Synthesis and evaluation of the antimicrobial activity of novel quinazolinones

Abstract

A simple and efficient microwave-assisted methodology for regioselective alkylation of exocyclic nitrogen of cyclic amidines was developed and novel N-alkylated 3,4-dihydropyrazino[2,1-b]quinazolin-6-ones were prepared. Although none of the molecules tested have any specific anti-quorum sensing (–QS) activity, our result validates the growth tests devised to control the bias of the anti-QS tests. Among the molecules studied, compound 2b exhibits interesting activity against the Gram-negative bacteria Escherichia coli and Shigella sonnei.

Keywords::

• Quinazolinones
• microwave-assisted chemistry
• quorum sensing
• anti-microbial activity

Introduction

The selective pressure exerted by antibiotic treatments has made many pathogenic bacteria multi-resistant to common antibiotics. New drugs are needed to resolve this serious public health problem. Searching for new biological targets, a special attention was recently paid to the bacterial communication system named quorum sensing (QS). The bacteria that possess a QS system release in their environment signal molecules called autoinducers. When a threshold autoinducer concentration is reached, all the producing bacteria trigger collectively the expression of target genes responsible for the production of virulence factors essential for the pathogenicity 1-3. Moreover, QS is dispensable for life functions, and it is well admitted that its specific inhibition does not implicate the development of bacterial resistances [4]. Molecules that inhibit QS have been called anti-pathogenic drugs since they can disturb the production of the virulence factors [5]. Recently, synthetic agents derived from algal halogenated furanones were shown to affect Pseudomonas aeruginosa pathogenicity in this way [6,7,8].

The occurrence of the quinazoline skeleton in various natural and synthetic products has generated interest of many groups because of their useful biological properties [9]. As a part of our ongoing research program dealing with the preparation and pharmacological evaluation of some original quinazoline derivatives 10-14, we recently described novel 3,4-dihydropyrazino[2,1-b]quinazolin-6-one derivatives (1a–d). During this preliminary work, we unambiguously established the 3D structure of compounds 1 and we confirmed the amidine isomerization in the solid state by X-ray crystallography.

After a rapid evaluation of these novel molecules on QS, it was suggested that compound 1a might possess a moderate anti-QS activity and that modification on the benzenic moiety of the quinazolinone ring seems to have no real incidence on the expected activity. Considering this 2,3-condensed quinazolin-4-one as a possible lead compound, we focused our efforts on the synthesis of various N-substituted amidines (e.g. 2 in Scheme 1) in which the substituent will be present specifically on the exogenic nitrogen atom.

Scheme 1 Structure of quinazolinones.

In the present work, we describe the synthetic route to the expected tricyclic compounds (3) and we first examine their anti-QS activities by optimising the protocol described by McLean et al [15]. We used three model Gram-negative bacteria that synthesise distinct acylated homoserine lactones (HSL) as autoinducer molecules: the opportunistic pathogen Chromobacterium violaceum synthesises a N-hexanoyl-homoserine lactone (C6-HSL) [16], the plant pathogen Agrobacterium tumefasciens a N-octanoyl-homoserine lactone (C8-HSL) [17], and the opportunistic pathogen P. aeruginosa two different HSLs, a N-butyrylhomoserine lactone (C4-HSL) and a N-3-oxododecanoyl-homoserine lactone (C12-oxo-HSL) [18]. Using our methodology, we are able to discriminate between the specific anti-QS activities and growth inhibitions. Finally, we characterised the anti-microbial properties of one quinazolinone derivative of interest.

Materials and methods

Chemistry

Instrumentation

Commercial reagents were used as received without additional purification. Melting points were determined using a Kofler melting point apparatus and were uncorrected. IR spectra were recorded on a Perkin-Elmer Paragon 1000PC instrument. ¹H and ¹³C-NMR were recorded using a JEOL NMR LA400 (400 MHz) spectrometer (Centre Commun d'Analyses, Université de la Rochelle). Chemical shifts (δ) were reported in part per million (ppm) downfield from tetramethylsilane (TMS) which was used as internal standard. Coupling constants J are given in Hz. The mass spectra (HRMS) were recorded on a Varian MAT311 spectrometer in the “Centre Régional de Mesures Physiques de l'Ouest” (CRMPO), Université de Rennes. Column chromatography experiments were performed by using Merck silica gel (70–230 mesh) at medium pressure. Light petroleum refers to the fraction boiling point 40–60°C. Other solvents were used without purification. Analytical thin layer chromatography (T.L.C.) was performed on Merck Kieselgel 60 F254 aluminium backed plates. Focused microwave irradiations were carried out with a Smith-Synthetizer™ (Personal Chemistry, AB) focused microwave reactor (300 W, 2450 MHz, monomode system). The Smith-Synthetizer™ (Personal Chemistry, AB) was a single mode cavity, producing controlled irradiation at 2450 MHz [19]. Reaction temperature and pressure were determined using the built-in, on-line IR and pressure sensors. Microwave-assisted reactions were performed in sealed Smith process vials (0.5–5 mL, total volume 10 mL) under air with magnetic stirring. The microwave output power was regulated by the software algorithm so that the selected maximum temperature was maintained for the desired reaction/irradiation time. After the irradiation period, the reaction vessel was cooled rapidly to ambient temperature by compressed air (gas-jet cooling). The minimal reaction times were determined by performing sequential series of identical reactions at constant temperature and with continuous heating, but with different irradiation times. Completion of the reaction was estimated by T.L.C. after each individual heating period.

Synthesis of 3,4-dihydropyrazino[2,1-b]quinazolin-6-one derivatives

Spectral data for compound 1a are consistent with results published in reference 14.

Synthesis of N-substituted quinazolinone derivatives

To a stirred suspension of quinazolinone 1a (0.2 mmol) and PTSA (0.22 mmol) in THF (3 mL) was added dropwise 0.4 mmol of alkylating agent. The mixture was irradiated for 5 min in a sealed tube. The irradiation was programmed to obtain a constant temperature (150°C). The solvent was removed under reduced pressure. Products 2a–e were obtained after purification by column chromatography with dichloromethane/methanol (99/1) as eluent.

1-Propylamino-3,4-dihydropyrazino[2,1-b]quinazolin-6-one (2a)

This compound was prepared from 1a.Yield: 95%, white solid, mp (84°C. (Found M⁺: 256.1323, C₁₄H₁₆N₄O requires 256.1324); υ_max (KBr)/cm^{− 1} 772, 1473, 1670, 2876, 2966, 3409 (NH); ¹H-NMR δ (400 MHz, d₆-DMSO) 0.90 (t, 3H, J 7.40 Hz, CH₃), 1.53–1.65 (m, 2H, CH₂–CH₂–CH₃), 3.18–3.24 (m, 2H, NH–CH₂–CH₂), 3.64 (t, 2H, J 6.00 Hz, CH₂), 4.00 (t, 2H, J 6.00 Hz, CH₂), 6.80 (s, 1H, NH), 7.59 (t, 1H, J 7.60 Hz, H arom.), 7.75 (d, 1H, J 7.60 Hz, H arom.), 7.86 (t, 1H, J 7.60 Hz, H arom.), 8.16 (d, 1H, J 7.60 Hz, H arom.); δ_C (100 MHz, d₆-DMSO) 11.83; 22.04; 42.55; 43.18; 121.56; 126.62; 128.16; 128.40; 135.21; 140.83; 146.29; 150.34; 160.45.

1-Isopropylamino-3,4-dihydropyrazino[2,1-b]quinazolin-6-one (2b)

This compound was prepared from 1a. Yield: 60%, white solid, mp (91°C. (Found M⁺: 256.1323, C₁₄H₁₆N₄O requires 256.1324); υ_max (KBr)/cm^{− 1} 692, 1243, 1474, 1671, 2972, 3398 (NH); ¹H-NMR δ (400 MHz, d₆-DMSO) 1.18 (d, 1H, 6.40 Hz, 2CH₃), 3.64 (t, 2H, J 6.40 Hz, CH₂), 3.93–4.01 (m, 3H, J 6.40 Hz, CH₂ and NH), 6.46 (d, 1H, J 8.40 Hz, CH), 7.59 (t, 1H, J 7.60 Hz, H arom.), 7.78 (d, 1H, J 8.00 Hz, H arom.), 7.87 (t, 1H, J 8.00 Hz, H arom.), 8.16 (d, 1H, J 8.00 Hz, H arom.); δ_C (100 MHz, d₆-DMSO) 22.14; 38.38; 41.54; 43.09; 121.38; 126.28; 127.69; 127.74; 134.62; 140.50; 145.99; 148.82; 159.80.

1-Butylamino-3,4-dihydropyrazino[2,1-b]quinazolin-6-one (2c)

This compound was prepared from 1a. Yield: 81%, white solid, mp (71°C. (Found M⁺: 270.1494, C₁₅H₁₈N₄O requires 270.1481); υ_max (KBr)/cm^{− 1} 774, 1330, 1682, 2860, 2925, 3421 (NH); ¹H-NMR δ (400 MHz, d₆-DMSO) 0.91 (t, 3H, J 7.10 Hz, CH₃), 1.25–1.40 (m, 2H, CH₂–CH₃), 1.50–1.60 (m, 2H, CH₂–CH₂–CH₂), 3.20–3.25 (m, 2H, NH–CH₂–CH₂), 3.60–3.65 (m, 2H, CH₂), 3.95–4.05 (m, 2H, CH₂), 6.75 (s, 1H, NH), 7.57 (t, 1H, J 8.00 Hz, H arom.), 7.73 (d, 1H, J 8.00 Hz, H arom.), 7.83 (t, 1H, J 8.00 Hz, H arom.), 8.11 (d, 1H, J 8.00 Hz, H arom.); δ_C (100 MHz, d₆-DMSO) 14.16; 20.24; 31.13; 38.88; 43.43; 121.76; 126.66; 128.10; 128.17; 135.03; 140.91; 146.42; 150.18; 160.23.

1-Benzylamino-3,4-dihydropyrazino[2,1-b]quinazolin-6-one (2d)

This compound was prepared from 1a. Yield: 66%, white solid, mp (110°C. (Found M⁺: 304.1333, C₁₈H₁₆N₄O requires 304.1324); υ_max (KBr)/cm^{− 1} 692, 774, 1474, 1586, 1672, 2858, 2938, 3406 (NH); ¹H-NMR δ (400 MHz, d₆-DMSOD₂O) 3.65 (t, 2H, J 6.00 Hz, CH₂), 4.01 (t, 2H, J 6.00 Hz, CH₂), 4.48 (d, 2H, J 6.00 Hz, CH₂–NH), 7.18–7.13 (m, 1H, H arom.), 7.30 (t, 2H, J 7.60 Hz, H arom.), 7.36 (d, 2H, J 7.60 Hz, H arom.), 7.56–7.63 (m, 1H, H arom.), 7.76 (d, 1H, J 7.60 Hz, H arom.), 7.83–7.90 (m, 1H, H arom.), 8.17 (dd, 1H, J 8.20 Hz et J 1.20 Hz, H arom.); δ_C (100 MHz, d₆-DMSO + D₂O) 38.27; 42.97; 43.65; 121.33; 126.13; 126.38; 127.10; 127.51; 127.58; 127.99; 134.42; 139.90; 140.33; 145.91; 149.56; 159.62.

1-[2-(1H-indol-3-yl)-ethylamino]- 3,4-dihydropyrazino[2,1-b]quinazolin-6-one (2e)

This compound was prepared from 1a. Yield: 58%, white solid, mp = 238°C. (Found M⁺: 357.1572, C₂₁H₁₉N₅O requires 357.1589); υ_max (KBr)/cm^{− 1} 1472, 1588, 1680, 2921, 3152, 3396 (NH); ¹H-NMR δ (400 MHz, CDCl₃) 3.16 (t, 2H, J 6.80 Hz, CH₂-indol), 3.68–3.76 (m, 2H, NH–CH₂), 3.82 (t, 2H, J 6.00 Hz, CH₂), 4.15 (t, 2H, J 6.00 Hz, CH₂), 6.47 (s, 1H, NH), 7.10–7.18 (m, 2H, H arom.), 7.12 (t, 1H, J 7.80 Hz, H arom.), 7.40 (d, 1H, J 8.00 Hz, H arom.), 7.53 (t, 1H, J 7.60 Hz, H arom.), 7.63 (d, 1H, J 8.00 Hz, H arom.), 7.70 (d, 1H, J 7.60 Hz, H arom.), 7.77 (t, 1H, J 7.60 Hz, H arom.), 8.53 (s, 1H, NH), 8.30 (d, 1H, J 8.00 Hz, H arom.).

1-Phenylamino-3,4-dihydropyrazino[2,1-b]quinazolin-6-one (2f)

In a sealed vial, a stirred mixture of quinazolinone (1a) (0.053 mg, 0.25 mmol) and aniline (0.50 mL, 0.45 mmol) in NMP (2 mL) was irradiated for 15 min. The irradiation was programmed to obtain a constant temperature (220°C). The solvent was removed under reduced pressure. The product was obtained by purification by column chromatography with dichloromethane/methanol (95/5) as eluent. Yield: 62%, white solid, mp (179°C. (Found (M-H⁺),: 289.1084, C₁₇H₁₄N₄O requires 289.1089); υ_max (KBr)/cm^{− 1} 1328, 1536, 1683, 2938, 3055, 3356 (NH); ¹H-NMR δ (400 MHz, CDCl₃) 3.96 (t, 2H, J 5.80 Hz, CH₂), 4.23 (t, 2H, J 5.80 Hz, CH₂), 7.06 (t, 1H, J 7.10 Hz, H arom.), 7.36 (t, 2H, J 7.60 Hz, H arom.), 7.55–7.63 (m, 1H, H arom.), 7.75–7.87 (m, 4H, H arom.), 8.34 (d, 1H, J 8.00 Hz, H arom.), 8.53 (s, 1H, NH); δ_C (100 MHz, CDCl₃) 38.22; 43.96; 119.04; 121.72; 122.69; 127.04; 127.93; 128.16; 128.95; 134.69; 139.28; 139.82; 145.88; 147.19; 160.62.

Pharmacology

Strains & growth conditions

The strains used for the anti-quorum sensing (–QS) tests are listed in Table I. Briefly, the biosensor strains C. violaceum CV017 and P. aeruginosa lasB-gfp(ASV) produce under QS activation a violet pigment (violacein) and a green fluorescent protein, respectively. A. tumefaciens NTLR4 is a mutant unable to synthesise its own autoinducer molecule (C8-HSL) but that produces the β-Galactosidase enzyme when QS is activated by exogenous C8-HSL. β-Galactosidase production was revealed by the appearance of a blue pigment in the presence of the X-Gal substrate. All the strains used in this study were grown aerobically on nutrient broth. Liquid cultures (5 mL) were performed in glass tubes under continuous stirring (160 rpm). A tumefasciens NTLR4 and P. aeruginosa lasB-gfp(ASV) constructions were maintained on nutrient agar 1.2% (w/v) supplemented with 25 μg/mL and 50 μg/mL of gentamycin (Gm) respectively. A. tumefasciens NTLR4 and C. violaceum CV017 were grown at 30 °C and P. aeruginosa lasB-gfp(ASV) at 37°C.

Table I.  Reporter strains used for the anti-quorum sensing (–QS) tests.

Strain	Description	Reference
Agrobacterium tumefasciens NTLR4	pTI cured, traG::lacZ fusion, Gm^r	Cha et al., 1998 [17]
Chromobacterium violaceum CV017	Smr mini-Tn5 Hgr, violacein overproducer	Winson et al., 1994 [22]
Pseudomonas aeruginosa PA01 lasB-gfp(ASV)	lasB::gfp(ASV) fusion,Gm^r	Hentzer et al., 2002 [5]

Anti-quorum sensing and anti-bacterial tests

Anti-quorum sensing (–QS) tests were systematically accompanied with anti-bacterial tests in order to confirm the effect of each molecule. For both tests, sterile glass wells (4 mm × 8 mm, Polylabo®) were placed on nutrient agar plates and pre-filled with 20 μL of sterilized milliQ water. This step was required for the good diffusion of the molecules added in the wells from stock solutions prepared in methanol. 20 μL of each stock solution (20 mM) were added to the wells. For diffusion and drying, plates were stored 3 h at ambient temperature. Wells were then removed and the plates were overlaid with 5 ml of soft agar (0,6% w/v) medium seeded with different volumes of stationary phase cultures of the indicator strains.

(i) For anti-QS tests, the overlays were seeded with 5 mL of the biosensors CV017 and NTLR4, and 5 μL of a 1:1000 dilution of PA01 lasB-gfp(ASV). The overlay seeded with NTLR4 was supplemented with 10 μL of C8-HSL (2 g/L) and 30 μL of X-Gal (20 g/L). Detection of the green fluorescent protein was performed using the Dark Reader® from Clare Chemical. (ii) For anti-bacterial tests, the overlays were seeded with 5 μL of stationary phase cultures diluted (1:10 for CV017 and NTLR4, and 1:1000 for PA01 lasB-gfp) or not (other strains). Using these inoculi, growth inhibition (partial or total) could be determined easier.

For both tests, plates were incubated overnight and analysed visually for the presence or not of halos of QS inhibition and/or growth inhibition. (i) A QS inhibition halo was defined as a clear zone where the biosensor does not produce its QS-dependent pigment or fluorescence, and (ii) a growth inhibition halo was a clear zone where growth was inhibited or absent.

Each molecule was tested in duplicate on three independent cultures.

Results and discussion

Chemistry

The synthesis of the 2,3-condensed (3H)-quinazolin-4-one precursor 1a was performed from methyl N-(4-chloro-5H-1,2,3-dithiazol-5-ylidene)-anthranilate itself obtained by condensation of 4,5-dichloro-1,2,3-dithiazolium chloride (Appel's salt) and methyl anthranilate (Scheme 2). Stirring of a solution of the intermediate imine and ethylenediamine at room temperature in tetrahydrofurane, for 1 or 2 h, gave a 58% yield of the expected 3,4-dihydropyrazino[2,1-b]quinazolin-6-one 1a (Scheme 2).

Scheme 2 Synthesis of 1a from methyl anthranilate.

The second step of the synthesis is an alkylation of the heterocyclic amidine present on the studied molecule. The usual methods for alkylation of this 1,3-dinucleophile usually afforded mixtures of both the N-endosubstituted and the N-exosubstituted products. Ten years ago Bourguignon and his co-workers have developed a regioselective alkylation of the exocyclic nitrogen atom via N-acyl derivatives and the method was successfully applied to the synthesis of analogs of chiral amino acids [20].

The method we developed is inspired by our previous studies of the Niementowski reaction for the synthesis of novel pentacyclic quinazolin-4-ones [13,14]. It consists of a strong microwave-assisted heating (5 min) of the reactants in a sealed tube, at 220°C in the presence of 10% by weight of graphite, one of the solids most efficiently heated by microwaves. In the case studied here, the use of graphite was not suitable since the association of the liquid reagents and graphite may create hot-spots at the surface of the reactor. After exploration of various experimental conditions, we found that microwave heating of a solution of the starting compound 1a and aliphatic amines in tetrahydrofurane (THF), in the presence of 1.1 equiv. of para-toluenesulfonic acid (PTSA), afforded good yields of the expected N-alkylated 3,4-dihydropyrazino[2,1-b]quinazolin-6-ones (Scheme 3).

Scheme 3 Conditions: PTSA (1.1 equiv.), THF, MW, 150°C, 5 min (for 2a–e); NMP, MW, 220°C, 15 min (for 2f).

The suggested mechanism of this reaction is an addition–elimination process in which the primary amine will first attack the carbon of the amidine function, and then, eliminate the ammoniac.

Applying the conditions described above with aromatic amines (e.g. aniline in Scheme 3) led to different results. The synthesis of the expected products implied a more intense heating (220°C) of the starting reactants in the presence of N-methylpyrrolidinone (NMP) a solvent which is particularly well adapted for microwave experiments (Scheme 3).

Here we developed a simple and efficient methodology for regioselective alkylation of exocyclic nitrogen of cyclic amidines. In our case microwave heating allows very short reaction times and clean conditions for work-up.

Pharmacology

Discrimination between anti-quorum sensing (–QS) and anti-bacterial activities is crucial to reject false anti-QS positives. As a matter of fact, if the drug added inhibits bacterial growth under the cell density threshold, QS will not be activated even though this drug does not possess anti-QS activity. We therefore optimised the methodology described by McLean et al [15] to discriminate between anti-QS and anti-bacterial activities (see Material & Methods). We first analysed the effects of 7 quinazolinone derivatives on three model Gram-negative bacteria. We found that 5 quinazolinones inhibited the production of the QS-dependent pigment violacein in the opportunistic pathogen C. violaceum CV017. These apparent effects were actually due to growth inhibitions (Table II).

Table II.  Effects of the quinazolinone derivatives synthesised on growth and quorum sensing of C. violaceum (CV017), A. tumefaciens (NTLR4), and P. aeruginosa [PA01 lasB-gfp(ASV)]. The symbol + indicates the presence of an inhibition halo (see Material & Methods).

	Quinazoline Derivatives
	1a	2a	2b	2c	2d	2e	2f
CV017 growth	+	+	+	+	+	0	0
CV017 quorum sensing	+	+	+	+	+	0	0
NTLR4 growth	0	0	+	0	0	0	0
NTLR4 quorum sensing	0	0	+	0	0	0	0
PA01 growth	0	0	0	0	0	0	0
PA01 quorum sensing	0	0	0	0	0	0	0

Consequently, none of the molecules tested have any specific anti-QS activity. However, this negative result validates the growth tests devised to control the bias of the anti-QS tests. None of the quinazolinones tested affect P. aeruginosa growth or QS and two of them had even no activity at all. Among the 5 quinazolinones that inhibit C. violaceum CV017 growth, the 2b derivative was also found to inhibit A. tumefaciens NTLR4 growth (Table II). The anti-bacterial potential of the 2b derivative was then analysed in more details using a panel of Gram-positive and Gram-negative bacteria. Results of Table III indicated that the 2b molecule had no effect on the Gram-positive bacteria tested. Conversely, this quinazolinone affected the growth of the Gram-negative bacteria Escherichia coli and Shigella sonnei. This is an attractive result since S. sonnei causes dysenteries (called shigellosis) hard to fight owing to multi-resistance phenomena [21]. Nevertheless, additional experiments are needed to determine whether 2b and other quinazolinone derivatives might be used to control these infections.

Table III.  Effect of the 2b derivative on growth of several Gram-positive and Gram-negative bacteria.

Strains	Origin	Inhibition
Gram-positive
Bacillus cereus ^(*)	Our laboratory	0
Bacillus megatherium	CIP 6620T	0
Bacillus subtilis ^(*)	CIP 5262	0
Enterococcus faecalis	CIP 103214	0
Lactobacillus acidophilus ^(*)	Our laboratory	0
Micrococcus luteus ^(*)	ATCC 4698	0
Staphylococcus aureus ^(*)	ATCC 25923	0
Staphylococcus saprophyticus	Our laboratory	0
Gram-negative
Citrobacter freundii	Our laboratory	0
Enterobacter cloacae	Clinical	0
Escherichia coli	Our laboratory	++
Escherichia coli K12	ATCC 23716	+
Hafnia alvei	CIP 5731T	0
Klebsiella pneumoniae subspecies. Pneumoniae	CIP 53153	0
Proteus mirabilis	Our laboratory	0
Salmonella anatum	CIP 5630	0
Salmonella enterica subspecies. Enterica serotype wien ^(*)	CIP 8122	0
Salmonella enteritidis	Clinical	0
Serratia marcescens	CIP 6755	0
Shigella flexeneri	CIP 5236	0
Shigella sonnei	CIP 5255	++

^(*) Incubated at 30°C instead of 37°C. The symbol +(+) indicates the presence of an inhibition halo (see Material & Methods)

In conclusion, developing a simple and very efficient microwave-assisted methodology for regioselective alkylation of exocyclic nitrogen of cyclic amidines, we prepared N-alkylated 3,4-dihydropyrazino[2,1-b]quinazolin-6-ones among which compound 2b exhibits interesting activity against the Gram-negative bacteria E. coli and S. sonnei. We believe that this family constitutes a scaffold from which more potent inhibitors could be designed, opening the door to various applications.

Acknowledgements

We thank the “Comité de Charente-Maritime de la Ligue Nationale Contre le Cancer” and the “Conseil Général de la Charente Maritime” for financial support ((M.-F. P. and R. C. PhD grant respectively). We are indebted to Pr Michael Givskov (Technical University of Denmark) for providing the PA01 lasB-gfp(ASV) strain. We are grateful to Dr Denis Faure, (Institut des Sciences du Végétal CNRS Gif sur Yvette France) for providing the Chromobacterium and the Agrobacterium strains.