Open Access Article
Antonella Goggiamani
*a,
Antonio Arcadi
b,
Alessia Cioglia,
Martina De Angelisa,
Stefano Dessalvia,
Giancarlo Fabrizia,
Federica Iavaronecd,
Antonia Iazzetti
cd,
Alessio Sferrazza*ae and
Roberta Zoppolia
aDipartimento di Chimica e Tecnologie del Farmaco, Sapienza, Università di Roma, P.le A. Moro 5, 00185 Rome, Italy. E-mail: alessia.ciogli@uniroma1.it; m.deangelis@uniroma1.it; stefano.dessalvi@uniroma1.it; giancarlo.fabrizi@uniroms1.it; antonella.goggiamani@uniroma1.it; a.sferrazza@irbm.com; roberta.zoppoli@uniroma1.it
bDipartimento di Dipartimento di Scienze Fisiche e Chimiche, Università degli Studi di L'Aquila, Via Vetoio, 67100 Coppito (AQ), Italy. E-mail: antonio.arcadi@univaq.it
cDipartimento di Scienze Biotecnologiche di base, Cliniche Intensivologiche e Perioperatorie, Università Cattolica del Sacro Cuore, L.go Francesco Vito 1, 00168 Rome, Italy. E-mail: federica.iavarone@unicatt.it; antonia.iazzetti@unicatt.it
dPoliclinico Universitario ‘A. Gemelli’ Foundation-IRCCS, Rome, 00168, Italy
eAlessio Sferrazza is currently a research scientist in IRBM S.p.A., Medicinal Chemistry Department, Pomezia, Roma, Italy
First published on 29th March 2023
The synthesis of 2,3-dihydropyrazino[1,2-a]indol-4(1H)-ones from the sequential reaction of amino acid methyl esters with readily available indole-2-ylmethyl acetates is described. The reaction proceeds in situ under basic conditions of highly unstable and reactive 2-alkylideneindolenines followed by Michael-type addition of α-amino acid methyl esters/intramolecular cyclization.
Then, the synthesis of polysubstituted pyrazino[1,2-a]indol-4-one derivatives through straightforward one-pot approaches from easily available building blocks would be particularly significant considering their great potential as molecular scaffolds in drug discovery.
During our studies in the synthesis of heterocyclic compounds,8 great interest has been devoted to constructing/functionalizing indole and indole-fused polycyclic systems through simple domino processes.9 Recently, we reported the metal-free synthesis of 2-(aminomethyl), (tosylmethyl), and (aryloxymethyl) indoles 2 starting from readily available 2-indolylmethylacetates 1 and N, O, and S soft nucleophiles. The reaction proceeds through in situ generation of highly reactive 2-alkylideneindolenines I under basic conditions as provided through ESI-MS and IRMPD spectroscopy analyses.10 Furthermore, we developed a straightforward assembly of polysubstituted 1,2-dihydro-3H-pyrrolo[1,2-a]indol-3-ones 3 through a domino palladium-catalyzed reaction of 2-indolylmethylacetates 1 with 1,3-dicarbonyl compounds (Scheme 1).11
Based on this background, we envisaged that the reaction of 2-indolylmethyl acetates 1 with α-amino acids 4 should achieve a general entry into the title target 5 through the in situ generation of 2-methide-2H-indole intermediate I/nucleophile Michael addition/cyclization cascade reaction according to the retrosynthetic analysis of Scheme 2.
Herein, we describe the scope and limitations of this approach to the synthesis of multi-substituted 2,3-dihydropyrazino[1,2-a]indol-4(1H)-one 5.
For our initial investigations, we examined the reaction of the indole-2-ylmethyl acetates 1a with two equivalents of methyl L-phenylalaninate 4a using K2CO3 in MeCN at 120 °C (Table 1, entry 1). The reaction proceeded to give the desired 3-benzyl-2,3-dihydropyrazino[1,2-a]indol-4(1H)-one 5a in 51% yield. A significant side reaction led to the formation of 2-((1H-indol-2-yl)methyl)-3-benzyl-2,3-dihydropyrazino[1,2-a]indol-4(1H)-one 8a in 26% yield. Very likely, the target product 5a is, also, prone to competitively give an aza-Michael addition on 2-alkylideneindolenine Ia to afford 8a (Scheme 3).
| Entryb | Solvent | T (°C) | Base | 4a (equiv.) | t (h) | 5a (%) | 8a (%) | 1a (%) |
|---|---|---|---|---|---|---|---|---|
| a Unless otherwise stated, reactions were carried out on a 0.404 mmol scale using 2 equiv. of base in 3.0 mL of solvent.b Yields are given for isolated products.c Under microwave irradiation.d The reaction was carried out in 9.0 mL of solvent. | ||||||||
| 1 | MeCN | 120 | K2CO3 | 2 | 23 | 51 | 26 | — |
| 2c | MeCN | 170, 150 W | K2CO3 | 2 | 0.25 | 54 | 25 | — |
| 3d | MeCN | 120 | K2CO3 | 2 | 32 | 34 | 16 | 27 |
| 4 | DMSO | 120 | K2CO3 | 2 | 2 | 30 | 6 | — |
| 5 | DMSO | 100 | K2CO3 | 2 | 2 | 37 | 17 | — |
| 6 | DMF | 120 | K2CO3 | 2 | 0.75 | 39 | 20 | — |
| 8 | MeCN | 120 | K3PO4 | 2 | 24 | 9 | 5 | 53 |
| 9 | MeCN | 120 | Cs2CO3 | 2 | 2 | 36 | 21 | — |
| 10 | MeCN | 120 | Na2CO3 | 2 | 22 | — | — | 90 |
| 11 | MeCN | 120 | K2CO3 | 3 | 24 | 54 | 20 | — |
| 12 | MeCN | 120 | K2CO3 | 4 | 22 | 58 | 14 | — |
| 13 | MeCN | 120 | K2CO3 | 5 | 18 | 73 | 8 | — |
The features of solvent (Table 1, entries 4–6), base (Table 1, entries 8–10), dilution of the reaction mixture (Table 1, entry 3), and microwave irradiation (Table 1, entry 2) didn't influence the ratio 5a/8a. However, the wished 5a was isolated in 73% yield by carrying out the reaction in MeCN at 120 °C under the presence K2CO3 as the base and 5 equivalents of 4a. The excess of 4a was found to play a key role to achieve better product selectivity control (Table 1, entries 11–13).
This protocol was then used when the process was extended to include other indoles 1 and α-amino acid methyl esters 4. Our preparative results are summarized in Table 2.
| Entryb | R1 | R2 | 1 | R3 | 4 | t (h) | 5 (%) | 8 (%) |
|---|---|---|---|---|---|---|---|---|
| a Unless otherwise stated, reactions were carried out on a 0.404 mmol scale at 120 °C using 5 equiv. of 4, 2 equiv. of K2CO3 in 3.0 mL of MeCN.b Yields are given for isolated products.c The reaction was carried out in presence of 2 equiv. of 4.d Yield was calculated from the 1H NMR analyses.e 1d was recovered in 15% of yield. | ||||||||
| 1c | H | H | 1a | CH2CH(Me)2 | 4b | 22 | 36 5b | 32 8b |
| 2 | H | H | 1a | CH2CH(Me)2 | 4b | 15 | 59 5b | 32 8b |
| 3c | H | H | 1a | CH(Me)2 | 4c | 19 | 59 5c | 29 8c |
| 4 | H | H | 1a | CH(Me)2 | 4c | 19 | 71 5c | 13 8c |
| 5 | Me | H | 1b | CH(Me)2 | 4c | 17 | 76 5d | 19 8d |
| 6 | Me | H | 1b | CH2Ph | 4a | 21 | 44 5e | 20 8e |
| 7 | OMe | H | 1c | CH2Ph | 4a | 6 | 67 5f | 15 8f |
| 8 | Cl | H | 1d | CH2Ph | 4a | 4.5 | 58 5g | 6 8gd,e |
| 9 | H | C6H5- | 1e | CH2Ph | 4a | 6.5 | 68 5h | 8 8h |
Several 2,3-dihydropyrazino[1,2-a]indol-4(1H)-one derivatives bearing a variety of useful functional groups have been prepared in good yields: in particular, the 5-chloro derivative (Table 2, entry 8) indicates that this protocol is a useful tool for obtaining more complex derivatives through subsequent transition metal-catalyzed cross-coupling reactions.
Indoles bearing electron-releasing and electron-donating groups at the C5 position show a comparable reactivity for the same α-amino acid 4 (Table 2, entries 6–8) in terms of overall yield and 5/8 ratio; this result suggests that the electrophilicity of 5-substituted 2-methide-2H-indole intermediates I, even modulated by the electronic effects of the substituents, do not affect the reaction outcome in presence of the strong neutral nitrogen nucleophiles 4 and 5.
Interestingly, the size of the R3 chain of the α-amino acid 4 seems to play a role in the achievement of the reaction: in fact, while a satisfactory yield was obtained with L-phenylalanine and L-valine methyl ester (Table 1, entry 13 and Table 2, entry 4), a significant lowering in the yield was observed with L-leucine (Table 2, entry 2).
In order to avoid the side formation of 8 using a lower molar excess of nucleophile, we decided to utilize the α-amino acids methyl esters as N-benzyl derivatives 9. The brief investigation done showed that compounds 10 were obtained in good yield (Table 3, entry 1–5) without the isolation of side products. As reported before, the correlation between the steric hindrance of the alkyl substituent R3 in 9 and the yield of the reaction was observed (Table 3, entry 1 vs. entry 2–3).
| Entryb | R1 | R2 | 1 | R3 | R4 | 9 | t (h) | 10 (%) |
|---|---|---|---|---|---|---|---|---|
| a Unless otherwise stated, reactions were carried out on a 0.290 mmol scale at 120 °C using 1.5 equiv. of 9, 2 equiv. of K2CO3 in 3.0 mL of MeCN.b Yields are given for isolated products. | ||||||||
| 1 | H | H | 1a | CH2Ph | Me | 9a | 24 | 84 10a |
| 2 | H | H | 1a | CH(CH3)2 | Et | 9b | 48 | 71 10b |
| 3 | H | H | 1a | CH(CH3)(C2H5) | Me | 9c | 16 | 67 10c |
| 4 | Cl | H | 1d | CH(CH3)(C2H5) | Me | 9c | 17 | 54 10d |
| 5 | OMe | H | 1c | CH2Ph | Me | 9a | 16 | 68 10e |
Finally, we briefly investigate the outcome of the reaction using L-proline 4d, to extend the scope of the process to the construction of the pirrolopirazinoindol-12-one tetracyclic core 11 (Table 4).
| Entryb | R1 | 1 | t (h) | 11 (%) | 12 (%) | 13 (%) |
|---|---|---|---|---|---|---|
| a Unless otherwise stated, reactions were carried out on a 0.404 mmol scale at 120 °C using 2 equiv. of 4d, 2 equiv. of K2CO3 in 3.0 mL of MeCN.b Yields are given for isolated products.c (1H-indol-2-yl)methanol was isolated in 14% yield. | ||||||
| 1 | H | 1a | 20 | 76 11a | 8 12a | — |
| 2c | OMe | 1c | 16 | Traces | 39 12c | 12 13c |
Table 4 shows the preliminary results obtained using two different indole precursors 1a and 1c. Interestingly, while the reaction between 1a and L-proline gave the target compound 11a in large excess over the side products 12a and 13a (Table 4, entry 1), using indole derivative 2b the outcome of the reaction is reversed, being the 12b and 13b products the principal, while 1b was isolated in traces (Table 4, point 2). A tentative explanation of these results could be made by taking into account the electronic effect of 5- substituent on the intermediate I: (Fig. 2) to this end, the HF calculations at 631G** level on Ia and Ib showed a significant difference in the partial positive charge on the methylidene carbon atom (see ESI† for more details).13 Since the carbanion derived from 11 may be considered highly delocalized, it is reasonable to believe that the greater electrophilicity of Ic prevents an accumulation of the product 11b but, on the contrary, favors its rapid transformation into 12c and 13c.
Because of the crucial role of chiral nitrogen heterocycles in medicinal chemistry, we also analyzed the stereochemical outcome of our synthesis (Table 5).
| Entryc | T °(C) | t (h) | 5f (%) | e.r.d | 8f (%) |
|---|---|---|---|---|---|
| a Unless otherwise stated, reactions were carried out on a 0.404 mmol scale using 5 equiv. of 4a, 2 equiv. of K2CO3 in 3.0 mL of MeCN.b The major enantiomer L is shown.c Yields are given for isolated products.d e.r was calculated from enantioselective HPLC analysis (see ESI for more details).e The starting indole 1c was recovered in 50% of yield.f The starting indole 1c was recovered in 60% of yield.g The starting indole 1c was recovered almost quantitatively. | |||||
| 1 | 120 | 6 | 67 | 68 : 32 |
15 |
| 2 | 90 | 6 | 69 | 85 : 15 |
— |
| 3e | 70 | 18 | 31 | 92 : 8 |
— |
| 4f | 70 | 6 | 27 | 95 : 5 |
— |
| 5g | 50 | 24 | — | — | |
It has long been noted that amino acids rapidly undergo racemization process when heated in acidic or basic conditions and subsequently our target pyrazoindole could undergo racemization at C3 via enolate.14
Indeed, when the reaction of 1c with methyl L-phenylalaninate 4a was carried out at 120 °C, HPLC analysis showed the formation of 5f as enantioenriched mixture (68
:
32, Table 5, entry 1), while 8f was obtained as racemate. The enantiomeric ratio and its dependence on reaction temperature were established by HPLC on Chiralpak IA column of purified pyrazinoindoles. When the reaction was performed at lower temperatures (Table 5, entry 2–4) the loss of enantiomeric purity was limited (up to 95
:
5, Table 5, entry 4).
:
20); IR (neat): 2989, 2796, 1444, 1272, 1183, 941, 734 cm−1; 1H NMR (400.13 MHz) (CDCl3): δ 8.36 (d, J = 8.0 Hz, 1H), 7.39 (d, J = 7.6 Hz, 1H), 7.27–7.16 (m, 7H), 6.17 (s, 1H), 4.11 (d, J = 16 Hz, 1H), 3.88 (d, J = 16 Hz, 1H), 3.79 (dd, J1 = 8.7 Hz, J2 = 4.0 Hz, 1H), 3.44 (dd, J1 = 14 Hz, J2 = 4.0 Hz, 1H), 3.06 (dd, J1 = 14.0 Hz, J2 = 8.7 Hz, 1H), 1.79 (s, 1H); 13C NMR (100.6 MHz) (CDCl3): δ 169.1 (C), 137.5 (C), 136.6 (C), 134.7 (C), 129.5 (CH), 128.8 (CH), 127.0 (CH), 124.5 (CH), 124.2 (CH), 120.2 (CH), 116.3 (CH), 103.1 (CH), 61.3 (CH), 41.9 (CH2), 36.3 (CH2). HRMS: m/z [M + H]+ calcd for C15H17NO2: 277.1335; found: 277.1336.
:
15); IR (neat): 3401, 2986, 2808, 1692, 1356, 1188, 691 cm−1; 1H NMR (400.13 MHz) (CDCl3): δ 8.40 (d, J = 7.9 Hz, 1H), 7.48 (d, J = 6.8 Hz, 1H), 7.41–7.18 (m, 10H), 7.06–7.02 (m, 1H), 7.00–6.94 (m, 2H), 6.36 (s, 1H), 6.11 (s, 1H), 4.43 (dd, J1 = 16.7, J2 = 1.8 Hz, 1H), 3.89–3.82 (m, 2H), 3.74–3.69 (m, 2H), 3.25 (dd, J1 = 14.1, J2 = 4.3 Hz, 1H), 3.09 (dd, J1 = 14.2, J2 = 11.5 Hz, 1H); 13C NMR (100.6 MHz) (CDCl3): δ 168.8 (C), 138.3 (C), 136.0 (C), 134.9 (C), 134.6 (C), 133.9 (C), 129.7 (C), 129.5 (CH), 128.8 (CH), 128.3 (C), 127.1 (CH), 124.9 (CH), 124.6 (CH), 121.7 (CH), 120.5 (CH), 120.2 (CH), 119.7 (CH), 116.5 (CH), 111.0 (CH), 105.8 (CH), 101.5 (CH), 64.6 (CH), 52.0 (CH2), 43.5 (CH2), 35.6 (CH2). HRMS: m/z [M + H]+ calcd for C27H24N3O: 406.1914; found: 406.1915.
:
4); IR (neat):3024, 1782, 1450, 1373, 695 cm−1; 1H NMR (400.13 MHz) (CDCl3): δ 8.54 (d, J = 7.9 Hz, 1H), 7.58 (d, 1H), 7.43–7.31 (m, 7H), 7.29–7.22 (m, 3H), 7.05–7.03 (m, 2H), 6.41 (s, 1H), 4.39 (dd, J1 = 16.9, J2 = 1.7 Hz, 1H), 3.93 (d, J = 4.4 Hz, 1H), 3.90 (d, 1H), 3.82 (d, J = 13.3 Hz, 1H), 3.66 (d, J = 13.3 Hz, 1H), 3.35 (dd, J1 = 14.4, J2 = 4.9 Hz, 1H), 3.25 (dd, J1 = 14.4, J2 = 10.1 Hz, 1H); 13C NMR (100.6 MHz) (CDCl3): δ 169.4 (C), 138.2 (C), 137.6 (C), 134.9 (C), 134.6 (C), 129.7 (C), 129.4 (CH), 128.8 (CH), 128.4 overlapping (CH), 127.5 (CH), 126.7 (CH), 124.6 (CH), 124.3 (CH), 120.3 (CH), 116.5 (CH), 105.4 (CH), 66.4 (CH), 58.1 (CH2), 43.1 (-CH2), 35.5 (CH2). HRMS: m/z [M + H]+ calcd for C25H23N2O: 367.1805; found: 367.1804.Footnote |
| † Electronic supplementary information (ESI) available. See DOI: https://doi.org/10.1039/d3ra01335a |
| This journal is © The Royal Society of Chemistry 2023 |