Synthetic strategies to a telomere-targeted pentacyclic heteroaromatic salt

Abstract

Three routes have been explored to synthesise the telomere-targeted agent 3,11-difluoro-6,8,13-trimethyl-8H-quino[4,3,2-kl]acridinium methosulfate 3. Application of a 6-(2-azidophenyl)phenanthridine precursor 11 gave an entry to the indazolo[2,3-f]phenanthridine ring system 12 not the required quino[4,3,2-kl]acridine. A six step synthesis starting from 2,6-dibromo-4-methylbenzonitrile 13via a 1-arylacridin-9(10H)-one intermediate, 19 or 21, gave the required 3 in low overall yield (<10%). The most efficient route entailed the one-pot (five step) conversion of 1,2-dimethyl-6-fluoroquinolinium methosulfate 23 to 3 in 33% yield employing triethylamine as base and nitrobenzene as solvent.

---

Introduction

Prior to initiation of our own work, the tetracyclic pyrido[4,3,2-kl]acridine skeleton was represented in the literature only by the natural product necatorone 1, elaborated by the fungus Lactarius necator and whose structure was deduced by Steglich and colleagues.² In previous papers in this series we have developed a range of synthetic approaches to tetra-, penta- and hexa-cyclic systems based on the necatorone scaffold. These methods have included thermal extrusion of nitrogen from triazolyl-substituted acridines,³ radical cyclisations of bromo-substituted 9-anilinoacridines⁴ and palladium(0)-mediated derivatisation of substituted acridin-9(10H)-ones.⁵ From these endeavours we have identified two series of bioactive pentacyclic systems with an affinity for DNA: topoisomerase II-inhibitory compounds such as the indolizino[7,6,5-kl]acridinium chloride 2⁶ which intercalates into duplex DNA sequences;⁷ and quaternary 8,13-dimethylquino[4,3,2-kl]acridinium salts exemplified by the methosulfate 3 (RHPS4)⁸ (Fig. 1) which binds to, and stabilises, DNA G-quadruplex structures, for example the (TTAGGG)_n sequences⁹ which characterise the ends (telomeres¹⁰) of chromosomes.

Fig. 1 Structures of bioactive polycyclic acridines and numbering of the quino[4,3,2-kl]acridine ring system.

The enzyme telomerase maintains telomere lengths, is upregulated in most human tumour types and is considered a relevant molecular target in the search for novel anticancer therapies.¹¹ The quinoacridinium salt 3 is a potent inhibitor of telomerase (IC₅₀ 0.33 µM)¹² and, in our recent pharmacological investigations, has been shown to compromise telomeric integrity and rapidly provoke the emergence of a senescent phenotype in human breast¹ and melanoma tumour cells.¹³ In an effort to enhance the ability of pentacyclic acridinium salts to stabilise G-quadruplex structures and increase their potency as telomere-targeted agents, we have exploited palladium(0) couplings to synthesise and biologically profile quinoacridinium salts related to 3 but bearing varied and multiple substituents at the 3-, 6- and 10-(or 11) positions.^5,14 (See Fig. 1 for numbering system). Of the many novel structures prepared compound 3 still retained the most desirable range of pharmacological^1,13 and pharmaceutical properties¹⁴ and has emerged as the clinical candidate from this series. In this paper we assess different synthetic routes which might be adaptable to a large-scale synthesis of 3.

Results and discussion

General synthetic considerations

In planning synthetic strategies to an appropriately-substituted quino[4,3,2-kl]acridine 4 or 5 which might be further processed by methylation to furnish salt 3, we have investigated two potential disconnections which could be applied generally to synthesise molecules of this class bearing varied substituents (X,Y,Z) in the three homocyclic rings. These involve cyclisations of 6-arylphenanthridine (Fig. 2; Route A) and 1-arylacridin-9(10H)-one (Route B) precursors. These routes might provide alternatives to the unlikely processing of a 1,2-methylquinolinium salt directly to 3 (Route C, see Scheme 3),⁸ an adaptation of a reaction reported by Ozczapowicz et al. in 1988.¹⁵

Fig. 2 Synthetic strategies to trisubstituted quino[4,3,2-kl]acridines.

Attempted synthesis of pentacyclic salt 3 from a 2-arylphenanthridine (Route A)

The tolylboronic acid 6 was coupled with 2-bromo-4-fluoroaniline under Suzuki conditions¹⁶ to afford biaryl 7 (68%) which reacted with 5-fluoro-2-nitrobenzoic acid under DCC coupling conditions to generate the anilide 8. This was cyclised under Bischler–Napieralski conditions to the nitroaryl-substituted phenanthridine 9 which was further routinely transformed to the corresponding amino- and azido-phenylphenanthridines 10 and 11, respectively. Thermolysis of 11 in boiling 1,2,4-trichlorobenzene at 214 °C gave a product with the molecular weight (by mass spectrometry: m/z 319 [M + 1]) expected for pentacycle 4. However, its physical properties were inconsistent with this assignment: significantly, the product was a pale yellow colour with very poor solubility in methanol, whereas an orange–red colouration and methanol-soluble character would be expected for compound 4. Moreover, in earlier work we have shown that the unsubstituted quino[4,3,2-kl]acridine ring system shows a long wavelength absorption at λ_max 443 nm in ethanol with a bathochromic shift to 488 nm on addition of HCl.⁴ Significantly, the product of thermolysis of 11 showed a UV absorption (λ_max) at 380 nm with no bathochromic shift on the addition of HCl. We have assigned the indazolo[2,3-f]phenanthridine structure 12 to this compound based on analogous cyclisations of o-azidoarenes bearing suitably disposed heteroaromatic residues.¹⁷ In the present example, nucleophilic displacement of nitrogen from the azido group by the phenanthridine nitrogen atom to form 12 is favoured over a nitrene insertion process which might have led to the required quino[4,3,2-kl]acridine 4 (Scheme 1).

Scheme 1 Attempted synthesis of quino[4,3,2-kl]acridine 4 from a 2-arylphenanthridine precursor (Route A). Reagents and conditions: (i) 2-bromo-4-fluoroaniline, tetrakis triphenylphosphine palladium (5 mol%), Na₂CO₃, DME–EtOH–H₂O under nitrogen, reflux; (ii) 2-nitro-5-fluorobenzoic acid, DCC in CHCl₃, 25 °C; (iii) POCl₃, 90 °C; (iv) tin(II) chloride dihydrate, EtOH, reflux, then NaOH; (v) NaNO₂, 1 M H₂SO₄, 5 °C, then NaN₃; (vi) 214 °C in 1,2,4-trichlorobenzene.

Synthesis of pentacyclic salt 3 from a 1-arylacridin-9(10H)-one (Route B)

The starting point to prepare a substituted 1-arylacridinone suitable for cyclisation to the methylated pentacycle 5 was 2,6-dibromo-4-methylbenzonitrile 13, available by a Sandmeyer reaction on diazotised 2,6-dibromo-4-methylaniline.¹⁸ Coupling between 13 and 4-fluoroaniline was achieved under Buchwald conditions¹⁹ mediated by Pd(OAc)₂–Cs₂CO₃–BINAP in refluxing toluene to give a mixture of the diarylamine 14 (65%), together with the disubstituted product 15 (20%) and unreacted starting material (15%). Cyclisation of 14 to the 1-bromoacridinone 16 was accomplished in 80% H₂SO₄ at 100 °C and methylation to the 10-methylacridinone 17 involved the use of NaH and dimethyl sulfate (DMS). On a medium scale it was possible to prepare 25 g of the methylacridinone 17 from 60 g of benzonitrile 13.

Protected 2-aminobenzeneboronic acids required for Suzuki coupling to the 1-bromo-10-methylacridinone 17 have been prepared from N-pivaloylanilines via directed ortho-lithiation with n-BuLi and trimethyl borate followed by an acid quench.^5,20 Yields of the required amino-protected fluorobenzeneboronic acid 18 from this route were very low (<5%). But encouragingly, Suzuki coupling between 17 and 18 with Pd(PPh₃)₄ catalyst and NaHCO₃–aqueous DME afforded the coupled pivaloylamine 19 which could be deprotected and cyclised in EtOH–5 M HCl to pentacyclic acridine 5 in 65% yield. An alternative coupling partner to 17, the boronic ester 20, could be prepared more efficiently from 3-fluoro-6-nitrophenol^21,22 and afforded the 1-(3-fluoro-6-nitroaryl)acridinone 21 (47%) with dichloro[1,1-bis(diphenylphosphino)ferrocene]palladium(II) and Na₂CO₃ in dioxane. Reductive cyclisation of 21 to the pentacycle 5 (65%) was effected with tin(II) chloride in EtOH, followed by 5 M HCl. Finally, methylation of 5 to the required 8,13-dimethylquinoacridinium salt 3 was accomplished with DMS in nitromethane (48%). Thus the required compound 3 is available in six steps from the benzonitrile 13 (Scheme 2) but in <10% overall yield.

Scheme 2 Synthesis of quino[4,3,2-kl]acridine 3 from a 1-arylacridin-9(10H)-one precursor (Route B). Reagents and conditions: (i) 4-fluoroaniline, Pd(OAc)₂, BINAP, Cs₂CO₃, toluene, 100 °C; (ii) 80% H₂SO₄, 100 °C; (iii) NaH in DMF, Me₂SO₄; (iv) 17 and 20 in dioxan, PdCl₂(dppf) (10 mol%), Na₂CO₃, 80 °C; (v) tin(II) chloride dihydrate, EtOH, reflux; (vi) Me₂SO₄ in MeNO₂, reflux.

Synthesis of pentacyclic salt 3 from 6-fluoro-2-methylquinoline (Route C)

Without the serendipitous discovery by Ozczapowicz and co-workers of the mechanistically-intriguing first synthesis of a 8,13-diethylquino[4,3,2-kl]acridinium salt from the conversion of an N-ethylquinaldinium salt with base¹⁵ it is unlikely that this approach would have been contemplated. We have earlier (Scheme 3) proposed a tentative reaction pathway for the Ozczapowicz synthesis initiated by addition of the carbanion 24, generated from 23 and base, to the iminium bond of the cation of 23 to form adduct 25.⁸ This could undergo an electrocyclic ring-opening to 26, thereby generating a conjugated system aligned to undergo cyclisation to the tetrahydro-quinoacridinium cation 27. Oxidation, presumably in situ, would give the salt 3.

Scheme 3 Synthesis of quino[4,3,2-kl]acridine 3 from 6-fluoro-2-methylquinoline (Route C). Reagents and conditions: (i) Me₂SO₄ at 100 °C; (ii) NEt₃ in nitrobenzene, 120 °C, 24 h.

We have now revisited this synthesis as we have been unable to replicate the high yields reported by the Polish workers.^15,23 Our efforts to secure a short route to 3 started with 6-fluoro-2-methylquinoline 22 which was methylated with DMS to afford the 1,2-dimethylquinolinium methosulfate 23. The presence of the highly-fluorescent maroon salt 3 in reaction mixtures can be readily monitored by TLC and the product isolated preferably by column chromatography on silica gel employing butan-1-ol–AcOH–H₂O (0.65 : 0.15 : 0.25) as eluting solvents, or by recrystallisation. Salt 3 has unusual physical properties, being soluble in both water and (to some extent) organic solvents; it has a disconcerting habit of disappearing into plastic cuvettes and tubing. We have attributed this amphiphilic nature to the fact that the delocalised positive charge of the iminopyridinium core is shielded by a hydrophobic aryl cladding.¹⁴

Standardising on the use of ethanol as solvent, it is clear that the nature of the base is critical to the efficiency of the Ozczapowicz conversion 23 → 3. Employing the primary aliphatic amine (n-butylamine), secondary aliphatic amines (pyrrolidine and morpholine), tertiary amines (triethylamine and Hünigs base) and heterocyclic bases pyridine and 4-methylpyridine as substitutes for piperidine, gave lower yields of the pentacyclic salt; whereas the use of DBU or the inorganic bases NaHCO₃, K₂CO₃, Cs₂CO₃ and even basic alumina gave only traces (TLC). The reaction is apparently sensitive to concentration effects and the stoichiometry of the base: the optimum conditions, determined after much experimentation, involved reacting 23 (5 g) and piperidine (3 mol. eq.) in boiling ethanol (150 mL) for 2 days which gave 0.6 g (15%) of product 3 when crystallised from acetone—and this on a good day! Increasing reaction time to 7 days did not increase the yield. Exploring a range of refluxing protic solvents gave yields of 3 in the order: ethanol > methanol > propan-2-ol > butan-2-ol ≫ trifluoroethanol; no product was detected (TLC) employing dioxane as solvent.

The proposed mechanism (Scheme 3) requires a final oxidation step to remove, formally, 4 atoms of hydrogen from cation 27. If the reaction is conducted in degassed or air-gassed ethanol, or under argon, formation of compound 3 is still observed. Evidently molecular oxygen is not the oxidant in this case, implying that reaction intermediates must be consumed as oxidants, thus accounting for the poor yield. Again, standardising on ethanol as solvent, incorporation of external oxidants MnO₂, SeO₂, p-chloroanil, benzoquinone or iodine gave no improvement in yields. In order to avoid potential premature oxidation of reaction intermediates by oxidants, aliquots of DDQ (0.1–0.2 mol. eq.) were titrated into the reaction mixture every 45, 90 or 120 min. Yields of the required salt 3 remained stubbornly in the 5–15% range. Mindful of the role of nitrobenzene as solvent and oxidant in the Scraup reaction, its use as external oxidant was explored. Application of nitrobenzene as solvent at 120 °C with either piperidine, pyridine, K₂CO₃ or Cs₂CO₃ as bases either gave no reaction (pyridine), or led to the formation of black mixtures where only traces of 3 could be detected (TLC). Finally, on the point of conceding defeat we discovered that reaction of 23 with triethylamine as a base in nitrobenzene at 120 °C for 15 h gave the required salt in 33% yield making this adaptation a feasible process for large-scale synthesis. Microwave irradiation (100 W) at 180 °C for 13 min gave a reduced yield.

Experimental

General details

Melting points were measured on a Galenkamp apparatus and are uncorrected. IR spectra were recorded on a Perkin Elmer Spectrum One FT-IR spectrometer as KBr discs and UV spectra on a Pharmacia Biotech Ultraspec 2000 UV/Visible spectrophotometer. Mass spectra were recorded on a Micromass LCT spectrometer. NMR spectra were recorded on a Bruker Avance 400 instrument at 400.13 MHz (¹H) and 100.62 MHz (¹³C) in [²H₆]DMSO or CDCl₃; coupling constants are in Hz. Merck silica gel 60 (40–60 µm) was used for column chromatography.

2-Amino-5-fluoro-3′-methylbiphenyl 7

2-Bromo-4-fluoroaniline (14 g, 0.073 mol) and m-tolylboronic acid (10 g, 0.073 mol) were dissolved in a solution of 1,2-dimethoxyethane (400 mL), ethanol (40 mL) and water (20 mL). Nitrogen gas was bubbled through the resulting solution for 10 min, after which, sodium carbonate (17 g, 0.16 mol) and tetrakis triphenylphosphine palladium (4.2 g, 5 mol%) were added. The resulting mixture was heated under reflux for 15 h, cooled, and solvent removed under vacuum. The residue was dissolved in chloroform (500 mL), washed with water (250 mL), dried (sodium sulfate) and evaporated to leave a yellow oil. The product was purified by vacuum distillation to leave 7 as a colourless oil (14.6 g, 98%), bp 130–132 °C (1.4 mbar) (Found: C, 77.3; H, 6.0; N, 6.9%. C₁₃H₂FN requires C, 77.6; H, 6.0; N, 7.0%); ν_max/cm⁻¹ 3356, 3370 (NH), 1606, 1500, 1272, 1201, 1163, 812; ¹H NMR δ_H (CDCl₃) 7.37 (1 H, t, J 7.6, ArH), 7.27–7.19 (3 H, m, ArH), 6.92–6.86 (2 H, m, ArH), 6.71 (1 H, dd, J 4.8, 9.2, ArH), 3.66 (2 H, br s, NH₂), 2.43 (3 H, s, CH₃); m/z (ES) 202 (M + 1).

5-Fluoro-N-(5-fluoro-3′-methylbiphenyl-2-yl)-2-nitrobenzamide 8

Dicyclohexylcarbodiimide (12.25 g, 0.054 mol) was added to a solution of biphenyl 7 (10.86 g, 0.054 mol) and 2-nitro-5-fluorobenzoic acid (10 g, 0.054 mol) in chloroform (200 mL) at room temperature. After stirring for 15 h, the precipitate was filtered from solution and the filtrate evaporated under vacuum. The residue was recrystallised from ethyl acetate–heptane to give 8 as a white crystalline solid (18.4 g, 93%), mp 151–153 °C (Found: C, 64.8; H, 3.8; N, 7.5%. C₂₀H₁₄F₂N₂ O₃ requires C, 65.2; H, 3.8; N, 7.6%); ν_max/cm⁻¹ 3310 (NH), 1654 (CO), 1587, 1530, 1354 (NO₂), 1209, 812; ¹H NMR δ_H ([²H₆]DMSO) 10.15 (1 H, br s, NH), 8.25 (1 H, dd, J 4.8, 9.2, ArH), 7.63 (1 H, dd, J 5.6, 8.8, ArH), 7.57 (1 H, ddd, J 2.8, 6.4, 9.2, ArH), 7.38–7.20 (7 H, m, ArH), 2.35 (3 H, s, CH₃); m/z (ES-HRMS) 369.1040 (M + 1). Calcd for C₂₀H₁₅F₂N₂ O₃ (M + 1) 369.1050.

2-Fluoro-6-(5-fluoro-2-nitrophenyl)-9-methylphenanthridine 9

2-Nitrobenzamide 8 (18 g, 0.048 mol) was dissolved in phosphorous oxychloride (100 mL) and heated at 90 °C for 2 h. After cooling, the reaction mixture was quenched with ice (500 g), the precipitate filtered from solution and washed with water (200 mL). The white solid was recrystallised from ethanol to give 9 as a white powder (16.1 g, 94%), mp 198–201 °C (Found: C, 68.6; H, 3.4; N, 7.9%. C₂₀H₁₂F₂N₂ O₂ requires C, 68.6; H, 3.4; N, 8.0%); ν_max/cm⁻¹ 1619 (C=N), 1583, 1352 (NO₂), 1210, 815; ¹H NMR δ_H ([²H₆]DMSO) 8.77 (1 H, s, ArH), 8.67 (1 H, dd, J 5.6, 8.0, ArH), 8.08 (1 H, dd, J 6.0, 9.2, ArH), 7.75–7.57 (5 H, m, ArH), 2.62 (3 H, s, CH₃); m/z (ES-HRMS) 351.0822 (M + 1). Calcd for C₂₀H₁₃F₂N₂ O₂ (M + 1) 351.0945.

6-(2-Amino-5-fluorophenyl)-2-fluoro-9-methylphenanthridine 10

Tin(II) chloride dihydrate (22.5 g, 0.1 mol) was added to a solution of phenanthridine 9 (14 g, 0.04 mol) in ethanol (250 mL). The resulting mixture was heated under reflux for 2 h, cooled and the solvent removed under vacuum. The residue was dissolved in ethyl acetate (400 mL), washed with 5 M sodium hydroxide solution (2 × 200 mL), followed by water (100 mL), and dried (sodium sulfate). The concentrated solution gave a yellow solid which was recrystallised from ethanol to give 10 as a white powder (10.5 g, 82%), mp 155–157 °C (Found: C, 74.7; H, 4.3; N, 8.7%. C₂₀H₁₄F₂N₂ requires C, 75.0; H, 4.4; N, 8.7%); ν_max/cm⁻¹ 3423, 3316, 3226 (NH), 1643, 1619, 1570, 1497, 1442, 1234, 829; ¹H NMR δ_H ([²H₆]DMSO) 8.72 (1 H, s, ArH), 8.63 (1 H, dd, J 2.8, 10.8, ArH), 8.12 (1 H, dd, J 5.6, 8.8, ArH), 7.75 (1 H, d, J 8.4, ArH), 7.65 (1 H, ddd, J 2.6, 8.8, 11.6, ArH), 7.58 (1 H, dd, J 1.2, 8.8, ArH), 7.10 (1 H, ddd, J 3.2, 8.8, 12.0, ArH), 7.00 (1 H, dd, J 3.2, 9.6, ArH), 6.87 (1 H, dd, J 4.8, 8.8, ArH), 4.86 (2 H, br s, NH₂), 2.61 (3 H, s, CH₃); m/z (ES) 321 (M + 1).

6-(2-Azido-5-fluorophenyl)-2-fluoro-9-methylphenanthridine 11

The aminophenylphenanthridine 10 (5 g, 0.016 mol) was stirred in 1 M sulfuric acid (200 mL) at room temperature for 3 h, then cooled to 5 °C in an ice bath. A solution of sodium nitrite (1.3 g, 0.019 mol) in water (10 mL) was added dropwise, and stirring continued for a further 60 min. A solution of sodium azide (1.22 g, 0.019 mol) in water (10 mL) was then added over a 15 min period. After stirring at 5 °C for a further 1 h, the resulting precipitate was filtered from solution and washed with water (100 mL) to leave 11 as an off-white solid (4.95 g, 92%), mp 189 °C (decomp.) (Found: C, 69.1; H, 3.5; N, 15.9%. C₂₀H₁₂F₂N₄ requires C, 69.4; H, 3.5; N, 16.2%); ν_max/cm⁻¹ 2126 (N₃), 1620 (C = N), 1489, 1204, 815; ¹H NMR δ_H ([²H₆]DMSO) 8.76 (1 H, s, ArH), 8.67 (1 H, dd, J 2.8, 10.4, ArH), 8.14 (1 H, dd, J 6.0, 9.2, ArH), 7.69 (1 H, ddd, J 2.8, 8.4, 11.2, ArH), 7.62–7.52 (4 H, m, ArH), 7.42 (1 H, dd, J 2.4, 8.8, ArH), 2.63 (3 H, s, CH₃); m/z (ES-HRMS) 347.1118 (M + 1). Calcd for C₂₀H₁₃F₂N₄ (M + 1) 347.1108.

6,13-Difluoro-3-methylindazolo[2,3-f]phenanthridine 12

6-(2-Azido-5-fluorophenyl)-2-fluoro-9-methylphenanthridine 11 (4.0 g, 0.011 mol) was suspended in 1,2,4-trichlorobenzene (100 mL) and heated at 214 °C for 3 h. The solvent was removed by vacuum distillation and the residue triturated with diethyl ether to precipitate the product, which was filtered from solution and recrystallised from chloroform to give 12 as a pale yellow solid (3.2 g, 87%), mp 247–249 °C (Found: C, 75.1; H, 3.8; N, 8.7%. C₂₀H₁₂F₂N₂ requires C, 75.2; H, 4.1; N, 8.8%); ¹H NMR δ_H (CDCl₃) 8.89 (1 H, dd, J 5.6, 9.2, ArH), 8.33 (1 H, d, J 8.0, ArH), 8.15 (1 H, s, ArH), 8.11 (1 H, dd, J 2.4, 10.0, ArH), 7.94–7.91 (2 H, m, ArH), 7.56 (1 H, dd, J 1.2, 8.0, ArH), 7.47 (1 H, dt, J 2.8, 7.6, ArH), 7.34 (1 H, dt, J 2.8, 8.8, ArH), 2.62 (3 H, s, CH₃); ¹³C NMR δ_C (CDCl₃) 161.2 (CF, J 244), 158.7 (CF, J 239), 146.4 (C), 137.9 (C), 130.7 (CH), 130.2 (CH), 129.1 (C, J 8), 125.5 (C, J 3), 124.5 (C, J 8), 123.3 (CH), 123.2 (CH), 120.2 (CH, J 9), 119.3 (CH, J 9), 118.5 (CH, J 28), 117.2 (CH, J 24), 116.4 (C, J 11), 109.1 (CH, J 24), 104.5 (CH, J 25), 30.0 (C), 22.0 (CH₃); m/z (ES-HRMS) 319.1006 (M + 1). Calcd for C₂₀H₁₃F₂N₂ (M + 1) 319.1047.

2-Bromo-6-(4-fluoroanilino)-4-methylbenzonitrile 14

2,6-Dibromo-4-methylbenzonitrile 13¹⁸ (60 g, 0.22 mol), 4-fluoroaniline (27.2 g, 0.24 mol), cesium carbonate (100 g, 0.28 mol), BINAP (2.7 g, 4.3 mmol) and palladium acetate (0.47 g, 2.1 mmol) were dissolved in toluene (800 mL) under nitrogen and heated at 100 °C for 12 h. The solvent was then removed under vacuum and the residue purified by column chromatography, initially eluting with 20% chloroform–hexane to remove unreacted 13, increasing to 50% chloroform–hexane to elute 14, and finally neat chloroform to elute 15. Benzonitrile 14 was obtained as white needles (43 g, 65%), mp 175–176 °C (Found: C, 54.9; H, 3.3; N, 9.1%. C₁₄H₁₀BrFN₂ requires C, 55.1; H, 3.3; N, 9.2%); ν_max/cm⁻¹ 3315 (NH), 2223 (C≡N), 1608, 1568, 1506, 1210, 820; ¹H NMR δ_H (CDCl₃) 7.21–7.10 (4 H, m, ArH), 6.89 (1 H, d, J 0.4, H-3), 6.67 (1 H, d, J 0.4, H-5), 6.28 (1 H, br s, NH), 2.25 (3 H, s, CH₃); m/z (APCI) 305.0/307.0 (M + 1). The di(4-fluoroanilino)benzonitrile 15 was obtained as a white solid (14.6 g, 20%), mp 170–171 °C (Found: C, 71.5; H, 4.5; N, 12.5%. C₂₀H₁₅F₂N₃ requires C, 71.6; H, 4.5; N, 12.5%); ν_max/cm⁻¹ 3321 (NH), 2201 (C≡N), 1607, 1578, 1511, 1463, 1219, 819; ¹H NMR δ_H (CDCl₃) 7.17 (4 H, dd, J 2.4, 6.8, ArH), 7.06 (4 H, dd, J 2.4, 6.8, ArH), 6.18 (2 H, d, J 0.2, ArH), 6.08 (1 H, br s, NH), 2.11 (3 H, s, CH₃); m/z (ES-HRMS) 336.1298 (M + 1). Calcd for C₂₀H₁₆F₂N₃ (M + 1) 336.1312.

1-Bromo-7-fluoro-3-methylacridin-9(10H)-one 16

Benzonitrile 14 (40 g, 0.13 mol) was suspended in 80% sulfuric acid (200 mL) and heated at 100 °C for 15 h. After cooling, the reaction mixture was quenched on ice (800 g). The resulting precipitate was collected, washed with water (200 mL) and allowed to dry. The yellow solid 16 (34 g, 85%) was used in the next step without further purification. Physical characteristics of 16: ν_max/cm⁻¹ 3394 (NH), 1629 (CO), 1517, 1205, 838; ¹H NMR δ_H ([²H₆]DMSO) 11.81 (1 H, br s, NH), 7.81 (1 H, dd, J 4.0, 12.0, ArH), 7.64 (1 H, ddd, J 4.0, 8.0, 12.0, ArH), 7.55 (1 H, dd, J 4.0, 8.0, ArH), 7.33 (1 H, d, J, ArH), 7.28 (1 H, d, J, ArH), 2.41 (3 H, s, CH₃); m/z (ES) 305.0/307.0 (M + 1).

1-Bromo-7-fluoro-3,10-dimethylacridin-9(10H)-one 17

1-Bromo-7-fluoro-3-methylacridin-9(10H)-one 16 (30 g, 0.1 mol) was dissolved in DMF (300 mL) and added dropwise over 30 min to a suspension of sodium hydride (2.6 g, 0.11 mol) in DMF (150 mL). After stirring at room temperature for 60 min, dimethyl sulfate (18.5 g, 0.147 mol) was added and stirring continued for a further 1 h. The reaction mixture was poured carefully into water (1 L) and the precipitate collected and dried. The resulting yellow solid was purified by column chromatography (chloroform) to give 17 as a bright yellow powder (25 g, 80%), mp 262–265 °C (Found: C, 56.3; H, 3.4; N, 4.4%. C₁₅H₁₁BrFNO requires C, 56.3; H, 3.5; N, 4.4%); ν_max/cm⁻¹ 1637, 1604, 1501, 1271, 815; ¹H NMR δ_H [²H₆]DMSO 7.83–7.88 (2 H, m, H-5, H-8), 7.69 (1 H, dt, J 9.6, 3.2, H-6), 7.62 (1 H, d, J 0.8, H-2), 7.40 (1 H, d, J 0.8, H-4), 3.87 (3 H, s, NCH₃), 2.45 (3 H, s, CH₃); m/z (ES) 319.7/321.7 (M + 1).

2-(5-Fluoro-2-nitrophenyl)-4,4,5,5-tetramethyl-1,3,2-dioxaborolane 20

Potassium acetate (6 g, 0.06 mol), bis(pinacolato)diborane (5.8 g, 0.022 mol) and dichloro[1,1′-bis(diphenylphosphino)ferrocene]palladium(II) (0.9 g, 3 mol%) were added to a solution of 5-fluoro-2-nitrophenyltrifluoromethane sulfonate²¹ (5.9 g, 0.02 mol) in dioxane (60 mL) under a nitrogen atmosphere. The resulting solution was stirred at 80 °C for 20 h. After cooling, the solvent was removed under reduced pressure and the residue purified by column chromatography (dichloromethane) to give 20 as a pale cream solid (3.8 g, 70%), mp 79–80 °C (from MeOH) (Found: C, 53.9; H, 5.6; N, 5.1%. C₁₂H₁₅BFNO₄ requires C, 53.9; H, 5.6; N, 5.2%); ν_max/cm⁻¹ 1578, 1526, 1411, 1346; ¹H NMR δ_H (CDCl₃) 8.21 (1 H, dd, J 4.8, 10.0, H-3), 7.17–7.28 (2 H, m, H-4, H-6), 1.43 (12 H, s, 4 × CH₃).

7-Fluoro-1-(5-fluoro-2-nitrophenyl)-3,10-dimethylacridin-9(10H)-one 21

Dichloro[1,1′-bis(diphenylphosphino)ferrocene]palladium(II) (90 mg, 10 mol%) was added to a solution of 1-bromo-7-fluoro-3,10-dimethylacridin-9(10H)-one 17 (0.36 g, 1.1 mmol) and dioxaborolane 20 (0.45 g, 1.7 mmol) in dioxane (10 mL) under nitrogen. 2 N sodium carbonate (2 mL) was added and the resulting solution heated at 80 °C for 24 h. The solvent was removed under vacuum and the residue purified by column chromatography (chloroform followed by dichloromethane) to give 21 as a yellow solid (0.2 g, 47%), mp 238–241 °C (Found: C, 66.0; H, 3.7; N, 7.1%. C₂₁H₁₄F₂N₂ O₃ requires C, 66.3; H, 3.7; N, 7.4%); ν_max/cm⁻¹ 1607, 1582, 1507, 1358, 1262, 1209; ¹H NMR δ_H (CDCl₃) 8.29 (1 H, dd, J 4.8, 8.8, ArH), 7.99 (1 H, dd, J 3.2, 9.2, ArH), 7.52–7.55 (3 H, m, ArH), 7.21 (1 H, dt, J 2.8, 7.2, 9.2, ArH), 7.00 (1 H, dd, J 2.8, 8.4, ArH), 6.84 (1 H, d, J 1.2, ArH), 3.98 (3 H, s, NCH₃), 2.58 (3 H, s, CH₃); m/z (ES) 380.9 (M + 1).

3,11-Difluoro-6,8-dimethyl-8H-quino[4,3,2-kl]acridine 5

Tin(II) chloride dihydrate (1.6 g, 7.1 mmol) was added to a solution of acridinone 21 (0.92 g, 2.3 mmol) in ethanol (10 mL), and the resulting solution heated under reflux for 2 h. 5 M HCl (10 mL) was then added, and refluxing continued for a further 15 h. After cooling, the ethanol was removed under vacuum, and the aqueous phase basified with 5 M sodium hydroxide (to pH 12). The resulting precipitate was collected from solution and purified by column chromatography (1% methanol–chloroform) to leave 5 as an orange solid (0.59 g, 75%), mp 256–258 °C (Found: C, 76.2; H, 4.2; N, 8.1%. C₂₁H₁₄F₂N₂ requires C, 75.9; H, 4.3; N, 8.4%); ν_max/cm⁻¹ 1619, 1511, 1473, 1236, 1184, 796; ¹H NMR δ_H [²H₆]DMSO 8.35 (1 H, dd, J 3.2, 10.0, ArH), 8.30 (1 H, dd, J 2.8, 10.8, ArH), 7.93 (1 H, s, ArH), 7.90 (1 H, dd, J 6.0, 9.2, ArH), 7.58 (1 H, dd, J 4.4, 9.2, ArH), 7.54–7.44 (2 H, m, ArH), 7.21 (1 H, s, ArH), 3.66 (3 H, s, NCH₃), 2.67 (3 H, s, CH₃); m/z (ES-HRMS) 333.1189 (M + 1). Calcd for C₂₁H₁₄F₂N₂ (M + 1) 333.1203.

3,11-Difluoro-6,8,13-trimethyl-8H-quino[4,3,2-kl]acridinium methosulfate 3 by methylation of 5

Dimethyl sulfate (0.28 g, 2.26 mmol) was added to a solution of 3,11-difluoro-6,8-dimethyl-8H-quino[4,3,2-kl]acridine 5 (0.15 g, 0.45 mmol) in nitromethane (5 mL) and heated under reflux for 15 h. After cooling, the solvent was removed under vacuum and the residue triturated with ice cold acetone. The precipitated solid was filtered from solution and washed with a small amount of ice cold acetone to leave 3 as a dark red solid (0.1 g, 48%), identical (TLC, ¹H NMR and MS spectrum) to an authentic sample.⁸

One-pot synthesis of 3,11-difluoro-6,8,13-trimethyl-8H-quino[4,3,2-kl]-acridinium methosulfate 3 (Route C) from 1,2-dimethyl-6-fluoroquinolinium methosulfate 23

Triethylamine (10 g, 0.1 mol) was added to a solution of 23 (10 g, 0.035 mol) in nitrobenzene (100 mL) and heated at 120 °C for 24 h. After cooling, the solvent was removed by vacuum distillation and the residue purified by column chromatography (1-butanol–acetic acid–water: 0.6 : 0.15 : 0.25) to leave a dark red solid (2.63 g, 33%), identical (TLC, ¹H NMR and MS spectrum) to an authentic sample.⁸

Acknowledgements

We thank Cancer Research Campaign UK for supporting this work.

References

1. Part 18. J. C. Cookson, F. Dai, V. Smith, R. A. Heald, C. A. Laughton, M. F. G. Stevens and A. M. Burger, Mol. Pharmacol., 2005, 68, 1551–1558.
2. B. Fugman, B. Steffan and W. Steglich, Tetrahedron Lett., 1984, 25, 3575–3578; C. S. Hilger, B. Fugman and W. Steglich, Tetrahedron Lett., 1985, 26, 5975–78; J.-D Klamann, B. Fugman and W. Steglich, Phytochemistry, 1989, 28, 3519–3523.
3. D. J. Hagan, E. Giménez-Arnau, C. H. Schwalbe and M. F. G. Stevens, J. Chem. Soc., Perkin Trans. 1, 1997, 2739–2746; D. J. Hagan, D. Chan, C. H. Schwalbe and M. F. G. Stevens, J. Chem. Soc., Perkin Trans. 1, 1998, 915–924; M. Julino and M. F. G. Stevens, J. Chem. Soc., Perkin Trans. 1, 1998, 1677–1684; M. J. Ellis and M. F. G. Stevens, J. Chem. Soc., Perkin Trans. 1, 2001, 3174–3179; M. J. Ellis and M. F. G. Stevens, J. Chem. Res., Synop., 2003, 75–77; I. Hutchinson, A. J. McCarroll, R. A. Heald and M. F. G. Stevens, Org. Biomol. Chem., 2004, 2, 220–228.
4. M. J. Ellis and M. F. G. Stevens, J. Chem. Soc., Perkin Trans. 1, 2001, 3180–3185.
5. R. A. Heald and M. F. G. Stevens, Org. Biomol. Chem., 2003, 1, 3377–3389.
6. J. Stanslas, D. J. Hagan, M. J. Ellis, C. Turner, J. Carmichael, W. Ward, T. R. Hammonds and M. F. G. Stevens, J. Med. Chem., 2000, 43, 1563–1572.
7. C. E. Bostock-Smith, E. Giménez-Arnau, S. Missailidis, C. A. Laughton, M. F. G. Stevens and M. S. Searle, Biochemistry, 1999, 38, 6723–6731.
8. R. A. Heald, C. Modi, J. C. Cookson, I. Hutchinson, C. A. Laughton, S. M. Gowan, L. R. Kelland and M. F. G. Stevens, J. Med. Chem., 2002, 45, 590–597.
9. E. Gavathiotis, R. A. Heald, M. F. G. Stevens and M. S. Searle, Angew. Chem., Int. Ed., 2001, 40, 4749–4751; E. Gavathiotis, R. A. Heald, M. F. G. Stevens and M. S. Searle, J. Mol. Biol., 2003, 334, 25–36.
10. E. H. Blackburn, Nature, 2000, 408, 53–56.
11. L. R. Kelland, Eur. J. Cancer, 2005, 41, 971–979; J. W. Shay and W. E. Wright, Carcinogenesis, 2005, 26, 867–874; R. S. Maser and R. A. DePinho, Science, 2002, 297, 565–569; M. A. Blasco, Nat. Rev. Cancer, 2002, 2, 627–632; J. Marx, Science, 2002, 295, 2348–2351; S. Neidle and M. A. Reid, G-quadruplexes as therapeutic targets, Biopolymers, 2001, 56, 195–208; G. N. Parkinson, M. P. Lee and S. Neidle, Crystal structure of parallel quadruplexes from human telomeric DNA, Nature, 2002, 417, 876–880.
12. S. Gowan, R. Heald, M. F. G. Stevens and L. R. Kelland, Mol. Pharmacol., 2001, 60, 981–988.
13. C. Leonetti, S. Amodei, C. D'Angelo, A. Rizzo, B. Benassi, A. Antonelli, R. Elli, M. F. G. Stevens, M. D'Incalci, G. Zupi and A. Biroccio, Mol. Pharmacol., 2004, 66, 1138–1146.
14. J. C. Cookson, R. A. Heald and M. F. G. Stevens, J. Med. Chem., 2005, 48, 7198–7207.
15. D. Ozczapowicz, J. Jaroszewska-Manaj, E. Ciszak and M. Gdaniec, Tetrahedron, 1988, 44, 6645–6650.
16. A. Suzuki, Proc. Jpn. Acad., Ser. B, 2004, 80, 359–371.
17. I. M. McRobbie, O. Meth-Cohn and H. Suschitzky, Tetrahedron Lett., 1976, 17, 929–932.
18. K. Paek, C. B. Knobler, E. F. Maverick and D. J. Cram, J. Am. Chem. Soc., 1989, 111, 8662–8671.
19. A. R. Muci and S. L. Buchwald, Top. Curr. Chem., 2002, 219, 131–209.
20. F. Gullier, F. Nivoliers, A. Goddard, F. Marais, G. Quéguiner, M. A. Siddiqui and V. Snieckus, J. Org. Chem., 1995, 60, 292–296.
21. S. Hamaoka, N. Kitazawa, K. Nara, A. Sasaki, A. Kamada and T. Okabe, PCT Int. Appl., 2004, 982 pp., WO 2004058682.
22. T. Ishiyama, Y. Itoh, T. Kitano and N. Miyaura, Tetrahedron Lett., 1997, 38, 3447–3450.
23. J. Jaroszewska-Manaj, D. Maciejewska and I. Wawer, Magn. Reson. Chem., 2000, 38, 482–485.

---

Footnote

† Part 19 in the series ‘Antitumour Polycyclic Acridines.’ See ref. 1 for Part 18.

---