Diacrylamides as selective G-quadruplex ligands in in vitro and in vivo assays

Adam Le Gresley; Ammara Abdullah; Deepak Chawla; Pratchi Desai; Uttam Ghosh; Uma Gollapalli; Munazza Kiran; Shehri Lafon; Alex Sinclair

doi:10.1039/C1MD00020A

View PDF VersionPrevious ArticleNext Article

DOI: 10.1039/C1MD00020A (Concise Article) Med. Chem. Commun., 2011, 2, 466-470

Show CompoundsShow Chemical Terms

Diacrylamides as selective G-quadruplex ligands in in vitro and in vivo assays†

Adam Le Gresley *, Ammara Abdullah , Deepak Chawla , Pratchi Desai , Uttam Ghosh , Uma Gollapalli , Munazza Kiran , Shehri Lafon and Alex Sinclair
School of Pharmacy and Chemistry, Kingston University, Kingston-upon-Thames, KT1 2EE, UK. E-mail: a.legresley@kingston.ac.uk; Fax: +44 (0) 20 8417 7497; Tel: +44 (0) 20 8417 7432

Received 19th January 2011 , Accepted 7th March 2011

First published on 23rd March 2011

Abstract

The synthesis of previously unreported anthracene-9-monoacrylamides and anthracen-9,10-bisacrylamides and their potential as G-quadruplex ligands, based on preliminary fluorometric and biological assay data, is discussed.

The G-quadruplex can be described as the macromolecular structure formed when a guanine rich oligonucleotide sequence forms several Hoogsteen bonded G-tetrads.¹ These guanine rich sequences, at the ends of chromosomes, are termed telomeres and their depletion in normal healthy somatic cells has been linked to cellular senescence, whereas their regeneration, via the telomerase pathway, has been linked to cancer cell “immortality” in 90% of tumour cells.²

The last 10 years has seen interest in the exploitation of telomerase-mediated tumour cell immortality grow exponentially, with approaches to inducing telomere mediated apoptosis, including the development of oncolytic viruses, vaccines for human Telomerase RNA and stabilising of the G-quadruplex structure using ligands.³

The importance of the G-quadruplex structure, as an attractive target for anticancer treatment, has been reported in a number of detailed reviews and reports.⁴ A number of classes of G-quadruplex ligands, ranging from acridones to porphyrins, have been reported.^5–7,9 Owing to various issues of toxicity and efficacy, however, no G-quaduplex ligands have come to market despite clinical trials.⁸ In terms of rationalisation of G-quadruplex binding, some structure–activity relationships have already been identified using biological assays, NMR, UV-vis and CD spectroscopy. These observations indicate:¹⁰

• An extended planar structure is required to aid intercalation of the ligand in the stacked Hoogsteen bonded G-tetrads;

• The presence of cationic or hydrogen bonding groups is necessary to bind to the anionic phosphate backbone.

To increase the scope for a large and diverse range of ligands to be synthesised and screened, any new class of compound should be amenable to further reversible functionalisation, as exploited by Balasubramanian in dynamic combinatorial libraries.^10–12

In light of the trend for planar, nitrogen rich, species to have been shown to interact with quadruplex DNA, the anthracene diacrylamide core found in 4h not only possesses the correct geometry to sit between the G-tetrads but also can undergo reversible conjugate addition with a range of nucleophiles.¹³

Results and discussion

An anthracene unit possessing several hydrogen bond donors/acceptors was conceived as a potential G-quadruplex ligand. A series of anthracene diacrylamides, possessing various binding motifs, was synthesised in an effort to consider the viability of this class of compound as a source of G-quadruplex ligands and to evaluate the scope of the Heck reaction in generating a small library.

A further aim was to attempt to understand the steric and intermolecular bonding that may affect the activity of this class of ligand and how it may interact both with duplex and quadruplex DNA motifs. Surprisingly, neither the synthesis nor the evaluation of these diacrylamides has been reported until now.

Acrylamides

Acrylamides 1a–e, 1f, 1k and 1l were synthesised using the method summarised in the experimental section. These acrylamides were chosen to represent a broad range of polar/lipophilic binding motifs (Scheme 1).


	Scheme 1 The synthesis of the acrylamide precursors 1a–l.

Ligands

The acrylamides 1a–1l were coupled using a standard Heck procedure to give a series of anthracence diacrylamides (Table 1 and Scheme 2). Acrylamides 1b and 1d–g failed to undergo substantial reaction within a reasonable timeframe, however, the remaining diacrylamides were obtained in good yield as the β-trans geometric isomer. It is thought that complexation of the Pd(OAc)₂ catalyst, especially in the case of 1b, 1d, and 1f, could interfere with the reaction pathway. Coupling of 1c resulted in a mixture of 3c and 4c, which could not be separated.

Table 1 Reaction data for acrylamides 1a–l in coupling reactions with 2 in a 8 [thin space (1/6-em)]

1 ratio

Acrylamide	Product (ratio)	Duration/h	Yield (%)
a Ratio of acrlyamide to 2 is 2:1. b Ratio of acrylamide to 2 is 6:1.
1a	4a	72	58
1b	—	150	—
1c	3c/4c (4:1)	150	5 (crude)
1d	—	150	—
1e	—	150	—
1f	—	150	—
1g	3g/4g (10:1)	150	5 (crude)
1h	3h ^a	96	70
1h	4h ^b	96	80
1i	4i	124	74
1j	3j ^a	72	55
1j	4j	110	81
1k	4k	118	69
1l	4l	72	72


	Scheme 2 Synthesis of mono/di acrylamide derivatives 3a–l and 4a–l.

Efforts to synthesise the N,N-dimethyl acrylamide derivative yielded poor results with trace quantities of 3g evidenced by ¹H NMR. A 2 [thin space (1/6-em)] :1 ratio of 1h:2 was shown to favour the formation of the monocoupled product 3h and it was necessary to use a significant excess (6:1) of acrylamide to achieve the di-coupled product 4h.

The equivalent reaction to generate 4k, 4a, 4i and 4l demanded a significant excess of acryloyl morpholine (8 [thin space (1/6-em)] :1), however, the good yield of 3j and 4j obtained indicated the reaction is not restricted to primary acrylamides despite longer reaction times being required in the case of the tertiary acrylamide.

Fluorescence assay

The binding affinities of potential ligands 3j, 4h–j and 4l to quadruplex (22AG) as opposed to duplex (ds26) DNA were compared, using a recently reported technique.¹⁴

This technique involves the displacement of thiazole orange from duplex and quadruplex forming DNA strands ds26 and 22AG resp.† The degree of displacement of the thiazole orange from the two different DNA types provides a simple, useful tool to evaluate potential ligands in cacodylate-buffered solution. Berberine was used as an authentic reference samples to validate the experimental K_D values. As shown in Fig. 1, the greatest degree of selectivity is apparent for compound 4h with compounds 4l and 3j having a small preference for quadruplex DNA.


	Fig. 1 Thiazole orange displacement curves for compounds 3j, 4h-j and 4l. Error bars represent the standard deviation of the mean.

The K_D values were obtained from fluorescence data using curve-fitting software (Graphpad Prism 5) and the values for selectivity obtained for berberine agree reasonably well with the literature values (Table 2). However, it is worthy of note that the degree of complexation has been shown to be dependent on the ratio of ligand to duplex/quadruplex DNA.¹⁵

Table 2 Tabulated selectivity data and fitted dissociation constants for compounds 3j, 4h–j and 4l. K_D Error calculated from Graphpad 5 curve-fitting software

Compound	3j	4h	4i	4j	4l	Berberine
a Ratio based of ligand concentration required to displace thiazole orange from 22AG to the same extent as percentage displacement from ds26 at 2.5 μM of compound. Error is calculated at 5%.¹ b Reliable curve fitting was not possible in this case.
Selectivity^a	5.0	33.3	0.0	0.0	5.6	0.1
K_D ds26/μM	1.76 (± 0.13)	2.09 (± 0.13)	22.60 (± 1.8)	0.29 (±0.03)	0.36 (± 0.08)	1.77 (± 0.07)
K_D 22AG/μM	0.17 (± 0.05)	0.03(± 0.09)	Ambiguous^b	20.15 (± 1.4)	0.07 (± 0.02)	216.40(± 1.5)

In the case of 4h, the K_D for binding to duplex DNA is two orders of magnitude greater than that for quadruplex DNA and possesses considerably greater selectivity for the quadruplex DNA when compared with 4l and 3j. The selectivity values for these diacrylamides vary considerably, however, they compare well with the range of quadruplex/duplex selectivities reported for quinacridines, N-methylated quinacridines, bisquinolinium and metallo-organic G-quadruplex ligands (5–27 quadruplex/duplex selectivity). It is further worthy of note than neither 4i nor 4j show any selectivity according to the assay and this will prove useful in discounting certain functional groups or geometries when designing the next group of compounds based on the anthracene core.

Biological activity

Based on the fluorescence assay data together with preliminary molecular modelling data (using Chem3D energy minimisation) and Log P determinations (around 5 in both cases), compounds 4h and 4j were selected for biological testing. By selecting the two compounds that showed substantially different selectivity for the quadruplex DNA, the aim was to observe any obvious variation in cytotoxicity.

Evaluation of 4h for cytotoxicity in MCF-7 and Caco-2 cell culture, using a MTT cell viability assay, shows no appreciable apoptosis after 48 h. A 10 day evaluation of 4h shows an IC₅₀ value of 2.5μM and over 80% loss of viability over the 10 day period. Interestingly, the MTT viability assay of 4j in MCF-7 culture shows a rapid loss of cell viability over only 24 h, with a 40% loss at concentrations as low as 2.5μM (Figures 2–4).


	Fig. 2 Dose response curves for MCF-7 cell line at 24 h, 5 days and 10 days for 4h in (PBS/0.1% v/v DMSO). Error bars represent the standard deviation of the mean.


	Fig. 3 Cell viability after 48 h for Caco-2 cell line for range of concentrations of 4h (PBS/0.1% v/v DMSO) when compared to treatment with actinomycin (10μM). Error bars represent the standard deviation of the mean.


	Fig. 4 Showing cell viability after 24 h for MCF-7 cell line for range of concentrations of 4j (PBS/0.1% v/v DMSO). Error bars represent the standard deviation of the mean.

The variation in the loss of viable cells between the two compounds suggests a differing mode of action and this is corroborated by the fluorescence assay data, which shows a lack of specificity for the G-quadruplex motif in vitro.

Clearly, further compounds based on this class should be developed and further evaluation of anti-proliferative as opposed to only cytotoxic effects needs to be carried out, however, the preliminary data indicates G-quadruplex stabilisation as a possible mode of action for 4h. Negligible cytotoxic effects were observed on control cells when compared to the MCF-7 cell line and the lack of immediate toxicity (as evidence for Caco-2 cell lines) suggests this class of ligand would benefit from further investigation.¹⁴

NMR

The in vitro formation of a tetramolecular G–quadruplex using a 7-mer strand TTGGGGT (Biomers, GmbH) has recently been reported and has been visualised using ¹H NMR in the presence and absence of a putative ligand.¹³ In order to further validate 4h as a potential G-quadruplex ligand a similar experiment was carried out.†

A 1mM concentration of TTGGGGT oligonucleotide was shown not to form a quadruplex in PBS D₂O/d₆-DMSO (9 [thin space (1/6-em)] :1) over 48 h. However, as evidenced by ¹H NMR, a signal correlating to the tetramolecular quadruplex in the presence of 4h was observed after only a few minutes, suggesting that 4h can induce and stabilise this tetramolecular quadruplex.^13,14 This was not observed in the case of 4j.

Conclusions

We have conducted preliminary optimisation of the conditions for these C–C coupling reactions and have fully characterised 8 previously unreported compounds that have the potential to afford a route to a range of conjugate addition products for the development of a G-quadruplex-templated dynamic combinatorial library.^{4,11,12,16,17} The promotion of quadruplex DNA formation, as observed in solution by NMR, coupled with compelling fluorescence data and initial biological testing, suggests the class of G-quadruplex ligand based on 4h possesses an anticancer activity, which renders this family of compounds worthy of further investigation in terms of both duplex/quadruplex selectivity, bioavailability and expansion of the series to better understand any QSAR that may exist. We are currently undertaking work in this area.

¹H, ¹³C, COSY and HETCOR NMR spectra were recorded on a JEOL ECP 400MHz. Chemical shifts are reported as δ values in ppm relative to TMS (δ 0.00). All coupling constants are quoted in Hz. Elemental analysis were made on a Leeman Labs CE440 Elemental Analyzer. Fluorescence experiments were carried out on a Perkin Elmer LS55 Luminescence Spectrometer with the HPLC purified ds26 and 22AG strands supplied by Biomers GmbH – all fluorescence experiments were performed five times after equilibration.

Infrared spectra were determined on a ThermoNicolet 380 FT-IR. The mass spectra (m/z) were recorded on a Varian CP-3800 Gas Chromatograph with Varian 1200L Quadrupole Mass Spectrometer.

Dissociation constants were calculated using the curve-fitting software Graphpad Prism 5. Acrylamides 1g–j were obtained from Sigma-Aldrich and used without further purification. All other chemicals were purchased from the Aldrich Chemical Company.

General preparation of acrylamide precursor

N-(Quinolin-6-yl)acrylamide (1c). Et₃N (8.23 ml, 59 mmol) and the amine (27 mmol) were stirred in DCM (75 ml) for 10 min at rt. Acryloyl chloride (2.66 ml, 33 mmol) was very slowly added with the temperature being maintained at room temperature. The reaction was monitored viaTLC or NMR. Upon completion, the reaction was quenched with iced water (200 ml) and the organic component extracted with DCM (3 × 50 ml). The combined organic extracts were washed with HCl (3 × 40 ml, 3 M), water (2 × 50 ml) NaOH (3 × 30 ml, 2 M) and again with water (2 × 50 ml). The organic extract was dried (Na₂SO₄), filtered and evaporated under a reduced pressure. Solid acrylamides could be recrystallised from hot methanol and all acrylamides were stored below 0 °C or used immediately. Yields 70% and above - characterisation data for 1cδ_H(400 MHz; d₆-DMSO) 10.47 (1H, s (broad), NH), 8.79 (1H, dd, ³J_HH = 2.5 Hz, ⁴J_HH = 1.0, Ar-H), 8.46 (1H, d, ³J_HH = 2.3 Hz, Ar-H), 8.29 (1H, dd, ³J_HH = 7.8 Hz, ⁴J_HH = 1.0 Hz, Ar-H), 7.99 (1H, d, ³J_HH = 9.0 Hz, Ar-H), 7.88 (1H, dd, ³J_HH = 9.0 Hz, ⁴J_HH = 2.3 Hz, Ar-H), 7.49 (1H, dd, ³J_HH = 7.8 Hz, ⁴J_HH = 2.5 Hz, Ar-H), 6.52 (1H, dd, ³J_HH = 17.0 Hz, ⁴J_HH = 10.0 Hz, COCH), 6.35 (1H, dd, ³J_HH = 17.0 Hz, ⁴J_HH = 2.0 Hz, CH_aH_b), 5.81 (1H, dd, ³J_HH = 10.0 Hz, ⁴J_HH = 2.0 Hz, CH_aH_b), δ_C (d₆-DMSO) 164.1, 149.7, 145.4, 137.5, 136.1, 132.3, 130.1, 128.9, 127.8, 124.0, 122.4, 115.9; m/z (EI) 198[M+].

Example preparation of acrylamide

3-[10-(2-Carbamoyl-vinyl)-anthracen-9-yl]-acrylamide (4h). Triethylamine (6.6 ml, 4.81 g, 47.6 mmol) and acrylamide (3.38 g, 47.6 mmol) were added to a stirred solution of 9,10-dibromo anthracene (2.00 g, 5.9 mmol) in dimethylformamide (50 ml), accompanied by vigorous stirring, at room temperature followed by addition of palladium acetate (0.27 g, 1.2 mmol) and 1,3-bis(diphenylphosphino)propane (0.49 g, 1.2 mmol). The mixture was heated to 90 °C for 48 h under an atmosphere of nitrogen. The reaction was allowed to cool and dilute HCl (15 ml, 2M) was added. The resulting precipitate was filtered and washed with water (3 × 20 ml) and diethyl ether (3 × 15 ml) and dried under vacuum before being recrystallised from ethanol to afford the diacrylamide 4h4h (1.49 g, 80%) as a yellow powder; mp- 338.2 °C; ν_max (Nujol)/cm⁻¹ 3375 (N–H), 3200 (N–H), 1663 (C [double bond, length as m-dash]

O), 1600 (C [double bond, length as m-dash]

C); δ_H(400 MHz; d₆-DMSO) 8.61 (4H, s (broad), NH₂), 8.27 (2H, d, ³J_HH = 15.8 Hz, Ar–CH=), 8.23 (4H, dd, ³J_HH = 6.3 Hz, ⁴J_HH = 2.8 Hz, Ar-H), 7.60 (4H, dd, ³J_HH = 6.3 Hz, ⁴J_HH = 2.8 Hz, Ar-H), 6.41 (2H, d, ³J_HH = 15.8 Hz, =CHCO), δ_C (d₆-DMSO) 166.0, 158.2, 149.4, 143.2, 142.2, 126.6, 125.8, 125.5, 123.4, 121.8, 121.4, 116.1, 72.4, 32.3, 31.4, 31.0, 19.3, 14.1; m/z (EI) 316[M+]. Found C, 75.49; H, 5.51%; N, 8.91. C₂₀H₁₆N₂O₂ requires C, 75.93%; H, 5.10%; N, 8.85%.

Notes and references

A. De Cian, L. Lacroix, C. Douarre, N. Temime-Smaali, C. Trentesaux, C. Riou and J. Mergny, Biochimie, 2008, 90, 131 CrossRef CAS.
A. Zahler, J. Williamson, T. Cech and D. Prescott, Nature, 1991, 350, 718 CrossRef CAS.
A. Siddiqui-Jain, C. L. Grand, D. J. Bearss and L. H. Hurley, Proc. Natl. Acad. Sci. U. S. A., 2002, 99, 11593 CrossRef CAS.
M. Bejugam, S. Sewitz, P. S. Shirude, R. Rodriguez, R. Shahid and S. Balasubramanian, J. Am. Chem. Soc., 2007, 4j9, 12926 CrossRef.
M. Read, R. Harrison, B. Romagnoli, F. A. Tanious, S. H. Gowan, A. Reszka, W. D. Wilson, L. Kelland and S. Neidle, Proc. Natl. Acad. Sci. U. S. A., 2001, 98, 4844 CrossRef CAS.
J. Reed, M. Gunaratnam, M. Beltran, A. Reszka, R. Vilar and S. Neidle, Anal. Biochem., 2008, 380, 99 CrossRef CAS.
A. De Cian, G. Cristofar, P. Reichenbach, E. Lemos, D. Monchaud, M. Teulade-Fichou, K. Shin-Ya, L. Lacroix, J. Lingner and J.-L. Mergny, Proc. Natl. Acad. Sci. U. S. A., 2007, 104, 17347 CrossRef.
S. Balasubramanian and S. Neidle, Curr. Opin. Chem. Biol., 2009, 13, 345–353 CrossRef CAS.
A. Burger, Cancer Lett., 2007, 245, 11 CrossRef CAS.
J. Redman, J. M. Granadino-Roldan, J. Schouten, S. Ladame, A. Reszka, S. Neidle and S. Balasubramanian, Org. Biomol. Chem., 2009, 7, 76 RSC.
P. Corbett, J. Leclaire, L. Vial, K. R. West, J. L. Wietor, J. K. M. Sanders and S. Otto, Chemistry Reviews, 2006, 106, 3711 Search PubMed.
N. Shi and M. Greaney, Chem. Commun., 2005, 886 RSC.
Q. Zhou, L. Li, J. Xiang, H. Sun and Y. Tang, Biochimie, 2009, 91, 304 CrossRef CAS.
D. Monchaud, C. Allain, H. Bertrand, N. Smargiasso, F. Rosu, V. Gabelica, A. De Cian, J.-L. Mergny and M. P. Teulade-Fichou, Biochimie, 2008, 90, 1207 CrossRef CAS.
M. Franceschin, L. Rossetti, A. D'Ambrosio, S. Schirripa, A. Bianco, G. Ortaggi, M. Savino, C. Schultes and S. Neidle, Bioorg. Med. Chem. Lett., 2006, 16, 1707 CrossRef CAS.
I. Huc and J.-M. Lehn, Proc. Natl. Acad. Sci. U. S. A., 1997, 94, 2106 CrossRef CAS.
J.-M. Lehn, Chem.–Eur. J., 1999, 5, 2455 CrossRef CAS.

Footnote

† Electronic supplementary information (ESI) available: HETCOR, 2D COSY NMR for 4h, full characterisation data for all synthesised compounds. G-quadruplex formation ¹H NMR spectrum, fluorescence assay experimental and general biological experimental. It is worthy of note that for our system a 10 min equilibration period was required for the blank fluorescent readings for the DNA-thiazole orange complex to become stable. See DOI: 10.1039/c1md00020a

Click here to see how this site uses Cookies. View our privacy policy here.