Screening of indeno[1,2-b]indoloquinones by MALDI-MS: a new set of potential CDC25 phosphatase inhibitors brought to light

Abstract

Quinones and quinones-like compounds are potential candidates for the inhibition of CDC25 phosphatases. The combination of MALDI-MS analyses and biological studies was used to develop a rapid screening of a targeted library of indeno[1,2-b]indoloquinone derivatives. The screening protocol using MALDI-TOFMS and MALDI-FTICRMS highlighted four new promising candidates. Biological investigations showed that only compounds 5c–f inhibited CDC25A and -C phosphatases, with IC₅₀ values around the micromolar range. The direct use of a screening method based on MALDI-MS technology allowed achieving fast scaffold identification of a new class of potent inhibitors of CDC25 phosphatases. These four molecules appeared as novel molecules of a new class of CDC25 inhibitors. Assessment of 5c–e in an MRC5 proliferation assay provided an early indicator of toxicity to mammalian cells. Compound 5d seems the most promising hit for developing new CDC25 inhibitors.

Keywords:

• CDC25 inhibitors
• indeno[1,2-b]indoloquinones
• irreversibility
• MALDI mass spectrometry

Introduction

Cell division cycle 25 phosphatases (A, B and C isoforms) are fundamental regulators of cell-cycle progression, driving each transition of the cycle1–3. CDC25 phosphatases are overexpressed in various cancers3,4, and many clinical studies emphasize the correlation between this overexpression and tumor aggressiveness, high-grade tumors and low vital prognosis for the patient5,6. The overexpression of CDC25 isoforms lead to checkpoint bypass and aberrant division of the cells. Thus, CDC25s represent very attractive but challenging targets for anticancer therapy. The literature currently references over one hundred CDC25 inhibitors from various chemical classes (electrophilic inhibitors, thiophene derivatives, quinonoids, phosphate surrogates, etc.)7. Some were tested on mice xenografted with breast cancer cells8–10, inducing the arrest of the tumor growth or the reduction of the tumor size. However, no inhibitor successfully passed the preclinical stage, mainly due to their organotoxicity. Then, it is of worthy interest to find new potent CDC25 inhibitors.

Quinones and quinones-like compounds seem to be very promising candidates for the inhibition of CDC25 phosphatases. They can be either reversible inhibitors such as the 2,5-dihydroxy-3-(indol-3-yl)benzoquinone11 and SV3712 or irreversible inhibitors such as BN820028 or NSC66328413 which is able to bind covalently the catalytic domain of the CDC25A. Evidencing new potential inhibitors by the mean of biological tests is expensive and time-consuming. In 2012, inspired by studies carried out by Hannewald et al. on tubulin14 and DHFR15, Sibille et al.16 proposed a rapid approach for screening CDC25 inhibitors using MALDI mass spectrometry. Based on the target-ligand interactions in which CDC25s are the target, this method allows highlighting the most promising candidates from a set of molecules. However, Sibille et al.16 completed the screening procedure with an additional step able to answer the question about the reversibility of the inhibition. Nevertheless, complementary biological tests, for instance cytotoxicity and IC₅₀, have to be carried out obviously but only for the most interesting candidates. A recent study17 confirms the great interest of this approach.

Indenoindoles18 and more precisely indeno[1,2-b]indoles have been recently studied to design bioactive molecules. For example, diverse subseries of indeno[1,2-b]indoles were developed as inhibitors of protein kinase CK219,20 and inhibitors of the breast cancer resistance protein ABCG221. In this paper, we selected compounds belonging to this family of indeno[1,2-b]indoles (see Figure 1). Assuming that a quinonic pattern may lead to interesting candidates22,23, we assessed nine compounds belonging to the subseries of indeno[1,2-b]indoloquinones19 as potential inhibitors for CDC25 phosphatases.

Figure 1. Exploration of the indeno[1,2-b]indoloquinone for the development of new CDC25 phosphatases.

Experimental section

General

Commercial reagents and starting materials were purchased from Sigma-Aldrich, Acros, or Alfa Aesar and were used without further purification. Melting points were measured in capillary tubes using a BUCHI 510 apparatus and were uncorrected. IR spectra were recorded on a Perkin Elmer 1310 spectrometer and a Spectrum One spectrometer using KBr pellets (ν cm⁻¹). ¹H NMR and ¹³C NMR spectra (broadband decoupling and DEPT-135) were recorded on a Bruker Avance 400 (400 MHz for ¹H and 100 MHz for ¹³C) or a Bruker Avance 500 spectrometer (500 MHz for ¹H and 125 MHz for ¹³C) using CDCl₃ or DMSO-d₆ as solvents. NMR analyses of compounds 4–7 were performed with the same experiments described by Alchab et al24. Chemical shifts (δ) are referred to that of the solvent. Low resolution mass spectra were recorded on an Agilent 1290 Infinity system equipped with an Agilent 1260 DAD detector and an Agilent 6120 Quadrupole mass detector with an ESI source in positive mode. HRMS spectra were performed on a Q-Tof Micro Waters with an ESI source in positive mode. Flash chromatography was performed on 230–400-mesh silica.

Ultrapure water was obtained from a Simplicity Personal Ultrapure water system (Millipore, Molsheim, France). High-performance liquid chromatography (HPLC)-grade ethanol was purchased from Sigma Chemicals (St. Louis, MO). Trypsin was purchased from Promega Corporation (Madison, WI). Bradykinin, insulin, ACTH, P14R, human Angyotensin II and 2,5-dihydroxybenzoic acid (DHB) were purchased from Sigma-Aldrich (Saint Quentin Fallavier, France). The SV37 was produced according to that illustrated by Bana et al12.

Chemistry

3-Isopropylamino-5,6-dimethyl-cyclohex-2-enone (1h)

Isopropylamine (2.48 g, 41.97 mmol) and 4,5-dimethyl-cyclohexane-1,3-dione (5.77 g, 41.97 mmol) were heated in 60 mL toluene at reflux for 6 h. Solvent was removed in vacuum and the crude product was treated with ethyl acetate to give the title compound as a yellow powder. Yield 39%. Mp 136 °C. IR (KBr v_max cm⁻¹) 3272 (N–H), 1574 (C=O). ¹H NMR (DMSO-d_6, δ ppm, J Hz) 6.74 (d, J = 5.1, 1H, N–H), 4.78 (s, 1H, H-2), 3.47 (m, 1H, CHMe₂), 2.30 (dd, 1H, J = 2.7 and 12.3, H-5), 2.10 (dd, 1H, J = 7.8 and 12.3, H-6), 1.60–1.77 (m, 2H, H-4), 1.10 (d, J = 2.4, 3H, CH₃), 1.09 (d, J = 2.4, 3H, CH₃), 1.00 (d, 3H, J = 5.1, NCHMe₂), 0.99 (d, 3H, J = 5.1, NCHMe₂). ¹³C NMR (DMSO-d₆, δ ppm) 195.9 (C-1), 161.3 (C-3), 93.7 (C-2), 46.0 (C-5), 42.9 (C-6), 36.3 (C-4), 34.9 (NCHMe₂), 21.6 (CH₃), 21.5 (CH₃), 19.8 (CH₃), 13.1 (CH₃).

3-Isopropylamino-6-methyl-5-oxo-cyclohex-3-enecarboxylic acid ethyl ester (1i)

Isopropylamine (0.741 g, 12.53 mmol) and 2-methyl-3,5-dioxo-cyclohexanecarboxylic acid ethyl ester (3 g, 12.53 mmol) were heated in toluene (40 mL) at reflux for 6 h. The title product was recovered as a yellow liquid after treatment with ethyl acetate. It was used for the next step without regioisomers separation.Yield 49%. IR (KBr v_max cm⁻¹) 3251 (N–H), 1731 (C=O_ester), 1551 (C=O). MS-ESI⁺(m/z): 240.2 [M + H]⁺.

4b,9b-Dihydroxy-5-isopropyl-7,8-dimethyl-4b,5,6,7,8,9b-hexahydro-indeno[1,2-b]indole-9,10-dione (2h)

Enaminone 1h (3.53 g, 10.35 mmol) and ninhydrin (1.84 g, 10.35 mmol) were introduced in 25 mL of methanol and the reaction mixture was stirred at room temperature for 22 h, during which time the reaction was monitored by TLC (CH₂Cl₂/acetone 1:2). Solvent is removed in vacuum and the crude product was treated with ethyl ether to give the title product as a yellow powder after filtration and drying. Yield 33%. Mp 193 °C. IR (KBr v_max cm⁻¹) 3215 (OH), 3174 (OH), 1722 (C=O). ¹H NMR (DMSO-d_6, δ ppm, J Hz) 7.97 (m, 1H, H-1), 7.83 (m, 1H, H-4), 7.73 (m, 1H, H-3), 7.61 (m, 1H, H-2), 6.79 and 6.69 (2 s, 1H, OH), 5.67 (s, 1H, OH), 4.61 (m, 1H, NCHMe₂), 2.73 (m, 1H, H-7 or H-8), 2.49–2.15 (m, 2H, H-6), 1.83 (m, 1H, H-7 or H-8), 1.49 (m, 3H, CH₃), 1.29 (m, 3H, CH₃), 0.92 (m, 6H, NCHMe₂). ¹³C NMR (DMSO-d₆, δ ppm) 198.3 and 198.0 and 197.8 and 197.7 (C-10), 192.4 and 191.9 and 191.2 and 190.5 (C-9), 162.4 and 163.4 and 163.3 (C-5a), 148.8 and 148.6 and 148.49 and 148.3 (C-4a), 135.9 and 135.8 and 135.8 (C-3), 135.3 and 135.1 and 135.0 (C-10a), 130.6 and 130.5 and 130.5 (C-2), 125.1 and 124.9 and 124.8 and 124.7 (C-4), 123.7 and 123.5 and 123.4 (C-1), 104.4 and 104.0 and 103.9 and 103.7 (C-9a), 96.2 and 96.2 and 96.1 and 96.0 (C-4b), 83.6 and 83.5 and 83.4 and 83.3 (C-9b), 47.0 and 46.4 (NCHMe₂), 45.3 and 45.2 and 45.1 and 45.1 (C-8), 36.1 and 35.6 (C-7), 33.4 and 32.8 and 32.4 and 31.6 (C-6), 23.2 and 23.1 and 23.1 and 23.0 and 22.8 and 22.7 and 22.7 (2 CH₃), 20.2 and 20.1 (CH₃), 16.5 and 15.4 and 14.0 and 13.1 (CH₃). MS-ESI⁺(m/z) 705.4 [2M + Na]⁺, 364.2 [M + Na]⁺, 342.2 [M + H]⁺.

4b,9b-Dihydroxy-5-isopropyl-8-methyl-9,10-dioxo-4b,5,6,7,8,9,9b,10-octahydro-indeno[1,2-b]indole-7-carboxylic acid ethyl ester (2i)

Enaminone 1i (4 g, 16.71 mmol) and ninhydrin (2.97 g, 16.71 mmol) were dissolved in 50 mL of methanol and the reaction mixture was stirred at room temperature for 21 h, during which time the reaction was monitored by TLC (CH₂Cl₂/acetone 1:2). Solvent is removed in vacuum and the crude product was treated with ethyl ether to give the title product as a yellow powder after filtration and drying. Yield 30%. Mp 120 °C. IR (KBr v_max cm⁻¹) 3393 (OH), 3260 (OH), 1721 (C=O). ¹H NMR (DMSO-d_6, δ ppm, J Hz) 7.95 (d, 1H, J = 7.7, H-1), 7.80 (m, 1H, H-4), 7.69 (m, 1H, H-3), 7.58 (m, 1H, H-2), 6.83 and 6.82 (2 s, 1H, OH), 5.68 (s, 1H, OH), 4.60 (m, 1H, NCHMe₂), 4.09 and 3.84 (2 m, 2H, CH₂CH₃), 2.95 (m, 1H, H-7), 2.85–2.62 (m, 2H, H-6), 2.41 (m, 1H, H-8), 1.47 (m, 3H, CH₃), 1.27 (m, 3H, CH₃), 1.19 and 0.8 (t, 3H, J = 7, CH₂CH₃), 1.06 and 0.95 (d, J = 7.1, 3H, CH₃). 13C NMR (DMSO-d₆, δ ppm) 197.4 and 197.2 (C-10), 188.4 and 188.3 (C-9), 172.7 and 172.4 (C=O ester), 163.7 and 161.7 (C-5a), 148.2 and 147.9 (C-4a), 135.5 and 135.3 (C-3), 134.7 and 134–6 (C-10a), 130.1 and 130.1 (C-2), 124.5 (C-1), 123.2 and 123.1 (C-4), 103.4 and 103.4 (C-9a), 95.8 and 95.8 (C-4b), 60.4 and 60.1 (CH₂CH₃), 45.6 and 45.5 (NCHMe₂), 44.8 and 44.8 (C-7), 41.6 and 41.5 (C-8), 24.7 and 24.2 (C-6), 22.6 and 22.4 (CH₃), 22.3 and 22.1 (CH₃), 14.8 and 14.5 (CH₃), 14.0 and 13.5 (CH₃). MS-ESI⁺(m/z) 422.1 [M + Na]⁺, 400.2 [M + H]⁺.

5-Isopropyl-7,8-dimethyl-5,6,7,8-tetrahydro-indeno[1,2-b]indole-9,10-dione (3h)

Compound 2h (2.5 g, 7.32 mmol) was dissolved in DMF (11.288 mL) and acetic acid (2.25 mL) was added. (i-Pr₂N)₂SO (TIPTA) (4.54 g, 18.30 mmol) was then added; a precipitate appeared after 22 h. Then, the mixture was poured into ice water (200 mL). After 1.5 h the precipitate was recovered by filtration to obtain 3h as an orange powder. Yield 13%. Mp 230 °C. IR (KBr v_max cm⁻¹) 1699 (C=O), 1666 (C=O). ¹H NMR (DMSO-d_6, δ ppm, J Hz) 7.33–7.33 (m, 2H, 2 Ar-H), 7.3 (d, 1H, J = 7.0, Ar-H), 7.2 (m, 1H, Ar-H), 4.7 (m, 1H, NCHMe₂), 3.07 (dd, 1H, J = 4.0 and J = 16.8, H-8), 2.63 (dd, 1H, J = 9.9 and J = 16,8, H-7), 2.11 (m, 1H, H-6), 1.96 (m, 1H, H-6), 1.59–1.54 (m, 6H, NCHMe₂), 1.11 (d, 3H, J = 6.4, CH_3–7 or CH_3–8), 1.08 (d, 3H, J = 6.5, CH_3–8 or CH_3–7). ¹³C NMR (DMSO-d₆, δ ppm) 193.0 (C-9), 183.4 (C-10), 151.5 (C-4b), 148.9 (C-5a), 138,1 (C-4a), 134.7 (C-10a), 132.8 (C-3), 128.1 (C-2), 122.8 (C-1), 119.4 (C-4 and C-9b), 115.9 (C-9a), 49.4 (N-CHMe₂), 47.2 (C-8), 36.2 (C-7), 30.5 (C-6), 21.4 (CH₃), 21.2 (CH₃), 19.7 (CH₃), 12.3 (CH₃). MS-ESI⁺(m/z) 637.3 [2M + Na]⁺, 330.2 [M + H]⁺. HRMS-ESI⁺(m/z) [M + Na⁺] calcd for C₂₀H₂₁ N Na O₂ 330.1465; found 330.1464.

5-Isopropyl-8-methyl-9,10-dioxo-5,6,7,8,9,10-hexahydro-indeno[1,2-b]indole-7-carboxylic acid ethyl ester (3i)

Compound 2i (3 g, 7.51 mmol) was dissolved in DMF (9.6 mL) and acetic acid (1.92 mL) was added. TIPTA was then added (4.66 g, 18.77 mmol); a precipitate appeared after 18 h. Then, the mixture was poured into ice water (200 mL). After 1.5 h, the precipitate was recovered by filtration to obtain 3i as an orange powder. Yield 44%. Mp 153 °C. IR (KBr v_max cm⁻¹) 1736 (C=O ester), 1701 (C=O), 1670 (C=O). ¹H NMR (CDCl_3, δ ppm, J Hz) 7.45 (m, 1H, Ar-H), 7.23 (m, 1H, Ar-H), 7.14–7.1 (m, 2H, 2 Ar-H), 4.6 (m, 1H, NCHMe₂), 4.18 (q, J = 7.1, 2H, CH₂CH₃), 3.31–1.72 (m, 4H, H-7, H-8 and 2 H-6), 1.66–1.63 (m, 6H, NCHMe₂), 1.29–1.12 (m, 6H, CH_3–8 and CH₂CH₃). ¹³C NMR (CDCl₃, δ ppm) 193.1 and 191.5 (C-9), 184.1 and 184.0 (C-10), 172.9 and 172.1 (C=O ester), 152.2 (C-4b), 145.8 and 145.4 (C-5a), 138.76 (C-4a), 135.2 (C-10a), 132.2 and 132.1(C-3), 128.3 and 128.2 (C-2), 123.9 (C-1), 121,0 (C-9b), 118.7 (C-4), 116,5 and 115,8 (C-9a), 61.3 and 61.1 (CH₂CH₃), 47.8 (N–CHMe₂), 44.6 (C-8 or C-7), 43,0 and 42.8 (C-8 or C-7), 25.3 (C-6), 22.0 and 21.9 (CH₃), 21.8 and 21.5 (CH₃), 14.2 and 14.1 (CH₃), 13.7 (CH₃). MS-ESI⁺(m/z): 753.4 [2M + Na]⁺, 388.1 [M + Na]⁺, 366.2 [M + H]⁺. HRMS-ESI⁺(m/z) [M + Na⁺] calcd for C₂₂H₂₄ N Na O₄ 388.1519; found 388.1519.

9-Hydroxy-5-isopropyl-7,8-dimethyl-5H-indeno[1,2-b]indol-10-one (4h)

Compound 3h (0.4 g, 1.30 mmol) was dissolved in diphenyl ether (7 mL) and Pd/C (0.4 g) was added. The reaction mixture was refluxed for 2 h, during this time the reaction was monitored by TLC (ethyl acetate/cyclohexane 1:2). After cooling, methanol (50 mL) was added. The mixture was then filtrated on Celite and the solution concentrated. The obtained dark red oil was purified by flash chromatography on silica gel (ethyl acetate/cyclohexane 1:2). The title compound was obtained as a dark red powder. Yield 50%. Mp 195 °C. IR (KBr v_max cm⁻¹) 3418 (OH), 1660 (C=O). ¹H NMR (CDCl_3, δ ppm, J Hz) 7.37 (d, 1H, J = 7.0, Ar–H), 7.22 (m, 1H, Ar–H), 7.16–7.10 (m, 2H, Ar–H), 6.75 (s, 2H, OH and H-6), 4.79 (sept, 1H, J = 7.0, NCHMe₂), 2.34 (s, 3H, CH₃), 2.20 (s, 3H, CH₃), 1.69–1.70 (d, 6H, J = 7.1, NCHMe₂). ¹³C NMR (CDCl₃, δ ppm) 185.9 (C-10), 154.9 (C-4b), 147.4 (C-9), 140.8 (C-5a), 140.4 (C-10a), 136.5 (C-4a), 134.8 (C-7), 132.0 (C-3), 129.0 (C-2), 123.2 (C-1), 119.6 (C-4), 116.2 (C-9b or C-8), 115.4 (C-8 or C-9b), 111.7 (C-9a), 105.4 (C-6), 49.5 (NCHMe₂), 21.8 (CH₃), 21.2 (CH₃), 11.0 (2CH₃). HRMS-ESI⁺(m/z) [M + H⁺] calcd for C₂₀H₂₀ N O₂ 306.1489 found 306.1496.

9-Hydroxy-5-isopropyl-8-methyl-10-oxo-5,10-dihydro-indeno[1,2-b]indole-7-carboxylic acid ethyl ester (4i)

Compound 3i (1 g, 2.73 mmol) was dissolved in diphenyl ether (15 mL), and Pd/C (1 g) was added. The reaction mixture was refluxed for 2 h, during this time the reaction was monitored by TLC (ethyl acetate/cyclohexane 1:2). After cooling, methanol (50 mL) was added. The mixture was then filtrated on Celite and the solution concentrated. The obtained dark red oil is purified by flash chromatography on silica gel (ethyl acetate/cyclohexane 1:2). The title compound was obtained as a dark red powder. Yield 20%. Mp 222 °C. IR (KBr v_max cm⁻¹) 3422 (OH), 1713 (C=O), 1660 (C=O), 1633 (C=O). ¹H NMR (CDCl_3, δ ppm, J Hz) 7.54 (s, 1H, H-6), 7.36 (d, 1H, J = 7. 0, Ar-H), 7.27–7.26 (m, 2H, Ar-H), 7.16 (m, 1H, Ar-H), 6.9 (br s, OH), 4.87 (sept, 1H, J = 7.0, NCHMe₂), 4.39 (q, 2H, CH₂CH₃), 2.46 (s, 3H, CH_3–8), 1.73 (d, 6H, J = 7.1, NCHMe₂), 1.43 (t, 3H, J = 7.1, CH₂CH₃). MS-ESI⁺(m/z) 7493 [2M + Na]⁺, 386.1 [M + Na]⁺, 364.2 [M + H]⁺. HRMS-ESI⁺(m/z) [M + H]⁺calcd for C₂₂H₂₂NO₄ 364.1543; found 364.1541.

5-Isopropyl-7,8-dimethyl-5H-indeno[1,2-b]indole-6,9,10-trione (5h)

Compound 4h (0.1 g, 0.32 mmol) and Co-salen (0.005 g, 0.017 mmol) were dissolved in DMF (5 mL) at room temperature. The reaction was carried out under oxygen for 43 h. The reaction mixture was then poured into ice water to afford after filtration the crude product. The filtrate was extracted with diethyl ether and combined with the precipitate. The title product was finally purified by flash chromatography on silica gel (ethyl acetate/cyclohexane 1:2) and obtained as a red powder. Yield 97%. Mp 287 °C. IR (KBr v_max cm⁻¹) 1719 (C=O), 1631 (C=O), 1606 (C=O). ¹H NMR (CDCl_3, δ ppm, J Hz) 7.58 (m, 1H, H-1), 7.43–7.37 (m, 2H, Ar-H), 7.27 (m, 1H, Ar-H), 5.82 (br s, 1H, NCHMe₂), 2.06–2.05 (m, 6H, 2 CH₃), 1.67 (d, 6H, J = 7.0, NCHMe₂). ¹³C NMR (CDCl₃, δ ppm) 183.6 (C-10), 181.6 (C-9), 178.6 (C-6), 155.4 (C-4b), 142.9 (C-5a), 141.2 (C-10a), 140.0 (C-8 or C-7), 139.9 (C-7 or C-8), 134.3 (C-4a), 133.9 (C-9b), 133.1 (C-4), 129.6 (C-2), 124.4 (C-1 or C-9a), 124.3 (C-9a or C-1), 121.30 (C-3), 50.1 (NCHMe₂), 20.9 (2CH₃), 12.6 (CH₃), 12.3 (CH₃). MS-ESI⁺(m/z) 661.3 [2 M + Na]⁺, 342.1 [M + Na]⁺, 320.2 [M + H]⁺. HRMS-ESI⁺(m/z) [M + H]⁺calcd for C₂₀H₁₈NO₃ 320.1281; found 320.1292.

5-Isopropyl-8-methyl-6,9,10-trioxo-5,6,9,10-tetrahydro-indeno[1,2-b]indole-7-carboxylic acid ethyl ester (5i)

According to the oxidative method using PIFA, 4i (0.2 g, 0.55 mmol, 1 eq) in solution in a acetonitrile/water 2:1 mixture (15 mL) was added dropwise at 0 °C to a flask containing PIFA (0.47 g, 1.10 mmol, 2 eq) solubilized in 15 mL of the same solvent mixture. The reaction was stirred at room temperature for 2 h then extracted with CH₂Cl₂. The title product was purified by flash chromatography on silica gel (ethyl acetate/cyclohexane 1:2), and obtained as a red powder. Yield 32%. Mp 166 °C. IR (KBr v_max cm⁻¹): 1733 (C=O), 1731 (C=O), 1646 (C=O). ¹H NMR (CDCl_3, δ ppm, J Hz) 7.62 (d, 1H, J = 7.0, Ar-H), 7.46–7.41 (m, 2H, Ar-H), 7.32 (dt, 1H, J = 2 and 7.3, Ar-H), 5.74 (br s, 1H, NCHMe₂), 4.42 (q, 2H, J = 7.2, CH₂CH₃), 2.06 (s, 3H, CH_3–8), 1.67 (d, 6H, J = 7.1, NCHMe₂), 1.40 (t, 3H, J = 7.2, CH₂CH₃). ¹³C NMR (CDCl₃, δ ppm) 183.3 (C-10), 180.9 (C-9), 175.0 (C-6), 168.0 (COO), 155.1 (C-4b), 144.9 (C-5a), 140.9 (C-10a), 139.8 (C-7), 138.3 (C-8), 133.4 (C-4a), 133.3 (C-9b), 132.8 (C-4), 130.0 (C-2 or C-9a), 129.8 (C-9a or C-2), 124.7 (C-1), 121.6 (C-3), 62.1 (CH₂CH₃), 50.6 (NCHMe₂), 20.9 (2CH₃), 14.2 (CH₃), 13.0 (CH₃). MS-ESI⁺(m/z) 777 [2M + Na]⁺, 400 [M + Na]⁺, 378 [M + H]⁺. HRMS-ESI⁺(m/z) [M + Na⁺] calcd for C₂₂H₁₉ N Na O₅ 400.1155; found 400.1152.

5-Isopropyl-8-methyl-6,9,10-trioxo-5,6,9,10-tetrahydro-indeno[1,2-b]indole-7-carboxylic acid ethyl ester (5i)

According to the oxidative method using Co-salen, 4i (0.1 g, 0.32 mmol) and Co-salen (0.005 g, 0.017 mmol) were dissolved in DMF (5 mL) at room temperature. The reaction was carried out under oxygen for 43 h. The reaction mixture was then poured into ice water to afford after filtration the crude product. The filtrate was extracted with diethyl ether and combined with the precipitate. The title product was finally purified by flash chromatography on silica gel (ethyl acetate/cyclohexane 1:2) and obtained as a red powder. Yield 72%.

MALDI-MS analyses

MALDI-TOFMS and MALDI-FTICRMS analyses

MALDI-TOFMS analyses were carried out on a Bruker UltraFlex II TOF/TOF mass spectrometer (Bruker Daltonic, Bremen, Germany) equipped with a nitrogen pulsed laser (337 nm). The laser output energy was 400 μJ/pulse (19% of the laser maximum energy). Positive ions were accelerated without any extraction delay at 20 kV. The reflector voltage was 23 kV. Positives ions were accelerated with a 200 ns extraction delay. Mass spectra were manually acquired using FlexControl software (Bruker Daltonic) by accumulating four series of 100 laser shots. All of the depositions were performed using the dried droplet method with DHB as matrix.

MALDI-Fourier transform ion cyclotron resonance (FTICR)-MS analyses were carried out on an IonSpec mass spectrometer (Varian, Palo Alto, CA) fitted with a 9.4 T shielded magnet. The external ProMALDI source was equipped with an Nd:YAG laser at 355 nm (third frequency, New Wave Research, Fremont, CA). All mass spectra were manually acquired with Omega software in Wizard mode (Varian). Thirty five laser shots were averaged with laser energy at 17% of the laser output maximum.

By using DHB matrix peaks and a mixture of standard peptides (Bruker Daltonic), external calibration was done. The following monoisotopic peaks were used: m/z 155.034 ([DHB + H]⁺), m/z 757.400 (bradykinin), m/z 1046.542 (human angiotensin II), and m/z 1533.860 (P14R). The calibration was successful if the root mean square (RMS) error was in the range of ± 1 to 2 ppm and ± 10 to 15 ppm for FTICRMS and TOFMS, respectively. Digest mass spectra were internally calibrated with two trypsin autolytic peaks: m/z 842.510 (VATVSLPR) and m/z 2211.104 (LGEHNIDVLEGNEQFINAAK).

Screening protocol

The screening of CDC25 inhibitors was already described by Sibille et al16. The first step consisted in the incubation at 37 °C of CDC25A or -C with the evaluated compound at a final concentration of 2 × 10⁻³ M. CDC25 phosphatases were solubilized in Tris A buffer (50 mM Tris, 50 mM NaCl, 1 mM ethylenediaminetetraacetic acid [EDTA], and 1 mM dithiothreitol [DTT], pH 8.0) at a concentration of approximately 2 × 10⁻⁵ M for isoform A and 2.2 × 10⁻⁵ M for isoform C. Then, 30 μL of CDC25A or 50 μL of CDC25C was incubated with 10 μL of the evaluated compound for 1 h. The next step was a ultracentrifugation of the incubate for 8 min at 12 000 rpm using a Microcon® centrifugal filter unit (Millipore, Cork, Irland) with a mass cutoff of 30 kDa (Millipore). The retentate was then washed three times by adding 100 μL of ultrapure water to remove all unbound species. The retentate is dissolved with 20 μL of ultrapure water and desalted using ZipTip C18. Reversible inhibitors were highlighted by a directly analysis in MALDI-TOFMS of 1 μL of retentate with DHB as matrix. Irreversible inhibitors were highlighted by checking the absence of the active site in the peptide mass fingerprint (PMF) after the digestion the CDC25-ligand covalent complex by trypsin (V5111, Promega, Madison, WI). Experiments have been carried out in duplicate.

Biological studies

Production and purification of recombinant human CDC25

Recombinant human GST-CDC25A and -C proteins were produced as previously described by Brault et al25. Briefly, E. coli strain BL21-DE3 pLys S was transformed by a plasmidic vector (pGEX 2T, Amersham, Ge-Healthcare, Vélizy, France) containing the sequences encoding full length CDC25A and -C. Production of recombinant proteins was induced by addition of isopropyl-thio-β-galactoside (Sigma Aldrich, Saint Quentin Fallavier, France) in the medium. Cells were lysed and centrifuged to recover the supernatant which was purified with a GSH-agarose column system (Sigma Aldrich), and recombinant GST-CDC25 proteins were eluted and collected in fractions. Activity, purity and protein concentration of the fractions were evaluated.

Evaluation of CDC25 enzymatic activity

CDC25 phosphatases activity is evaluated in vitro by a fluorimetric method previously described by Valente et al26. Briefly, samples are prepared in 96-well plates with [50 mM Tris–HCl, 50 mM NaCl, 1 mM EDTA and 0.1% SAB, pH 8.1] buffer containing 3-O-methylfluorescein phosphate as substrate. After 2 h at 30 °C, 3-O-methylfluorescein fluorescent emission is measured with a CytoFluor system (Perspective Applied Biosystems; excitation filter: 475 nm; emission filter: 510 nm). Assays were performed in triplicate, and the experiment was independently performed three times. The results are expressed as residual activity of CDC25 phosphatase in percentage (compared to DMSO control) in the presence of the tested compound, naphtoquinone (20 μM) was used as a positive inhibition control.

Statistics and analytical models

CDC25 IC₅₀ values determined by in vitro fluorimetric assays were evaluated with sigmoid curves plotted by using a nonlinear approximation model based on the least square method (GraphPad Prism software, La Jolla, CA). Data are presented as mean of at least three independent experiments ± SD.

MRC5 proliferation assay

Cell viability is evaluated through the MTT colorimetric assay. The MTT assay is based on protocol described by Mossmann27. The assay was optimized for cell line used in the experiment. MRC5 cells (Lung Fibroblast Human CCL-171 ATCC) are plated in triplicated at a density of 6000 cells/well into 96-well culture plates. Cells are incubated overnight at 37 °C in 5% CO₂ in MEM (modified Eagle’s medium) media supplemented with 10% FBS (fetal bovine serum). The following day, cells are treated with the compounds (10–100 μM) or with vehicle control (dimethylsulfoxide, DMSO). After 72 h, the cells are incubated with 1 mg/mL of MTT (Sigma Aldrich) for 3 h at 37 °C. The medium is then removed and 100 μL of 0.01 M HCL in isopropanol is added in each well for 15 min. Absorbance is measured by a plate reader at 570 nm and the value measured at 690 nm was subtracted. Data are the mean ± SD of at least three independent experiments.

Results and discussion

Synthesis of the targeted indeno[1,2-b]indoloquinones

The general outline of the synthesis of the derivatives is shown in Scheme 1. Key enaminones 1 h,i were synthetized as previously described20,28. The synthesis of indeno[1,2-b]indoloquinones 5a–g (see also Table 1 for the nature of the substitution group R¹ and R²) have been previously described by Alchab et al28. Compounds 5 h,i were synthetized using salcomine (5 h,i) or PIFA (5i). All structures were ascertained by spectroscopic methods (¹H NMR, 13C NMR, IR, HRMS).

Scheme 1. Synthesis of indeno[1,2-b]indoloquinones 5a–i. Reagents and conditions: (a) MeOH, RT, 21–22 h; (b) TIPTA (1.5 eq.), DMF, AcOH, RT, 18–22 h; (c) Pd/C, diphenyl ether, reflux, 2 h; (d) salcomine (1.8 eq.), DMF, O₂, RT, 43 h or PIFA (2 eq.), acetonitrile/H₂O, RT, 2 h.

Table 1. Nature of the evaluated indeno[1,2-b]indoloquinones and results of the screening test by MALDI mass spectrometry. Results obtained with CDC25 isoforms A and C are the same.

			Analysis by MALDI-MS
Compounds	R^1	R^2	Detection with direct MS	Detection of the target active site	Conclusion of the test
5a	CH3	H	No	Yes	Not significant
5b	i-C3H7	H	No	Yes	Not significant
5c	4-F-C6H4	H	No	No	Irreversible
5d	2′-Furanyl	H	No	No	Irreversible
5e	–C6H5	H	No	No	Irreversible
5f	H	CH3	No	No	Irreversible
5g	H	i-C3H7	No	Yes	Not significant
5h	CH3	CH3	No	Yes	Not significant
5i	COOC2H5	CH3	No	Yes	Not significant

Screening by MALDI mass spectrometry

MALDI Mass spectrometry is used to determine whether a molecule has a chance to be a good candidate for the inhibition of CDC25s. This procedure requires two steps. The reversible binding mode of a given compound on CDC25A and -C proteins is evaluated by direct MALDI-TOFMS. Following the conclusion of Sibille et al.16, the direct detection of the candidate in the mass spectrum leads to the conclusion that this molecule binds the target but not covalently. Thus, it behaves as a reversible inhibitor. If nothing is detected by direct MS analysis, a second MS analysis is performed after the digestion of the target protein by trypsin. Then, a peptide mass fingerprint (PMF) of the digest is obtained by MALDI-FTICRMS. Combining the results of direct analysis and PMF, two cases can be distinguished:

1. The [M + H]⁺ ion corresponding to the active site of the CDC25A (m/z 1551.7624, VIVVFHCEFSSER) or -C (m/z 1579.794, IIIVFHCERFSSER) is observed in the PMF report (one missed cleavage of CDC25s by the trypsin enzyme is also accepted) leading to the conclusion that the molecule does not significantly bind the target. This molecule is not an interesting candidate for the inhibition of the target.

2. The active site does not appear in the PMF report leading to the conclusion that the molecule covalently binds the target. This compound behaves as an irreversible inhibitor.

The results of the MS screening procedure applied to the 9 indeno[1,2-b]indoles 5a to 5i are displayed in Table 1. Compounds 5a, 5b and 5 g to 5i were not observed by direct MALDI-TOFMS and the active site of the target proteins are observed in PMF leading to the conclusion that they does not significantly bind the target protein. These compounds are not interesting enough for further investigation. The case of compounds 5c–f is different. They do not appear in direct MS and the active sites of the target proteins are not observed in PFM when they have been incubated with CDC25A and -C. Thus, these compounds appear to be promising candidates and further investigation can be carried out. Although the MS screening test is unable to determine biological activity such as the IC₅₀ value, we can assume that their binding mode is covalent and they are good candidates to irreversibly inhibit CDC25s.

Correlation of MS results with biological tests and determination of IC₅₀ values of the most active compounds

The above described MS screening test allows a rapidly discrimination between potentially active compounds and the less active ones. Such a procedure is particularly interesting because the evaluation of inactive compounds by expensive and time-consuming procedures can be avoided. However, the MALDI-MS test is not quantitative and parameters such as IC₅₀ values cannot be deduced from the screening results. Consequently, the nine indeno[1,2-b]indole derivatives were all submitted to a test of residual enzymatic activity of CDC25A and -C phosphatases at 50 μM (concentration of incubation during the screening test). The results are displayed in Figure 2.

Figure 2. Residual enzymatic activity of CDC25A and -C phosphatases in the presence of compounds 5a–i at 50 μM (results expressed in % compared to the DMSO control).

Indenoindole compounds of the series are found to inhibit recombinant full length CDC25A and -C. The biochemical screening highlights four compounds with inhibitory activity up to 75% (residual enzymatic activity under 25%) at 50 μM: 5c, 5d, 5e and 5f which, noteworthy, also corresponds to the most interesting structures pointed out by mass spectrometry. Yet, the MS screening test fails to evidence that compounds such as 5 g or 5i which percentage of inhibition are 70% and 73%, respectively.

MALDI-TOFMS experiments show that none of these compounds behaves as a reversible inhibitor. Such results can be explained by the fact that the detection of the active sites remains possible if only more than 25% of these sites remain unbound to the inhibitor (these 25% constitute somehow their limits of detection in PFM). Thus, the test is able to discriminate compounds that are unable to bind the target protein as previously described, but also to discriminate the weakest inhibitors allowing staying focused on only the most active compounds.

As previously mentioned, the MALDI-MS screening method remains a nonquantitative approach and the IC₅₀ values have to be determined separately. IC₅₀ values (concentrations inhibiting 50% of enzyme activity) were determined for the four most active compounds 5c–f (Table 2). These results exhibit IC₅₀ values in the low micromolar range (IC₅₀=1.68–2.73 μM and 1.72–2.55 μM for CDC25A and -C, respectively) highlighting these compounds as inhibitors of CDC25 phosphatases.

Table 2. Evaluation of IC₅₀ values for compounds 5c–f.

		IC50 (μM)
Compound	Structure	CDC25A	CDC25C
5c		2.73 ± 0.79	2.55 ± 0.64
5d		1.68 ± 0.54	1.72 ± 0.11
5e		2.01 ± 0.29	2.00 ± 0.48
5f		2.00 ± 0.24	2.51 ± 0.51

MRC5 cells were used to evaluate cytotoxicity of compounds 5c–e. This evaluation is a good indicator of toxicity to mammalian cells29. Compound 5d showed the less toxicity against MRC5, with an IG₅₀ value of 85.5 μM (Table 3).

Table 3. Results on MRC5 cell line.

Compound	IG50 (μM)	TR
5c	<10
5d	85.5 ± 7.5	49.7
5e	<10

TR: therapeutic ratio.

Conclusions

MALDI-MS technology was integrated in this project as screening tool to accelerate the discovery of new hit/scaffolds for designing new CDC25 inhibitors. The combination of MALDI-MS analyses and biological studies leads to the identification of four indeno[1,2-b]indoloquinone derivatives as new CDC25 inhibitors. These four molecules constitute a new starting point for the development of small molecules targeting CDC25 phosphatases and especially compound 5d. Our MALDI-MS approach could also be applied to other small molecule libraries for discovering new scaffolds faster.

A recent work focused the particular interest to develop small molecule inhibitor targeting CDC25 dual-specificity phosphatases30. Further pharmacomodulation works on indeno[1,2-b]indoloquinone will lead to a new collection of inhibitors. The best ones will be tested on human malignant cells (melanoma cell lines30 A2058 and SAN, acute myeloid leukemia cell lines31 Molm13, MV4–11 and IPC-Bcl-2).

Declaration of interest

The authors report no conflicts of interest. The authors alone are responsible for the content and writing of this article.

We acknowledge financial support to Dr. Faten Alchab by the “Bonus Qualité Recherche” (BQR) of the University Claude Bernard Lyon 1, by the “Cluster 5 Chimie Durable et Chimie pour la Santé” of the Region Rhône-Alpes and by the “ARC 1 Santé” of the Region Rhône-Alpes. The “Institut des Sciences Pharmaceutiques et Biologiques” (ISPB) is also gratefully acknowledged for the funding of an uHPLC/DAD/MS system. Pr. Marc Le Borgne would like to thank “Cancéropôle Lyon Auvergne Rhône-Alpes” (CLARA) and “Université France Allemagne” (UFA) for their funding of ChemBioInteract network.

Pr. Patrick Chaimbault would like to thank the Region of Lorraine for funding.

Acknowledgements

Dr. Faten Alchab is grateful to Mrs Christine Ranquet for her assistance in the stock management of the small molecule library of EA 4446 B2C. We also thank ChemAxon for providing us a license to their cheminformatics software Instant JChem. Pr. Patrick Chaimbault would like to thank Mrs. Mariana Noronha for technical assistance.

This paper is dedicated to Professor Gilbert Kirsch without whom no chance that any of us would ever have met. We thank him for his tremendous energy and his enthusiasm.