, 3b, 3d, 3h, 3i, 3j and 3l) will decrease the time and cost of further biological assays that intend to use these structures as a starting point in the discovery and development of new bioactive agents.

Experimental

Chemistry

Infrared (FT-IR) spectra were obtained on a Lumex Infralum FT-02 spectrometer, ν_max in cm⁻¹. The bands were characterized according to the functional group. ¹H NMR spectra were obtained on a Bruker Avance-400 (400 MHz) spectrometer. Chemical shifts are reported in ppm using the solvent resonance as an internal standard (CDCl₃: δ 7.26 ppm; DMSO-d₆: δ 2.50 ppm). Data were reported as follows: chemical shift, multiplicity (s = singlet, d = doublet, t = triplet, dd = doublet of doublets, br = broad, m = multiplet), coupling constants (Hz) and integration (see ESI†). ¹³C NMR spectra were obtained on a Bruker Avance-400 (400 MHz) spectrometer with complete proton decoupling. Chemical shifts are reported in ppm from the solvent resonance used as an internal standard (CDCl₃: δ 77.00 ppm; DMSO-d₆: δ 40.45 ppm). In the DEPT-135 spectra, the signals of CH₃ and CH carbons are shown as positive (+) and CH₂ carbons are shown negative (−). Quaternary carbons are not shown. A Hewlett Packard 5890a Series II Gas Chromatograph interfaced to an HP 5972 Mass Selective Detector (MSD) with an HP MS ChemStation Data system was used for MS identification at 70 eV using a 60 m capillary column coated with HP-5 [5%-phenylpoly(dimethylsiloxane)]. Accurate mass data were obtained on a Micromass Q-TOF using electrospray ionisation (ESI). Melting points were measured on a Fisher Johns melting point apparatus and are uncorrected.

Unless otherwise noted, all reactions have been carried out with distilled and dried solvents under atmosphere pressure. All work-up and purification procedures were carried out with reagent grade solvents (purchased from Aldrich and Merck) in air. Thin-layer chromatography (TLC) was performed using Merck silica gel 60 F254 precoated plates (0.25 mm). Column chromatography was performed using silicagel 60 (0.063–0.200 mm) 70–230 mesh.

General procedure for the synthesis of the 2-methyl-THQs 3a–l

To a stirred MeCN solution (20 mL) of Ce(SO₄)₂ (15% mol), the starting aniline 1a–d was added (2 mmol), after its dissolution, the corresponding N-vinyl amide 2a–c (4.2 mmol) was added to the reaction mixture and it was stirred at room temperature for the indicated time. After completion of the reaction as indicated by TLC, the mixture was treated with a saturated aqueous solution of NaHCO₃ and extracted with ethyl acetate (3 × 30 mL), dried over anhydrous Na₂SO₄ and evaporated. Crude product was purified by column chromatography on silica gel using a mixture of petroleum ether and ethyl acetate as the eluent to obtain the THQ derivatives. The compounds that resulted were found to be poorly soluble in common solvents (CH₂Cl₂, CHCl₃ and EtOAc) and very soluble in DMSO, MeCN or MeOH, the solvents used for their synthesis, NMR experiments and biological studies.

Toxicity testing and phenotypic screening of 2-methyl-THQs 3a–l using the zebrafish embryo model

Wild-type adult zebrafish of both sexes were separated in two tanks (30 L each), according to their gender, at 26 ± 2 °C under natural light–dark photoperiods. The fishes were fed twice daily and the water quality was recorded weekly, to acclimate the fishes for at least two weeks before experiments begin. For the reproduction of the adult fishes, small breeding tanks were set up in the evening previous to experiment, each containing three males and one female specimen. The tanks were isolated until next morning when the lights were switched on and natural mating occurs, without any perturbation.

The adult fishes were returned to their corresponding tank and the embryos were collected, pooled and washed with E3 medium and transferred to a 92 mm glass Petri dish. Furthermore, dead, delayed, malformed and unfertilized embryos were identified under a dissecting microscope and removed by select aspiration with a pipette. This last procedure was repeated at 12 and 20 hpf to remove the unfit embryos. Throughout this period of time, the embryos were kept at 28 ± 2 °C in an incubator under natural light–dark photoperiods.

The selected embryos of 24 hpf from the Petri dish were gently distributed into 96-well plates, placing a single embryo and 200 μL of E3 medium per well.

Adult zebrafish were cared for and used according to the Guide of the National Institute of Health for Care and Use of Laboratory Animals, keeping them healthy and free of any signs of disease. The Ethics and Research Committee of the Heart Institute of Bucaramanga approved the protocol under the Acta Number 050 of May 26 of 2012.

Determination of zebrafish embryo LC₅₀

For this experiment, in total 72 embryos were required per sample to run three independent experiments in three different plates and each compound was evaluated three times in the same plate, allowing the evaluation of four samples per plate. Compounds 3a–l were diluted into the E3 screening medium with 2% v/v of DMSO and aliquots of 200 μL were prepared at concentrations starting from 12.5 and finishing in 1250 μM (geometric series). The LC₅₀ determination (expressed in μmol of compound per L of solution) was based on the cumulative mortality after 72 hours of chemical exposure (96 hpf). Each embryo was examined under a dissecting microscope and the statistical analysis was made using Regression Probit analysis with SPSS for windows version 19.0. Data are expressed as the standard error of the mean (SEM) of three different experiments in triplicate.

Phenotypic screening using the zebrafish embryo model

Compounds 3a–l were diluted into the E3 screening medium with 2% v/v of DMSO and aliquots of 200 μL were prepared at concentrations from 5–400 μM, depending on the LC₅₀ of each THQ. The surrounding medium (200 μL) was carefully removed from the embryonic plates using an 8-multichannel pipette and then the appropriate chemical aliquot of each compound (200 μL), previously prepared, were added into the corresponding well of the embryonic plate. Eight controls wells were used peer plate, each containing E3 medium with 2% v/v of DMSO. The embryonic plates were incubated at 28 °C and examined at 48, 72 and 96 hours post-fertilization (hpf) using an OPTIKA zoom stereo microscope (trinocular version of model SZM-1).

Acknowledgements

This study was financially supported by the Colombian Institute for Science and Research (COLCIENCIAS) under the project No. RC-0346-2013. CAMB thanks Colciencias (Jovenes Investigadores e Innovadores) for the fellowship. LYVM thanks Universidad Santo Tomas for the financial support and CEPG acknowledges the fellowship given by the doctoral program COLCIENCIAS-Conv. 617.

References

1. For a detailed review in the field see: V. Sridharan, P. A. Suryavanshi and J. C. Menéndez, Chem. Rev., 2011, 111, 7157.
2. (a) A. Hagfeldt, G. Boschloo, L. Sun, L. Kloo and H. Pettersson, Chem. Rev., 2010, 110, 6595; (b) R. Chen, X. Yang, H. Tian, X. Wang, A. Hagfeldt and L. Sun, Chem. Mater., 2007, 19, 4007; (c) Y. Hao, X. Yang, J. Cong, A. Hagfeldt and L. Sun, Tetrahedron, 2012, 68, 552; (d) K. Rakstys, J. Solovjova, T. Malinauskas, I. Bruder, R. Send, A. Sackus, R. Sens and V. Getautis, Dyes Pigm., 2014, 104, 211.
3. R. J. Pagliero, S. Lusvarghi, A. B. Pierini, R. Brun and M. R. Mazzieri, Bioorg. Med. Chem., 2010, 18, 142.
4. R. N. Asolkar, D. Schröder, R. Heckmann, S. Lang, I. Wagner-Döbler and H. Laatsch, J. Antibiot., 2004, 57, 17.
5. R. Hiessböck, C. Wolf, E. Richter, M. Hitzler, P. Chiba, M. Kratzel and G. Ecker, J. Med. Chem., 1999, 42, 1921.
6. J. Liu, Y. Wang, Y. Sun, D. Marshall, S. Miao, G. Tonn, P. Anders, J. Tocker, H. L. Tang and J. Medina, Bioorg. Med. Chem. Lett., 2009, 19, 6840.
7. W. A. Kinney, C. A. Teleha, A. S. Thompson, M. Newport, R. Hansen, S. Ballentine, S. Ghosh, A. Mahan, G. Grasa, A. Zanotti-Gerosa, J. Dingenen, C. Schubert, Y. Zhou, G. C. Leo, D. F. McComsey, R. J. Santulli and B. E. Maryanoff, J. Org. Chem., 2008, 73, 2302.
8. G. Li, J. Ren, P. Yao, Y. Duan, H. Zhang, Q. Wu, J. Feng, P. C. K. Lau and D. Zhu, ACS Catal., 2014, 4, 903.
9. (a) H.-G. Imrich, J. Conrad, D. Bubrin and U. Beifuss, J. Org. Chem., 2015, 80, 2319; (b) C. R. Borel, L. C. Almeida Barbosa, C. R. Álvares Maltha and S. A. Fernandes, Tetrahedron Lett., 2015, 56, 662.
10. (a) D. R. Merchán Arenas and V. V. Kouznetsov, J. Org. Chem., 2014, 79, 5327; (b) V. V. Kouznetsov, Tetrahedron, 2009, 65, 2721; (c) S. Kobayashi, H. Ishitani and S. Nagayama, Synthesis, 1995, 1195.
11. G. Masson, C. Lalli, M. Benohouda and G. Dagousseta, Chem. Soc. Rev., 2013, 42, 902.
12. D. Tejedor and F. García-Tellado, Chem. Soc. Rev., 2007, 36, 484.
13. V. V. Kouznetsov, D. R. Merchán Arenas, C. Ortiz Areniz and C. Meléndez Gómez, Synthesis, 2011, 4011.
14. D. R. Merchán Arenas, C. A. Martínez Bonilla and V. V. Kouznetsov, Org. Biomol. Chem., 2013, 11, 3655.
15. S. Basu and C. Sachidanandan, Chem. Rev., 2013, 113, 7952.
16. V. E. Fako and D. Y. Furgeson, Adv. Drug Delivery Rev., 2009, 61, 478.
17. (a) C. E. Puerto Galvis and V. V. Kouznetsov, Org. Biomol. Chem., 2013, 11, 407; (b) C. E. Puerto Galvis and V. V. Kouznetsov, Org. Biomol. Chem., 2013, 11, 7372.
18. V. Sridharan, C. Avendaño and J. C. Menéndez, Tetrahedron, 2007, 63, 673.
19. V. Sridharan and J. C. Menéndez, Chem. Rev., 2010, 110, 3805.
20. K. Karnakar, J. Shangkar, S. N. Murthy, K. Ramesch and Y. V. D. Nageshwar, Synlett, 2011, 1089.
21. C. A. Horiuchi, T. Fukushima, N. Furuta, W. Chai, S. J. Ji, Y. Saito, C. Hashimoto, T. T. Takahashi, T. Sugiyama, A. Muto, Y. Sakata and S. Nozaki, J. Chem. Res., 2003, 270.
22. D. González-Calderón, C. A. González-González, A. Fuentes-Benítez, E. Cuevas-Yáñez, D. Corona-Becerril and C. González-Romero, Tetrahedron Lett., 2013, 54, 7164.
23. N. P. Selvam and P. T. Perumal, Can. J. Chem., 2009, 87, 698.
24. B. Han, X.-D. Jia, X.-L. Jin, Y.-L. Zhou, L. Yang, Z.-L. Liu and W. Yu, Tetrahedron Lett., 2006, 47, 3545.
25. C. E. Kefalidis, L. Castro, L. Perrin, I. Del Rosal and L. Maron, Chem. Soc. Rev., 2016 10.1039/c5cs00907c.
26. B. Han, X.-D. Jia, X.-L. Jin, Y.-L. Zhou, L. Yang, Z.-L. Liu and W. Yu, Tetrahedron Lett., 2006, 47, 3545.
27. P. Buonora, J. C. Olsen and T. Oh, Tetrahedron, 2001, 57, 6099.
28. S. Hermitage, D. A. Jay and A. Whiting, Tetrahedron Lett., 2002, 43, 9633.
29. T. Sato, N. Morita and M. Seno, Eur. Polym. J., 2001, 37, 2055.
30. (a) V. Lucchini, M. Prato, G. Scorrano, M. Stivanello and G. Valle, J. Chem. Soc., Perkin Trans. 2, 1992, 259; (b) O. Jiménez, G. de la Rosa and R. Lavilla, Angew. Chem., Int. Ed. Engl., 2005, 44, 6521.
31. D. Bello, R. Ramón and R. Lavilla, Curr. Org. Chem., 2010, 332.
32. S. Ali, H. G. J. van Mil and M. K. Richardson, PLoS One, 2011, 6, e21076.
33. Embryos were classified as dead if they present: (i) coagulation, (ii) lack of somite formation, (iii) lack of detachment of the tail-bud from the yolk sac (YS), (iv) their heart was not beating, or/and (v) they were no longer moving after external stimulation.
34. I. Juneidi, M. Hayyan and M. A. Hashim, RSC Adv., 2015, 5, 83636.
35. C. Zhang, C. Willett and T. Fremgen, in Current protocols in toxicology, John Wiley & Sons, Hoboken, NJ, USA, 2003, ch. 1, unit 1.7.
36. (a) S. Scholz, S. Fischer, U. Gündel, E. Küster, T. Luckenbach and D. Voelker, Environ. Sci. Pollut. Res., 2008, 15, 394; (b) K. Henn and T. Braunbeck, Comp. Biochem. Physiol., Part C: Toxicol. Pharmacol., 2011, 153, 91.
37. (a) D. Raldúa, M. André and P. J. Babin, Toxicol. Appl. Pharmacol., 2008, 228, 301; (b) A. Holmberg, C. Olsson and S. Holmgren, J. Exp. Biol., 2006, 209, 2472.
38. J.-H. He, S.-Y. Guo, F. Zhu, J.-J. Zhu, Y.-X. Chen, C.-J. Huang, J.-M. Gao, Q.-X. Dong, Y.-X. Xuan, C.-Q. Li and J. Pharmacol, Toxicol. Methods, 2013, 67, 25.
39. K.-H. Kim, D. S. Antkiewicz, L. Yan, K. W. Eliceiri, W. Heideman, R. E. Peterson and Y. Lee, Dev. Biol., 2007, 308, 494.
40. Y.-J. Ding and Y.-H. Chen, Toxicol. Appl. Pharmacol., 2012, 261, 59.

---

Footnote

† Electronic supplementary information (ESI) available: Detailed experimental procedures, characterization data, ¹H NMR, ¹³C NMR, DEPT-135, COSY, HMBC and HSQC NMR spectra and all photographic records of the phenotypic changes exhibited during zebrafish embryo development. See DOI: 10.1039/c6ra04325a

---