Naphthalene-dioxygenase catalysed cis-dihydroxylation of bicyclic azaarenes

Abstract

The scope of the biotransformation of a series of azaarene compounds by recombinant whole-cells of Escherichia coli JM109(DE3)(pDTG141) expressing the naphthalene dioxygenase system (NDOS) from Pseudomonas sp. NCIB 9616-4 was explored. The present study establishes that several bicyclic azaarenes are good substrates in the NDO-catalysed reaction, giving cis-dihydrodiol derivatives; monohydroxylated products were also observed.

---

1. Introduction

The bacterial metabolism of aromatic compounds often proceed via an initial dihydroxylation catalyzed by Rieske non-heme iron dioxygenases to yield cis-dihydrodiol derivatives. These multicomponent bacterial systems, mainly toluene dioxygenase (TDOS), naphthalene dioxygenase (NDOS) and biphenyl dioxygenase (BPDOS), have been frequently utilized to produce, by environmentally friendly routes, hundreds of such cis-1,2-dihydrodiols, several of which have been used as chiral synthons for the synthesis of biologically active products and value-added chemicals.^1–3

In recent years, direct evidence for the cis-dihydroxylation of a 2-pyridone ring by NDO has been reported.^4–6 A major dihydroxylated metabolite, (5S,6S)-5,6-dihydroxy-1-methyl-5,6-dihydropyridin-2(1H)-one was obtained from the corresponding N-methyl-2-pyridone substrate, together with small amounts of its 3,4-dihydroxy isomer when E.coli JM109(DE3)(pDTG141) expressing the NDO system from Pseudomonas sp. NCIB 9616-4 (a strain developed by Gibson⁷) was used as a whole-cell biocatalyst. We have also shown that 1-methyl quinolone was transformed by this strain more efficiently than 1-methyl pyridone (62% transformation vs. 35–40%); probably due to a better accommodation of 1-methyl quinolone by the active site of NDO. Azaheterocyclic cis-dihydrodiols with a high enantiomeric purity are formed by this bacterial dioxygenase and their absolute configuration was directly and definitively determined.⁶ The effect of substituents on 2-pyridone and 2-quinolone rings in the NDO-catalysed reaction was also investigated.

The cis-dihydroxylation of bicyclic fused azaarenes (quinoline, isoquinoline, quinoxaline, quinazoline, substituted quinolines) has been reported, using the multicomponent bacterial toluene dioxygenase (TDOS).^8–10 More recently Boyd et al. have compared different types of dioxygenase enzymes (TDO, NDO and BPDO) for the regio- and stereo-selectivity of the cis-dihydroxylation of several monocyclic azaarene analogues to biphenyl and tricyclic azaarenes, showing that the larger active site of BPDO was more efficient to accommodate sterically demanding tricyclic azaphenantrenes.¹¹

Azaarenes (often quoted as NPAHs) are widely distributed throughout the global environment where they are persistent. Some compounds of this group exert toxic effects.¹² Quinoline is one of the most abundant and has been shown to be both carcinogenic and mutagenic.^13,14 Bioremediation is one of the strategies developed to rehabilitate polluted environments.^15,16 Several bacterial strains are able to metabolise some polycyclic azaarenes.¹⁷ Among them, Pseudomonas putida UV4, which expressed toluene dioxygenase (TDO),^9,10 has been shown to transform quinoline and other bicyclic azaarenes, but in a very low extent perhaps because of its better specificity toward monocyclic arenes. Nevertheless, higher yields were reported when using another (unspecified) strain of P. putida.⁸ As quinoline is almost isosteric to naphthalene, a bacterial strain expressing naphthalene dioxygenase should be a better choice. NDO-catalysed biotransformation of fused bicyclic azaarenes should be expected to occur in good yields, providing less toxic compounds as metabolites. From another point of view, quinoline or isoquinoline moieties are present in many classes of biologically-active compounds. New metabolites could give access to modifications of the quinoline or isoquinoline moieties which can trigger interesting biological activities.

In this study we have used E.coli JM109(DE3)(pDTG141) overexpressing the NDO system from Pseudomonas sp. NCIB 9616-4 for the oxidative biotransformation of fused bicyclic azaarenes reported in Scheme 1.

Scheme 1 Azaarenes used as substrates for the biotransformations and numbering.

2. Results

The previously used E. coli JM109(DE3)(pDTG141), a strain that overexpresses the NDO system from Pseudomonas sp. NCIB 9816-4, was used as a whole-cell biocatalyst. Biotransformation of each of the compounds (1–13) was conducted in the experimental conditions precedently optimised for 1-methyl pyridone.⁶ HPLC-ESIMS analysis (positive mode) of aliquots of the incubation mixtures withdrawn at intervals shows the initial substrate (M + H⁺), dihydrodiol(s) (M + 34 + H⁺) and/or monohydroxylated derivative(s) (M + 16 + H⁺) (see Table 1 and 2). Conversion yields were evaluated by comparing the integrated mass peak areas of the products to that of remaining substrate. The initial conversion rate remained in a lower range (1 : 2–1 : 60) compared to naphthalene dihydroxylation in the same conditions. The conversion yield (after 30 h) was in the 20–100% range (see Tables 1 and 2). Preparative experiments were then conducted with each substrate. The supernatants of incubation media were lyophilized and extraction with aceton of the dry powder gave the crude organic extracts. The oxidation products were isolated from the crude organic extract by flash chromatography on silicagel or a prepacked C18 column, and characterised by NMR analysis. The relative yield of isolated products was estimated from the ¹H NMR spectra of the crude organic extracts using integration of representative signals.

Table 1 Biotransformations of quinoline and isoquinoline derivatives by NDO system from Pseudomonas sp. NCIB 9816-4 (27 °C, 30 h)

Substrate	Total conversion (%) and relative yield of isolated products (%)			m/z expected	Identified major products
	NDO^a	TDO^b	P. putida (mutant)^c
a This work, E. Coli JM109(DE3)(pDTG141). b From ref. 9, P. putida UV4. c From ref. 8, mutant of P. putida.
	66	<10	100
1 (Quinoline)				130	Quinoline
	55	33	33	164	5R,6S-Dihydrodiol (1d1)
	35	13	13	146	Quinolin-3-ol (1h3)
	6	n.d.	n.d.	164	cis-7,8-Dihydrodiol (1d2)
	4	27	27	146	Quinolin-8-ol (1h8)
	0	27	27	136 (ESI-)	Anthranilic acid
2 (2-Methylquinoline)	>99
				144	2-Methylquinoline
	92			178	5R,6S-Dihydrodiol (2d1)
	8			160	2-Methylquinolin-8-ol (2h8)
3 (4-Methylquinoline)	20
				144	4-Methylquinoline
	80			160	Quinolin-4-ylmethanol (3e)
	20			178	cis-5,6-Dihydrodiol (3d1)
					Quinolin-4-ylmethanol
4 (6-Methylquinoline)	45
				144	6-Methylquinoline
	100			160	6-Methylquinolin-3-ol (4h3)
5 (8-Methylquinoline)	78
				144	8-Methylquinoline
	72			160	Quinolin-8-ylmethanol (5e)
	28			178	cis-5,6-Dihydrodiol (5d1)
	74	5	100
6 (Isoquinoline)				130	Isoquinoline
	60–65	42	45	146	Isoquinolin-5-ol (6h5)
	16	58	47	164	cis-7,8-Dihydrodiol (6d2)
	6–8		n.d	164	cis-5,6-Dihydrodiol (6d1)
	16	n.d.	n.d.	146	Isoquinolin-4-ol (6h4)
	0	n.d.	n.d.	146	Isoquinolin-8-ol and -1-ol
7 (1-Methylisoquinoline)	52
				144	1-Methylisoquinoline
	71			160	Isoquinolin-1-ylmethanol (7e)
	29			178	7S,8R-Dihydrodiol (7d2)
8 (3-Methylisoquinoline)	>99
				144	3-Methylisoquinoline
	73			178	cis-7,8-Dihydrodiol (8d2)
	27			160	3-Methylisoquinolin-5-ol (8h3)

---

Table 2 Biotransformations of fused bicyclic bisazaarenes by the NDO system from Pseudomonas sp. NCIB 9816-4 (27 °C, 30 h)

Substrate	Total conversion (%) and relative yield of products (%)			m/z expected	Identified major products
	NDO^a	TDO^b	P. putida (mutant)^c
a This work, E. Coli JM109(DE3)(pDTG141). b From ref. 9, P. putida UV4. c From ref. 8, mutant of P. putida.
9 (Phthalazine)	20
				131	Phthalazine
	10			165	Dihydrodiol
	90			147	Phthalazin-5-ol (9h5)
10 (Cinnoline)	25
				131	Cinnoline
	n.d			165	Dihydrodiol
	n.d.			147	Cinnolinols
	>99	<10	100
				131	Quinazoline
11 (Quinazoline)	56	55	55	165	5R,6S-cis-Dihydrodiol (11d1)
	0			167	5,6-cis-Tetrahydrodiol
	0	45	10	147	Quinazolin-4-ol
	43	—	—	147	Quinazolin-8-ol (11h8)
12 (Quinoxaline)	>99	<10	100
				131	Quinoxaline
	100	47	45	165	5R,6S-cis-Dihydrodiol (12d1)
	0	44	40	147	Quinoxalin-5ol
	0	9	20	147	Quinoxalin-2ol
13 (1,5-Naphthyridine)	20
				131	1,5-Naphthyridine
	n.d.			165	Dihydrodiol
	n.d.			147	Naphthyridinols

---

2.1 Quinoline, isoquinoline and C-methyl derivatives

Quinoline (1). Quinoline (1) was not completely transformed after 30 h by E. coli JM109(DE3)(pDTG141) but the conversion was better than previously reported for the biotransformations conducted with Pseudomonas putida UV4 (expressing TDO).⁹ Nevertheless the products formed are similar, respectively cis-5,6-dihydroquinoline-5,6-diol (1d₁), cis-7,8-dihydroquinoline-7,8-diol (1d₂), quinolin-3-ol (1h₃), and quinolin-8-ol (1h₈), except that 2-quinolone or anthranilic acid were not found in our incubation extracts (HPLC-MS, positive and negative modes).

Concerning the regioselectivity of the reaction, assuming a molecular similarity with naphthalene, two different cis-dihydrodiols were expected on the carbo ring; the 5,6-position is favoured (55% yield) compared to the 7,8-position, even if the small amount of quinolin-8-ol (1h₈) formed may come from a spontaneous dehydration of the 7,8-dihydrodiol (1d₂) (10% yield). For the aza ring two positions for dioxygenation should be considered. As expected, due to nitrogen in position 1, no transformation on the 1 and/or 2 position was observed. The cis-3,4-dihydrodiol was not isolated, but its spontaneous dehydration could produce the quinolin-3-ol (1h₃) formed in 35% yield. The optical rotation of the purified 5,6-dihydrodiol 1d₁ ([α]_D +204; c 0.032, MeOH) shows that it is a single enantiomer, identical to that previously isolated,⁹ (5R,6S)-5,6-dihydroquinoline-5,6-diol. The oxidation is not specific of the carbo ring of quinoline.

2-Methylquinoline (quinaldine) (2). Quinaldine (2) was completely transformed after 20 h. Two products were isolated and their structures established by NMR analysis. cis-2-Methyl-5,6-dihydroquinoline-5,6-diol (2d₁) was obtained quite exclusively (92% yield) and 2-methylquinolin-8-ol (2h₈) (8% yield) might come from a spontaneous dehydration of a 7,8-dihydrodiol during incubation or workup. The optical rotation of the 5,6-dihydrodiol 2d₁ ([α]_D + 140; c 0.028, MeOH) shows that the biotransformation was enantioselective . In order to establish its absolute configuration, a 4-bromophenyl boronate derivative of 2d₁ (2d₁B) was prepared, crystallised and its structure resolved by X-ray crystallographic analysis (Fig. 1).

Fig. 1 X-Ray structure of the 4-bromophenylboronate derivative of (5R,6S)-2-methyl-5,6-dihydroquinoline-5,6-diol (2d₁B).

Analysis of the structure (Fig. 1) unequivocally revealed a cis-diaxial substitution with an absolute (5R,6S) configuration based on the clear indication of the Flack index at the end refinements.^18,19 Put together with the high positive value of the optical activity, this also suggested that the dihydrodiol 2d₁ was the enantiopure (5R,6S)-2-methyl-5,6-dihydroquinoline-5,6-diol.

4-Methylquinoline (lepidine) (3). Biotransformation of this substrate was limited to about 20% and afforded only 4% (20% yield) of cis-4-methyl-5,6-dihydroquinoline-5,6-diol (3d₁) as shown by NMR analysis. Its stereochemistry was not investigated further. The major product (80% yield) was identified as quinolin-4-ylmethanol (3e) by NMR analysis.

6-Methylquinoline (4). This substrate was transformed with a moderate conversion (45%) and the only product isolated was 6-methylquinolin-3-ol (4h₃), perhaps because of the poor stability of the 3,4-dihydrodiol as precedently stated for quinoline (1).⁹ The structure of 4h₃ was established by NMR analysis. The substitution by a methyl group in position 6 prevents the formation of 5,6- and 7,8-dihydrodiol and oxidation takes place only on the aza ring of (4).

8-Methylquinoline (5). 8-Methylquinoline (5) is a pretty good substrate for this biotransformation (78% conversion). Oxidation of the methyl group to give the quinolin-8-ylmethanol (5e), identified by NMR analysis, takes place in a large extent (72% yield). Nevertheless the 5,6-dihydrodiol (5d₁), identified by NMR analysis, was also produced (28% yield). The value of the coupling constant (³J_5–6 = 4.7 Hz) and the stoichiometric formation of a 4-bromophenylboronate derivative (5d₁B) agree with a cis-8-methyl-5,6-dihydroquinoline-5,6-diol. This cis-dihydrodiol is optically active, ([α]_D + 222; c 0.33, MeOH), but its absolute configuration still remains to be determined.

Isoquinoline (6). As above for quinoline (1) the biotransformation of isoquinoline (6) by E. coli JM109(DE3)(pDTG141) was observed and conversion reached about 74%. Four products were identified by NMR of the complex crude organic extract, by comparing with previously published NMR spectra.⁹ Isoquinolin-5-ol (6h₅) (65% yield) was the only one isolated with a good purity. cis-5,6-Dihydroisoquinoline-5,6-diol (6d₁) is formed (5.5% yield) but it gives spontaneously 6h₅ during the incubation and/or purification on silica gel; this instability was previously reported by Boyd et al.,⁸ and impairs a high recovery. A second dihydrodiol was isolated and identified to cis-7,8-dihydroisoquinoline-7,8-diol (6d₂) (15% yield); its optical rotation evaluated from that of a mixture with 6d₁ is + 134 (assuming that 6d₁ is the same enantiomer as obtained by Boyd et al. with TDO⁹). Isoquinoline-4-ol (6h₄), isolated by reversed phase chromatography on a prepacked C18 cartridge and identified by NMR analysis represents 7.5% of the products. Isoquinoline-8-ol was not detected in the crude extract. These results show that the enzyme can accommodate three positions of the substrate in the active site for oxidation, as observed for quinoline.

1-Methylisoquinoline (7). The conversion of this substrate reached 53% after 30 h and was in the same range to that of isoquinoline (6). Isoquinolin-1-ylmethanol (7e),²⁰ was formed (29% yield) and the 7,8-dihydrodiol (7d₂), was isolated as the major product (71% yield). It is optically active ([α]_D + 25; c 0.35, MeOH) and its cis configuration suggested by the value of ³J_7,8 is confirmed by the formation of a 4-bromophenylboronate derivative (7d₂B). 7d₂B was crystallised and the structure resolved by X-ray crystallographic analysis (Fig. 2).

Fig. 2 X-Ray structure of the 4-bromophenylboronate derivative of (7S,8R)-1-methyl-7,8-dihydroisoquinoline-7,8-diol (7d₂B).

As for the derivative of 2d₁B, the crystal structure analysis (Fig. 2) revealed a cis-diaxial derivative, with the (7S, 8R)-absolute configuration, based on the anomalous scattering of the bromine atom. This suggested that dihydrodiol 7d₂ was the enantiopure (7S,8R)-1-methyl-7,8-dihydroisoquinoline-7,8-diol.

3-Methylisoquinoline (8). 3-Methylisoquinoline was completely transformed after 30 h and two products were isolated, a dihydrodiol (8d₂) and a monohydroxylated derivative (8h₅). The yield of the dihydrodiol is pretty high (72% yield). Its structure was established by NMR analysis and a cis configuration suggested by the value of ³J_7,8, was confirmed by the formation of a 4-bromophenylboronate derivative (8d₂B). Thus this dihydrodiol is cis-3-methyl-7,8-dihydroisoquinoline-7,8-diol. It is optically active ([α]_D + 192; c 3.4, MeOH), but its absolute configuration still remains to be determined. The structure of the obtained monohydroxylated derivative (8h₅) (27% yield) was established by NMR analysis and agrees with that of 3-methylisoquinolin-5-ol. This shows that oxidation takes place only on the carbo ring of the substrate.

The results obtained for the biotransformations of compounds 1–8 are summarized in Table 1.

2.2 Bis-azaarenes

Quinazoline (11). Biotransformation of quinazoline in the usual conditions was complete after 30 h and gave similar amounts of dihydrodiol (11d₁) and a monohydroxylated compounds of phenol type (11h₈) which were isolated. The ¹H NMR spectrum of 11d₁ showed that it is cis-5,6-dihydroquinazoline-5,6-diol previously described by Boyd et al.,⁹ and the cis configuration was confirmed by the formation of 4-bromophenylboronate derivative (11d₁B). The dihydrodiol 11d₁ was optically active ([α]_D + 197; c 7.4, MeOH) but the absolute configuration was not established. Nevertheless the high positive value previously reported for 1d₁ and 12d₁,⁹ suggests a 5R,6S-configuration. The phenol was identified to be quinazolin-8-ol (11h₈) by NMR analysis. NDO was more efficient and more regioselective for the biotransformation of quinazoline than TDO.

Quinoxaline (12). Quinoxaline was completely transformed after only 5 h into one dihydrodiol (12d₁). The ¹H NMR spectrum of 12d₁ showed that it was cis-5,6-dihydroquinoxaline-5,6-diol previously described by Boyd et al.⁹ The cis configuration is confirmed by the formation of a 4-bromophenylboronate derivative (12d₁B). The value of the optical activity found for the dihydrodiol, ([α]_D + 208; c 7.5, MeOH), close to + 210 reported by Boyd et al.⁹ shows that 12d₁ is the single enantiomer, 5R,6S-dihydroquinoxaline-5,6-diol. Obviously NDO exhibits a high activity toward the stereospecific dihydroxylation of quinoxaline.

Other bisazaarenes. Phthalazine (9), cinnoline (10) and naphthyridine (13) were also used as substrate for biotransformation, but in each case the conversion was low (≈ 20%) with only a small amount of dihydrodiol formed. For these compounds, it was not possible to isolate any dihydrodiol. Nevertheless phthalazin-5-ol (9h₅) was obtained and identified by NMR analysis from the biotransformation of (9). Monohydroxylations of 10 and 13 were not regiospecific and no purified products could be isolated for their identification.

The results obtained for the biotransformation of compounds 9–13 are summarized in Table 2.

3. Discussion

The substrates tested (1–13) are transformed by E. coli JM109(DE3)(pDTG141) into one or several oxidized products with the following efficiencies: quinazoline (11) > 2-methylquinoline (quinaldine) (2), 3-methylisoquinoline (8), 8-methylquinoline (5), quinoxaline (12) > isoquinoline (6), quinoline (1), 1-methylisoquinoline (7) > 6-methylquinoline (4) > 4-methylquinoline (lepidine) (3), phthalazine (9), naphthyridine (13), cinnoline (10). cis-Dihydrodiols have been detected from all of them except with 4 were only 6-methylquinolin-3-ol was obtained, and 13. The cis configuration of most of the dihydrodiols was chemically confirmed by the formation of their bromophenyl boronate derivatives which were also used for the determination of their absolute configurations by X-ray crystallography. Other monohydroxylated phenol type compounds have also been recovered in the biotransformations of 1, 2, 6, 8, 10, and 11 either resulting from the spontaneous dehydration of a primarily formed cis-dihydrodiol or directly formed via a monohydroxylation reaction.²¹ A third reaction type mediated in the active site results in the hydroxylation of methyl substituents, exclusively when the methyl substitution of the substrate is adjacent to the ring junction, as precedently reported for 1-methylnaphthalene.²² Referring to quinoline or isoquinoline, and depending on its position on the carbo- or aza-ring, a methyl substituent is able to affect in a large extent the biotransformation yield and the regioselectivity of the reactions as previously described for benzofurane or benzothiophene.²³

Regarding the dihydroxylation reaction, the published crystallographic data from Karlsson et al. of the enzyme (PDB ID: 1O7G) suggests that the positioning of a substrate toward iron in the active site can be deduced from the structure and the stereochemistry of the produced dihydrodiol(s).²⁴ We propose a description of our and other results where we consider that a bicyclic substrate can occupy in the active site the same place as naphthalene. Due to the asymmetry of the structure of the active site including the catalytic iron, four dihydroxylated products could be obtained, each of them corresponding to a distinct orientation (or “pose”) of the substrate in the active site (referred to as A, B, C and D in Scheme 2). We have to note that the stereochemistry of the dihydrodiol is undissociable from its location on the structure. The number of products formed can be lowered due to the geometry of the substrate. With naphthalene where the four poses are identical, only one product, (1R,2S)-cis1,2-dihydro-1,2-dihydroxy-naphthalene can be recovered. Anthracene, where also the four poses are identical due to the symmetry of the molecule, similarly gives a single dihydrodiol (1R,2S)-1,2-dihydroanthracene-1,2-diol.²⁵

Scheme 2 A: Relative positioning of 1,2-dihydrodihydroxy-naphthalene with respect to mononuclear iron in the active site of NDO (from crystallographic data²⁶). B: The four possible poses for a N-monosubstituted bicyclic substrate in the active site: no subscript refers to naphthalene, N subscripts refer to quinoline and iN subscripts refer to isoquinoline.

The regio- and stereochemical analysis of the products formed from a series of monosubstituted naphthalenes using P. fluorescens TTC1 (expressing NDO) have been previously performed by Bestetti et al.²² If we refer to our model of the active site (Scheme 2), the regio- and stereochemistry of the dihydrodiols they obtained for the 2-substituted naphthalenes correspond to pose B, but trace amounts of other products suggest that pose A is also adopted for 2–Cl and 2–NO₂ naphthalenes. Thus for 2-substituted naphthalenes, pose B ≫ pose A > pose C, and pose D). With 1-substituted naphthalenes two dihydrodiols are usually obtained for each substituted compound corresponding to pose B and pose A (pose B > pose A). Nevertheless, with a methyl or ethyl substituent, only one dihydrodiol is formed, namely (1R,2S)-8-methyl or (1R,2S)-8-ethyl-1,2-dihydronaphthalene-1,2-diol corresponding to pose B. We have to note that no dihydrodiol is formed corresponding to the substrate in pose C or in pose D. Phenanthrene,²⁷ and benzo(h)quinoline and benzo(f)quinoline recently studied by Boyd et al.,²⁸ can be considered as 1,2-disubstituted naphthalenes, and should adopt pose B in the active site; accordingly, the major bioproducts obtained were respectively (3S,4R)-3,4-dihydrophenanthrene-3,4-diol, (9S,10R)-9,10-dihydrobenzo[h] quinoline-9,10-diol, and (9S,10R)-9,10-dihydrobenzo[f] quinoline-9,10-diol, corresponding to the usual regioselectivity observed for the bay region and with the expected stereospecificity. Minor cis-dihydroxylated products, quoted “abnormal” by Boyd et al. because they are formed at a non-bay region of the substrate, are also isolated in these biotransformations with NDO,²⁸ respectively (1R,2S)-1,2-dihydrophenanthrene-1,2-diol, and (7R,8S)-7,8-dihydro-benzo[h]quinoline-7,8-diol and agree with a pose A of these substrates in the active site (pose B > pose A). The proposed scheme is validated by the structure and the stereochemistry of the dihydrodiols isolated by the different authors and the high enantiomeric excess found for these products.

In the particular case of 1-methyl naphthalene, a minor different metabolite is isolated, naphthalen-1-yl methanol formed enzymatically by benzylic oxidation, revealing for this type of reactivity a pose for the substrate close to pose D (perhaps shifted to allow the methyl group to be close enough to iron in the catalytic pocket).

From a geometrical point of view, it should be expected that quinoline (1) and isoquinoline (6) occupy the same four poses as naphthalene in the active site, but due to the presence of a nitrogen atom, they are not equivalent. The dihydrodiols obtained by biotransformation of quinoline (1) have been identified to (5R,6S)-5,6-dihydroquinoline-5,6-diol (1d₁) and 7,8-dihydroquinoline-7,8-diol (1d₂), which implies the two poses A and B for the substrate (further referred to as A_N and B_N in Scheme 2). Accordingly, the absolute configuration of (1d₁) is 5R,6S, as observed by Boyd et al. in the biotransformation with TDO.⁹ If we consider that the absolute configuration of the 7,8-dihydrodiol (1d₂) obtained with NDO is 7S,8R as previously established with TDO, then dihydroxylation occurs according with the substrate in pose B_N. Assuming that quinolin-3-ol (1h₃) is produced from the cis-3,4-diol by spontaneous dehydration, that means that pose C would be also allowed for the substrate (referred to as C_N in Scheme 2). No hydroxylated product has been detected which could have been formed with the substrate at pose D (referred to as D_N in Scheme 2). For the efficiency of the dihydroxylation reaction, we have observed that pose A_N ≈ pose C_N ≫ pose B_N ≫ pose D_N. The regio- and stereo-specificity for the 5R,6S-dihydrodiol 12d₁ we observe for the dihydroxylation of quinoxaline (12), shows as expected, that the substrate occupies only pose A_N, and that a pose similar to D_N is not accepted.

The products identified after biotransformation of isoquinoline (6), (5R,6S)-5,6-dihydroisoquinoline-5,6-diol (6d₁) (and isoquinolin-5-ol (6h₅)), cis-7,8-dihydroisoquinoline-7,8-diol and isoquinolin-4-ol suggest that the allowed poses for the substrate are also respectively A_N, B_N and C_N (referred to as A_iN , B_iN and C_iN in Scheme 2), without any product corresponding to dihydroxylation in pose D_N (referred to as D_iN in Scheme 2). For the efficiency of the dihydroxylation reaction, we have obtained pose A_iN ≫ pose B_iN > pose C_iN ≫ D_iN.

For quinoline and isoquinoline, specific interactions of nitrogen with the active site could be involved for the positioning of the substrate favouring poses A_N and A_iN , and excluding poses D_N and D_iN for cis-dihydroxylation reaction. Quinazoline (11) can be considered either as a quinoline or as an isoquinoline. The products obtained, the 5,6-dihydrodiol 11d₁ and quinazolin-8-ol (11h₈) are those expected from each of the preferred poses (pose A_N and pose A_iN) of quinoline and isoquinoline respectively. Therefore the stereochemistry of 11d₁ should be 5R,6S as also suggested by the high positive value of the optical activity.

For methyl substituted quinolines and isoquinolines the structure of the dihydrodiol formed is due to the position of the methyl substituent (as with substituted naphthalene) and to the position of nitrogen. In the case of quinaldine (2), the dihydrodiol expected from a major role of nitrogen (pose A_N) would be the 5R,6S-dihydrodiol while the effect of the methyl substituent would favour the 7S,8R-dihydrodiol (pose B). As the major product obtained is the 5R,6S-dihydrodiol (2d₁), as expected for a director role of nitrogen in the positioning of this substrate at the active site (pose A_N). With lepidine (3) and 8-methylquinoline (5), the regioselective formation of the 5,6-dihydrodiols 3d₁ and 5d₁ is observed (pose A_N) but the major way of transformation is the benzylic oxidation of the methyl substituent (in pose D), which occurs as a minor way with 1-methyl-naphthalene. For 6-methylquinoline (4), the 3-hydroxy derivative 4h₃ may be produced by direct monohydroxylation of the substrate or by the spontaneous dehydration of the 3,4-dihydrodiol. Therefore the substrate should be in pose B (as 2-methylnaphthalene) corresponding to pose C_N of quinoline and excluding pose A_N. For 1-methylisoquinoline (7) the 7S,8R cis-dihydrodiol obtained (7d₂) is not that expected from an isoquinoline (pose A_iN) but from 1-methylnaphthalene (pose B). The low value of its α_D prompted us to investigate its chirality by CD spectrum, versus the cis-dihydrodiol from naphtalene used as a reference.²⁹ The same Cotton effect was observed suggesting equivalent absolute configurations and hence favoring pose B in the active site. The absolute configuration was definitely proven by X-ray diffraction (Fig. 2). In addition, the benzylic oxidation of the methyl substituent also occurs (pose D). Biotransformation of 3-methylisoquinoline (8) is one of the most efficient in this series of compounds. No 5,6-dihydrodiol is observed and 5-hydroxy derivative is only a minor product whereas, in the pose A_iN for isoquinoline or pose B for 2-methylnaphthalene, such hydroxylation should be favoured. The formation of cis-7,8-dihydrodiol (8d₂) as the major product is not expected and suggests a specific interaction within the active site for the methyl substituent of 8 in pose A. Thus the stereochemistry of the cis-7,8-dihydrodiol of 3-methylisoquinoline (8d₂) should be 7S,8R, as suggested also by the large positive value of the optical activity of this compound.

From this description of our results we can note for the regiospecific dihydroxylation of isoquinoline and C-methyl derivatives of quinolines (excepted for 6-methylquinoline), better efficiency is observed with substrate in poses A_N and A_iN, pose A_N ≈ pose C_N for quinoline, pose C_N for 6-methyl quinoline, and pose B_iN for C-methyl derivatives of isoquinolines.

We have also observed an enhanced efficiency of dihydroxylation of quinaldine (2) compared to that of quinoline (1) put together with the unexpected formation of the cis-7,8-dihydrodiol (8d₂) as the major product of oxidation of 3-methylisoquinoline (8) this suggests a specific interaction within the active site for the methyl substituent of these substrates in pose A. It would be interesting to identify the amino acid(s) of the active site involved in this interaction and substrate docking simulations would be helpful.

Benzylic hydroxylation is the major way of biotransformation for 3, 5, and 7. This reaction implies pose D for the substrate and is not dependent of the location of nitrogen on the bicyclic structure. As for 1-methylnaphthalene, a regiospecific cis-dihydroxylation is in competition with this hydroxylation.

4. Conclusion

The results of the present investigation clearly show that quinoline, isoquinoline and their frequently associated methyl derivatives as environmental pollutants (often quoted as NPAHs) are transformed by the whole cells of E. coli JM109(DE3)(pDTG141) expressing the naphthalene dioxygenase system from Pseudomonas sp. NCIB 9616-4. Optically active cis-dihydrodiols and/or the corresponding phenol derivatives of the carbocycle are generally produced. Nevertheless, it appears also that when the methyl substituent is at the vicinity of the junction of the bicyclic structure its monohydroxylation occurs preferentially (3, 5 and 7), independent of the nitrogen position. An important set of new products with valuable stereochemical and/or functional properties was generated by these biotransformations.

5. Experimental

5.1 Substrates and chemicals

Substrates were from commercial origin : quinoline, quinaldine, 1-methylisoquinoline, 3-methylisoquinoline, cinnoline hydrochloride, and naphthyridine hydrochloride from Aldrich, lepidine, isoquinoline, 6-methylquinoline, phthalazine, quinoxaline, and quinazoline, from Acros Organics and, 8-methylquinoline from Lancaster, and used without further purification except quinoline which was distilled before use.

5.2 Analytical procedures

¹H NMR and ¹³C NMR spectra were recorded in (CD₃)₂CO (unless stated otherwise) on Bruker ARX250 (250 MHz for ¹H) or Bruker Avance500 (500 MHz for ¹H and 125 MHz for ¹³C) spectrometers. Chemical shifts (δ) are reported in ppm, and coupling constants are given in Hz. ¹H signals were assigned using 2D homonuclear ¹H-¹H COSY and TOCSY experiments. The TOCSY and the COSY spectra were used to distinguish the spin systems and to confirm assignments of the hydrogen resonances. The ³J_1H,1H coupling constants were determined by homodecoupling experiments of 1D ¹H spectra. ¹³C signals were assigned using 2D heteronuclear ¹H-¹³C HSQC and HMBC experiments. Additionally, 2D-NOESY experiments were used to assign the spin systems closely related. Optical rotation measurements ([α]_D) were given in 10⁻¹ deg cm² g⁻¹, and were carried out with a Perkin Elmer 241 polarimeter at ambient temperature (ca. 20 °C). HPLC-ESIMS analyses were performed with a Thermo Finnigan LCQ Advantage Instrument (electrospray positive mode) equipped for HPLC with a Polarity dC18 column (waters, 4.6 × 250 mm, 5 μm) equilibrated with the appropriate water–CH₃CN mobile phase. Products were eluted with a CH₃CN gradient containing formic acid (999 : 1) at 40 °C and 0.4 cm³ min⁻¹ flow rate. Purification of the organic extract of biotransformation was achieved by SP-flash chromatography on Kieselgel Si60 (Merck, 40–63 μm) columns equilibrated with CH₂Cl₂ and eluted with CH₂Cl₂/(CH₃)₂CO/CH₃OH in the appropriate ratios and reversed-phase (RP)-flash chromatography over a prepacked C18 column (7.5 cm × 1.5 cm; AIT, Houilles, France), using a gradient from 0/100 to 60/40 of CH₃CN/H₂O mixture. The CD spectra were recorded on a JASCO J-815 spectrometer in methanol solutions. The two X-ray structure coordinates of 2d₁B and 7d₂B were deposited with the Cambridge Crystallographic Data Centre (CCDC) 12 Union road, Cambridge CB2 1EZ, UK under reference codes 815425 and 815426, respectively, they are available upon request to http://www.ccdc.cam.ac.uk/.

5.3 Micro organism and growth conditions

Escherichia coli JM109(DE3)(pDTG141) was kindly provided by D. T. Gibson (Iowa University). Cells were grown as previously described,⁶ and their dihydroxylating activity toward naphthalene routinely examined in initial rate conditions as described for analytical biotransformations (see below). Activity was found to be about 1–2 mM 1,2-dihydroxydihydronaphthalene formed/hour/OD₆₀₀ unit.

5.4 General procedure for analytical biotransformations

In optimised conditions, frozen cells were suspended in 20 cm³ of 0.1 M phosphate buffer pH 7.2 (0.5 g of cells, OD₆₀₀ about 7), containing sodium ampicillin (0.27 mM, 0.1 g L⁻¹) and glucose (0.2%) in 125 cm³ conical flasks. After addition of 5 mg of substrate, the biotransformation was conducted at 27 °C with orbital shaking at 250 rpm for 30 h. Aliquots (0.7 cm³) were withdrawn at intervals, centrifuged and micro filtered for direct HPLC-MS analysis.

5.5 General procedure for preparative biotransformations

Frozen cells (about 12 g) were resuspended in 500 cm³ of 0.1 M phosphate buffer (pH 7.2) containing sodium ampicillin (0.27 mM) and glucose (0.2%) in 2 L conical flasks. After addition of 150 mg of substrate, biotransformations were conducted at 27 °C with orbital shaking at 250 rpm for 30 h. Aliquots (0.7 cm³) were withdrawn at intervals, centrifuged and micro filtered for direct HPLC-MS analysis.

5.6 Characterization and identification of the bioproducts

5.6.1 Biotransformation of quinoline (1). HPLC-MS analysis of the incubation medium revealed two metabolites with m/z = 164 and two metabolites with m/z = 146, with a biotransformation reaching 66% (initial rate around 0.04 mM h⁻¹/OD₆₀₀ unit). Accordingly, flash chromatography afforded four products which were respectively identified as (5R,6S)-5,6-dihydroquinoline-5,6-diol (1d₁), cis-7,8-dihydroquinoline-7,8-diol (1d₂), quinolin-3-ol (1h₃), and quinolin-8-ol (1h₈) by ¹H NMR previously published by Boyd et al.⁹ Their relative yields, evaluated by HPLC-MS analysis of the incubation medium when the biotransformation was stopped were 55, 6, 35 and 4% respectively. Optical rotation measurement of 1d₁ ([α_D]) was + 204 (c 0.032, MeOH), literature ref. 9: + 220 (THF).

5.6.2 Biotransformation of quinaldine (2). HPLC-MS analysis of the incubation medium revealed one major metabolite with m/z = 178 (71%, dihydrodiol) and one monohydroxylated metabolite (m/z = 160, 18%) with a biotransformation reaching 100% in analytical experiments (initial rate around 0.09 mM h⁻¹/OD₆₀₀ unit). Another dihydrodiol (<10%) and trace amounts of quinaldic acid (m/z = 174) and aldehyde (m/z = 158) were also detected (<1%).

Two products were isolated by flash chromatography of preparative experiments and identified as cis-2-methyl-5,6-dihydroquinoline-5,6-diol (2d₁) and 2-methylquinolin-8-ol (2h₈) by ¹H and ¹³C NMR analysis.

cis-2-Methyl-5,6-dihydroquinoline-5,6-diol (2d₁). δ _H (500 MHz, (CD₃)₂CO): 7.05 (1H, d, J_3–4 = 7.6, 3-H), 7.68 (1H, d, J_4–3 = 7.6, 4-H), 4.64 (1H, d, J_5–6 = 4.9, 5-H), 4.30 (1H, t, J = 4.9, 6-H), 6.27 (1H, dd, J_7–8 = 9.9, J_7–6 = 4.3 7-H), 6.52 (1H, d, J_8–7 = 9.9, 8-H), 2.42 (3H, s, CH₃); δ_C (125 MHz, (CD₃)₂CO): 157.3 (C2), 122.1(C3), 135.7 (C4), 130.5 (C4a), 70.5 (C5), 67.2 (C6), 134.7 (C7), 131.0 (C8), 151.7 (C8a), 23.8 (CH₃). Optical rotation measurement of 2d₁ ([α_D]) was + 140 (c 0.028, MeOH).
(4-Bromo)phenylboronate derivative (2d₁B). 2d₁ (85 mg, 0.48 mmol) and 4-bromophenyl boronic acid (105 mg, 1.1 equiv.) were reacted in CH₂Cl₂/(CH₃)₂CO (5 cm³) 1/1 at room temperature during 1 h. The product formed was purified by flash chromatography on silica with CH₂Cl₂ as eluent. δ_H (250 MHz, CDCl₃): 7.82 (1H, d, J_4–3 = 7.8, 3-H), 7.21 (1H, d, J_4–3 = 7.8, 4-H), 6.63 (1H, d, J_6–5 = 10.0, 5-H), 6.30 (1H, dd, J_6–5 = 10.0, J_6–7 = 3.3, 6-H), 5.56 (1H, ddd, J_8–7 = 7.8, J_7–6 = 3.3, ⁴J_7–5 = 1.2, 7-H), 5.81 (1H, d, J_8–7 = 8.3, 8-H), 2.47 (3H, s, CH₃), 7.62–7.75 (4H, m, Ar–H). Crystals suitable for X-ray analysis were grown by slow evaporation from a CH₂Cl₂ solution as thin prisms, they were recorded on a Bruker AXS CCD diffractometer (Mo-Kα wavelength). Crystal data for 2d₁B: monoclinic, P2₁, with parameters: a = 4.320(5); b = 10.637(4); c = 16.090(4) Å; β = 94.11(5)° and Z = 2. Structure was refined to R = 3.7% (1495 observed Fo) and R = 5.3% (all 1799 Fo data).
2-Methylquinolin-8-ol (2h₈). δ _H (250 MHz, (CD₃)₂CO): 8.21 (1H, d, J_4–3 = 8.5, 4-H), 7.46 (1H, d, J_3–4 = 8.5, 3-H), 7.42 (1H, t, J_5–6 = 7.8, 6-H), 7.38 (1H, d, J_5–6 = 7.8, 5-H), 7.11 (1H, d, J_7–6 = 7.8, 7-H), 2.72 (3H, s, CH₃); δ_C (125 MHz, (CD₃)₂CO): 159.1 (C2), 124.8 (C3), 138.3 (C4), 128.9 (C4a), 119.6 (C5), 128.7 (C6), 111.9 (C7), 154.3 (C8), 139.8 (C8a), 26.1 (CH₃).

5.6.3 Biotransformation of lepidine (3). HPLC-MS analysis of the incubation medium revealed one major metabolite with m/z = 160 and one metabolite with m/z = 178 with a biotransformation reaching only 63% in analytical experiments (initial rate around 0.03 mM h⁻¹/OD₆₀₀ unit). Two other metabolites with m/z = 158 (<2%) and m/z = 174 (<1%) were also detected.

Two products were isolated by flash chromatography of preparative biotransformations and identified as quinolin-4-ylmethanol (3e) and cis-4-methyl-5,6-dihydroquinoline-5,6-diol (3d₁) by ¹H and ¹³C NMR spectroscopy.

Quinolin-4-ylmethanol (3e). δ _H (500 MHz, (CD₃)₂CO): 8.92 (1H, d, J_2–3 = 3.8, 2-H), 7.68 (1H, d, J_3–2 = 4.4, 3-H), 7.63 (1H, t, J_6–5 = 7.6, 6-H), 7.78 (1H, t, J_7–8 = 7.6, 7-H), 8.11 (2H, t, J = 7.1, 5-H and 8-H), 5.24 (2H, s, CH₂); δ_C (125 MHz, (CD₃)₂CO): 152.0 (C2), 119.7 (C3), 149.6 (C4), 127.6 (C4a), 131.4 (C5), 128.0 (C6), 130.6 (C7), 131.6 (C8), 149.4 (C8a), 62.0 (CH₂).
cis-4-Methyl-5,6-dihydroquinoline-5,6-diol (3d₁). δ _H (500 MHz, (CD₃)₂CO): 8.25 (1H, d, J_2–3 = 5.2, 2-H), 7.02 (1H, d, J_3–2 = 5.2, 3-H), 4.84 (1H, dd, J_5–6 = 5.2, ⁴J_5–7 = 1.5, 5-H), 4.52 (1H, m, 6-H), 6.09 (1H, dt, J_7–8 = 9.8, ⁴J_7–5 = 1.8, 7-H), 6.43 (1H, dd, J_8–7 = 10.0, ⁴J_8–6 = 2.7, 8-H), 2.41 (3H, s, CH₃); δ_C (125 MHz, (CD₃)₂CO): 149.3 (C2), 125.1 (C3), 147.0 (C4), 129.6 (C4a), 66.2 (C5), 70.3 (C6), 137.7 (C7), 128.6 (C8), 152.1 (C8a), 17.7 (CH₃).

5.6.4 Biotransformation of 6-methyl quinoline (4). HPLC-MS analysis of the incubation medium revealed one metabolite with m/z = 160 with a biotransformation reaching only 45% in analytical experiments (initial rate around 0.03 mM h⁻¹/OD₆₀₀ unit).

The product was isolated by flash chromatography of preparative experiments and identified as 6-methylquinolin-3-ol (4h₃) by ¹H and ¹³C NMR.

6-Methylquinolin-3-ol (4h₃). δ _H (500 MHz, (CD₃)₂CO): 8.59 (1H, d, ⁴J_2–4 = 2.5, 2-H), 6.09 (1H, d, ⁴J_4–2 = 2.5, 4-H), 7.54 (1H, br s, 5-H), 7.38 (1H, dd, J_7–8 = 8.6, ⁴J_7–5 = 1.7, 7-H), 7.87 (1H, d, J_8–7 = 8.6, 8-H), 2.51 (3H, s, CH₃); δ_C (125 MHz, (CD₃)₂CO): 144.3 (C2), 152.6 (C3), 116.6 (C4), 138.1 (C4a), 127.0 (C5), 131.1 (C6), 129.8 (C7), 130.5 (C8), 143.8 (C8a), 22.4 (CH₃).

5.6.5 Biotransformation of 8-methyl quinoline (5). HPLC-MS analysis of the incubation medium revealed one metabolites with m/z = 178 and one metabolite with m/z = 160 with a biotransformation reaching 90% in analytical experiments (initial rate around 0.14 mM h⁻¹/OD₆₀₀ unit) and respective yields around 25% and 75%.

Flash chromatography afforded two products which were identified by ¹H and ¹³C NMR as cis-8-methyl-5,6-dihydroquinoline-5,6-diol (5d₁) and quinolin-8-ylmethanol (5e). They were also obtained by RP-chromatography over a prepacked C18 column and elution with H₂O–CH₃CN in the appropriate ratios.

cis-8-Methyl-5,6-dihydroquinoline-5,6-diol (5d₁). [α_D] + 222 (c 0.035, MeOH); δ_H (250 MHz, (CD₃)₂CO): 8.42 (1H, br d, J_2–3 = 4.4, 2-H), 7.21 (1H, dd, J_3–4 = 7.5, J_2–3 = 4.8, 3-H), 7.84 (1H, d, J_4–3 = 7.4, 4-H), 4.67 (1H, d, J_5–6 = 4.7, 5-H), 4.23 (1H, t, J = 4.8, 6-H), 6.17 (1H, d, J_7–6 = 3.6, 7-H), 2.11 (3H, br s, CH₃); δ_C (125 MHz, (CD₃)₂CO): 148.6 (C2), 123.1 (C3), 134.5 (C4), 134.2 (C4a), 71.0 (C5), 67.0 (C6), 130.1 (C7), 137.5 (C8), 153.7(C8a), 18.3 (CH₃).
(4-Bromo)phenylboronate derivative (5d₁B). 5d₁ (85 mg, 0.48 mmol) and 4-bromophenyl boronic acid (105 mg, 1.1 equiv.) were reacted in CH₂Cl₂/(CH₃)₂CO 1/1 (5 cm³) at room temperature during 1 h. The product formed was purified by flash chromatography with CH₂Cl₂ as eluent. δ_H (250 MHz, (CD₃)₂CO): 8.56 (1H, dd, J_2–3 = 4.8, ⁴J_2–4 = 1.6, 2-H), 7.34 (1H, dd, J_3–4 = 7.7, J_3–2 = 4.8, 3-H), 7.93 (1H, dd, J_4–3 = 7.7, ⁴J_2–4 = 1.5, 4-H), 5.49 (1H, d, J_5–6 = 9.0, 5-H), 5.27 (1H, dm, J_6–5 = 9.0, 6-H), 6.21 (1H, dt, J_7–6 = 3.8, ⁴J_7-CH3 = 1.6, 7-H), 2.16 (3H, t, J_CH3-7 = 1.6 CH₃), 7.51–7.62 (4H, m, Ar–H). The crystals grown by slow evaporation from the CH₂Cl₂ solution were not suitable for X-ray analysis.
Quinolin-8-ylmethanol (5e). δ _H (500 MHz, (CD₃)₂CO): 8.89 (1H, dd, J_2–3 = 4.3, ⁴J_2–4 = 1.8, 2-H), 7.53 (1H, dd, J_3–4 = 8.2, J_3–2 = 4.2, 3-H), 8.34 (1H, dd, J_4–3 = 8.1, ⁴J_2–4 = 1.6, 4-H), 7.82 (1H, d, J_5–6 = 7.7, 5-H), 7.58 (1H, dd, J_6–7 = 8.1, J_6–5 = 7.7, 6-H), 7.85 (1H, d, J_7–6 = 8.1, 7-H), 5.25 (2H, s, CH₂); δ_C (125 MHz, (CD₃)₂CO): 150.2 (C2), 122.1 (C3), 137.3 (C4), 129.1 (C4a), 127.3 (C5), 127.2 (C6), 127.6 (C7), 140.8 (C8), 147.1(C8a), 62.5 (CH₂).

5.6.6 Biotransformation of isoquinoline (6). HPLC-MS analysis of the incubation medium revealed two metabolites with m/z = 164 and several one with m/z = 146 with a biotransformation yield larger than 74% (initial rate around 0.08 mM h⁻¹/OD₆₀₀ unit). ¹H NMR of the crude extract revealed also several products.

SP-flash chromatography isoquinolin-5-ol (6h₅), with a small amount of an other phenol (6h₄) were eluted with CH₂Cl₂; cis-7,8-dihydroisoquinoline-7,8-diol (6d₂) and cis-5,6-dihydroisoquinoline-5,6-diol (6d₁) were eluted with CH₂Cl₂/CH₃OH (0→10%). 6h₅, 6d₁, and 6d₂ were identified by comparing their ¹H NMR spectra to published data.⁹ The structure of 6h₄ was deduced from ¹H and ¹³C NMR analysis and assigned to isoquin-4-ol. The yield of 6h₅, 6d₂, 6h₄, and 6d₁ for this biotransformation were evaluated to 55, 6, 5, and 4%, respectively by NMR analysis of the crude organic extract.

Isoquin-4-ol (6h₄). δ_H (500 MHz, (CD₃)₂CO): 8.82 (1H, s, 1-H), 8.12 (1H, s, 3-H), 8.21 (1H, d, J_5–6 = 8.3, 5-H), 7.73 (1H, t, J = 7.5, 6-H), 7.66 (1H, t, J = 7.5, 7-H), 8.03 (1H, d, J_8–7 = 8.3, 8-H); δ_C (125 MHz, (CD₃)₂CO): 144.5 (C1), 127.9 (C3), 149.0 (C4), 127.2 (C4a),121.7 (C5), 129.7 (C6), 128.3 (C7), 127.6 (C8), 130.4 (C8a).

As our attempts to obtain a pure sample of cis-7,8-dihydroisoquinoline-7,8-diol (6d₂) were not successful, the sample used to determine the optical activity of 6d₂ was a mixture of 6h₅, 6d₁, and 6d₂, in the proportions of 8, 21, and 71% respectively determined by ¹H NMR. The value of the optical activity of the sample was [α]_D + 97.5 (c 0.56, MeOH); using the value of + 163 for the optical activity of 6d₁,⁹ the estimated value for 6d₂ is then + 134.

5.6.7 Biotransformation of 1-methyl isoquinoline (7). HPLC-MS analysis of the incubation medium revealed one metabolite with m/z = 178 and one metabolite with m/z = 160 with respective yields of 29 and 71% and a biotransformation reaching only 52%.

RP-flash chromatography afforded two products which were identified by ¹H and ¹³C NMR analysis as cis-1-methyl-7,8-dihydroisoquinoline-7,8-diol (7d₂) and isoquinolin-1-ylmethanol (7e) according to Weitgenant et al.²⁰ As shown by NMR analysis of the crude organic extract, only these two products are formed in the biotransformation of 1-methyl isoquinoline.

cis-1-Methyl-7,8-dihydroisoquinoline-7,8-diol (7d₂). [α]_D + 25 (c 0.35, MeOH); δ_H (500 MHz, (CD₃)₂CO): 8.31 (1H, d, J_3–4 = 5.0, 3-H), 6.92 (1H, d, J_4–3 = 5.0, 4-H), 6.38 (1H, dd, J_5–6 = 9.8, J = 2.7, 5-H), 6.08 (1H, dt, J_6–5 = 9.8, J = 1.6, 6-H), 4.51 (1H, m, 7-H), 4.79 (1H, dd, J_8–7 = 5.2,J = 1.3, 8-H), 2.57 (3H, s, CH₃); δ_C (125 MHz, (CD₃)₂CO): 158.1 (C1), 149.9 (C3), 119.5 (C4), 140.2 (C4a), 125.4 (C5), 138.2 (C6), 70.6 (C7), 65.9(C8), 128.4 (C8a), 21.1 (CH₃).
(4-Bromo)phenylboronate derivative of cis-1-methyl-7,8-dihydroisoquinoline-7,8-diol (7d₂B). 7d₂ (85 mg, 0.48 mmol) and 4-bromophenyl boronic acid (105 mg, 1.1 equiv.) were reacted in CH₂Cl₂ (5 cm³) at room temperature during 1 h. The product formed was purified by flash chromatography with CH₂Cl₂ as eluent; δ_H (250 MHz, (CD₃)₂CO): 8.42 (1H, d, J_3–4 = 4.9, 3-H), 7.02 (1H, d, J_4–3 = 4.9, 4-H)), 6.56 (1H, dd, J_5–6 = 9.9, ⁴J_5–7 = 1.5, 5-H), 6.21 (1H, dd, J_6–5 = 10.0, J_6–7 = 3.0, 6-H), 5.60 (1H, dm, J_7–8 = 9.6, 7-H), 5.92 (1H, d, J_8–7 = 9.6, 8-H), 2.71(3H, s, CH₃), 7.56–7.76 (4H, m, Ar–H). Crystals suitable for X-ray analysis were grown by slow evaporation from the CH₂Cl₂ solution. They were obtained as a mass of thin needles 10 micron thick. After dispersion in paraffin oil, one needle was selected, fished out using a standard cryo-loop and frozen in liquid nitrogen. Diffraction data were recorded at ESRF synchrotron (Grenoble), beam line BM 30A (FIP). Wavelength was set to 0.914 Å so as to get a significant anomalous signal from the bromine K-edge. Crystal data for 7d₂B: monoclinic, P2₁ with parameters: a = 12.683(2); b = 4.267(2); c = 14.339(2) Å; β = 107.93(5)° and Z = 2. The structure was refined to R = 8.5% (1629 observed Fo) and R = 8.6% (all 1655 Fo data).

5.6.8 Biotransformation of 3-methyl isoquinoline (8). HPLC-MS analysis of the incubation medium revealed metabolites with m/z = 178 and m/z = 160 with a biotransformation larger than 99% (initial rate around 0.09 mM h⁻¹/OD₆₀₀ unit) and respective yields of 72 and 27%.

RP-flash chromatography afforded one product which was identified by ¹H and ¹³C NMR as cis-3-methyl-7,8-dihydroisoquinoline-7,8-diol (8d₂).

cis-3-Methyl-7,8-dihydroisoquinoline-7,8-diol (8d₂). [α]_D + 192 (c 3.4, MeOH); δ_H (500 MHz, (CD₃)₂CO): 8.43 (1H, s, 1-H), 6.84 (1H, s, 4-H), 6.38 (1H, dd, J_5–6 = 9.8, J = 2.7, 5-H), 6.08 (1H, dt, J_6–5 = 9.8, J = 1.6, 6-H), 4.37 (1H, m, 7-H), 4.65 (1H, d, J_7–8 = 4.9, 8-H), 2.44 (3H, s, CH₃); δ_C (125 MHz, (CD₃)₂CO): 148.8 (C1), 159.3 (C3), 120.3 (C4), 140.4 (C4a), 126.6 (C5), 136.6 (C6), 68.6 (C7), 68.5(C8), 129.0 (C8a), 24.2 (CH₃).

A monohydroxylated metabolite was isolated by SP-flash chromatography and identified by ¹H and ¹³C NMR as 3-methylisoquinolin-5-ol (8h₅).

3-Methyl-isoquinolin-5-ol (8h₅). δ _H (500 MHz, (CD₃)₂CO): 9.10 (1H, s, 1-H), 7.84 (1H, s, 4-H), 7.09 (1H, d, J_7–8 = 7.6, 6-H), 7.37 (1H, t, J = 7.9 7-H), 7.51 (1H, d, J_8–7 = 8.1, 8-H), 2.64 (3H, s, CH₃); δ_C (125 MHz, (CD₃)₂CO): 152.2 (C1), 151.3 (C3), 113.8 (C4), 129.2 and 129.1 (C4a, C8a), 152.9 (C5), 112.8 (C6), 127.7 (C7), 119.1 (C8), 24.14 (CH₃).
(4-Bromo)phenylboronate derivative of cis-3-methyl-7,8-dihydroisoquinoline-7,8-diol (8d₂B). 8d₂ (85 mg, 0.48 mmol) and 4-bromophenyl boronic acid (105 mg, 1.1 equiv.) were reacted in CH₂Cl₂ (5 cm³) at room temperature during 1 h. The product formed was purified by flash chromatography with CH₂Cl₂ as eluent. δ_H (250 MHz, (CD₃)₂CO): 8.59 (1H, s, 1-H), 7.04 (1H, s, 4-H), 6.56 (1H, d, J_5–6 = 9.9, 5-H), 6.21 (1H, dd, J_6–5 = 9.9, J_6–7 = 3.2, 6-H), 5.55 (1H, ddd, J_7–8 = 9.1, J_7–6 = 3.2, ⁴J_7–5 = 1.4, 7-H), 5.78 (1H, d, J_8–7 = 9.0, 8-H), 2.48 (3H, s, CH₃), 7.56–7.73 (4H, m, Ar-H). Crystals grown by slow evaporation from the CH₂Cl₂–Et₂O were not suitable for X-ray analysis.

5.6.9 Biotransformation of phthalazine (9). HPLC-MS analysis of the incubation medium showed metabolites with m/z = 165 and m/z = 147 in a 1 : 9 ratio, with a biotransformation which reached only ≈ 20%. There is no evidence of the presence of a dihydrodiol in the crude extract. Its purification by flash chromatography with CH₂Cl₂ as eluent gave only a monohydroxylated compound identified as phthalazin-5-ol (9h5) by NMR analysis.
Phthalazin-5-ol (9h5). δ _H (500 MHz, (CD₃)₂CO): 9.50 (1H, d, ⁵J_1–4 = 1.4, 1-H), 9.72 (1H, s, 4-H), 7.35 (1H, dd, J_6–7 = 7.5, J = 1.0, 6-H), 7.83 (1H, dd, J_7–8 = 7.9, J_7–6 = 7.5, 7-H), 7.57 (1H, d, J_8–7 = 7.9, 8-H); δ_C (125 MHz, (CD₃)₂CO): 151.8 (C1), 147.1 (C4), 118.4 (C4a), 154.7 (C5), 116.7 (C6), 135.2 (C7), 117.7 (C8), 128.6 (C8a).

5.6.10 Biotransformation of cinnoline (10). HPLC-MS analysis of the incubation medium at various time showed one metabolite with m/z = 165 and two with m/z = 147. After 30 h, the biotransformation was limited to within 20%. There was no evidence of the presence of a dihydrodiol in the crude extract of preparative experiments and its purification by flash chromatography with CH₂Cl₂ as eluent have given a mixture of monohydroxylated compounds.

5.6.11 Biotransformation of quinazoline (11). As shown by HPLC-MS analysis of the incubation medium, two metabolites with m/z = 165 and one metabolite with m/z = 147 are formed all along the biotransformation and quinazoline was completely transformed within 20 h (100% conversion, initial rate around 0.10 mM h⁻¹/OD₆₀₀ unit). ¹H NMR of the crude extract of a preparative experiment showed only two products, a dihydrodiol (56% yield) and a monohydroxy derivative (44% yield). Flash chromatography (CH₂Cl₂ : CH₃OH 100 : 0 to 90 : 10) afforded two products, identified by ¹H and ¹³C NMR analysis as quinazolin-8-ol (11h8) and the previously described cis-5,6-dihydroquinoxaline-5,6-diol (11d₁).⁹ This dihydrodiol is opticaly active: [α]_D + 197 (c 0.35, MeOH).
(4-Bromo)phenylboronate derivative of cis-5,6-dihydroquinazoline-5,6-diol (11d₁B). 11d₁ (15 mg, 0.09 mmol) and 4-bromophenyl boronic acid (20 mg, 1.1 equiv.) were reacted in CH₂Cl₂/(CH₃)₂CO 1/1 (2 cm³) at room temperature during 1 h. The product formed was purified by flash chromatography with CH₂Cl₂ as eluent. δ_H (250 MHz, (CD₃)₂CO): 8.90 (1H, s, 2-H), 9.08 (1H, s, 4-H), 5.93 (1H, d, J_5–6 = 9.4, 5-H), 5.65 (1H, dd, J_6–5 = 9.4, ⁴J_6–8 = 2.2, 6-H), 6.50–6.70 (2H, m, 7-H and 8-H), 7.50–7.80 (4H, m, Ar–H). The crystals grown by slow evaporation from the CH₂Cl₂ solution were not suitable for X-ray analysis.
Quinazolin-8-ol (11h₈). δ _H (500 MHz, (CD₃)₂CO): 9.22 (1H, s, 2-H), 9.50 (1H, s, 4-H), 7.58 (1H, dd, J_5–6 = 8.3, J_5–7 = 1.4, 5-H), 7.62 (1H, dd, J_6–5 = 7.9, J_6–7 = 7.4, 6-H), 7.37 (1H, dd, J_7–6 = 7.4 and J_7–5 = 1.4, 7-H); δ_C (125 MHz, (CD₃)₂CO): 154.5 (C2), 161.2 (C4), 126.6 (C4a), 118.3 (C5), 129.9 (C6), 116.1 (C7), 153.4 (C8), 141.1 (C8a).

5.6.12 Biotransformation of quinoxaline (12). HPLC-MS analysis of the incubation medium showed only one metabolite with m/z = 165 and biotransformation reached 100% after 5 h (initial rate around 0.39 mM h⁻¹/OD₆₀₀ unit). ¹H NMR spectrum of the crude extract of preparative experiments shows only one product, isolated by flash chromatography with CH₂Cl₂. It was identified by ¹H NMR (lit, ⁹), and optical activity ([α]_D + 208; c 7.5, MeOH) (literature ref. 9 + 210) to be the single enantiomer cis-5R,6S-dihydroquinoxaline-5,6-diol (12d₁).

5.6.13 Biotransformation of 1,5-naphthyridine (13). HPLC-MS analysis of the incubation medium at various time showed that the biotransformation was slow and bounded. Preparative experiments were not conducted.

Acknowledgements

We are indebted to Pr Gibson (Iowa state University) for the strain E.coli JM109(DE3)(pDTG141). We would like to thank Dr R. Azerad and Dr D. Mansuy (UMR 8601, Université Paris Descartes, Paris) who made valuable suggestions on reading the draft of this manuscript, and the staff of the BM30A beamline at the European Synchrotron Facility, Grenoble for beamtime allocation.

References

1. T. Hudlicky, D. Gonzalez and D. T. Gibson, Aldrichimica Acta, 1999, 32, 35–62.
2. M. A. Endoma, V. P. Bui, J. Hansen and T. Hudlicky, Org. Process Res. Dev., 2002, 6, 525–532.
3. D. R. Boyd and T. D. H. Bugg, Org. Biomol. Chem., 2006, 4, 181–192.
4. L. Modyanova and R. Azerad, Tetrahedron Lett., 2000, 41, 3865–3869.
5. D. R. Boyd, N. D. Sharma, L. V. Modyanova, J. G. Carroll, J. F. Malone, C. C. R. Allen, J. T. G. Hamilton, D. T. Gibson, R. E. Parales and H. Dalton, Can. J. Chem., 2002, 80, 589–600.
6. C. Chopard, R. Azerad and T. Prangé, J. Mol. Catal. B: Enzym., 2008, 50, 53–60.
7. S. M. Resnick, D. S. Torok and D. T. Gibson, FEMS Microbiol. Lett., 1993, 113, 297–302.
8. D. R. Boyd, R. A. S. McMordie, H. P. Porter, H. Dalton, R. O. Jenkins and O. W. Howarth, J. Chem. Soc., Chem. Commun., 1987, 1722–1724.
9. D. R. Boyd, N. D. Sharma, M. R. Dorrity, M. V. Hand, R. A. S. Mcmordie, J. F. Malone, H. P. Porter, H. Dalton, J. Chima and G. N. Sheldrake, J. Chem. Soc., Perkin Trans. 1, 1993, 1065–1071.
10. D. R. Boyd, N. D. Sharma, J. G. Carroll, J. F. Malone, D. G. Mackerracher and C. C. R. Allen, Chem. Commun., 1998, 683–684.
11. D. R. Boyd, N. D. Sharma, G. P. Coen, F. Hempenstall, V. Ljubez, J. F. Malone, C. C. R. Allen and J. T. G. Hamilton, Org. Biomol. Chem., 2008, 6, 3957–3966.
12. E. A. Bleeker, S. Wiegman, P. de Voogt, M. Kraak, H. A. Leslie, E. de Haas and W. Admiraal, Rev. Environ. Contam. Toxicol., 2002, 173, 39–83.
13. D. L. Tullis and S. Banerjee, Cancer Lett., 1984, 23, 241–244.
14. E. J. LaVoie, S. Dolan, P. Little, C. X. Wang, S. Sugie and A. Rivenson, Food Chem. Toxicol., 1988, 26, 625–629.
15. M. Dua, A. Singh, N. Sethunathan and A. K. Johri, Appl. Microbiol. Biotechnol., 2002, 59, 143–152.
16. J. Aislabie, D. J. Saul and J. M. Foght, Extremophiles, 2006, 10, 171–179.
17. S. Fetzner, B. Tshisuaka, F. Lingens, R. Kappl and J. Huttermann, Angew. Chem., Int. Ed., 1998, 37, 577–597.
18. H. D. Flack, Acta Crystallogr., 1983, A39, 876–881.
19. G. M. Sheldrick, Acta Crystallogr., 2008, A64, 112–122.
20. J. A. Weitgenant, J. D. Mortison and P. Helquist, Org. Lett., 2005, 7, 3609–3612.
21. S. M. Resnick, K. Lee and D. T. Gibson, J. Ind. Microbiol., 1996, 17, 438–457.
22. G. Bestetti, D. Bianchi, A. Bosetti, P. DiGennaro, E. Galli, B. Leoni, F. Pelizzoni and G. Sello, Appl. Microbiol. Biotechnol., 1995, 44, 306–313.
23. D. R. Boyd, N. D. Sharma, I. N. Brannigan, S. A. Haughey, J. F. Malone, D. A. Clarke and H. Dalton, Chem. Commun., 1996, 2361–2362.
24. A. Karlsson, H. Eklund, S. Ramaswamy, J. V. Parales, R. E. Parales and D. T. Gibson, J. Inorg. Biochem., 2003, 96, 164.
25. D. M. Jerina, H. Selander, H. Yagi, M. C. Wells, J. F. Davey, V. Mahadevan and D. T. Gibson, J. Am. Chem. Soc., 1976, 98, 5988–5996.
26. A. Karlsson, J. V. Parales, R. E. Parales, D. T. Gibson, H. Eklund and S. Ramaswamy, Science, 2003, 299, 1039–1042.
27. R. E. Parales, S. M. Resnick, C. L. Yu, D. R. Boyd, N. D. Sharma and D. T. Gibson, J. Bacteriol., 2000, 182, 5495–5504.
28. D. R. Boyd, N. D. Sharma, L. Sbircea, D. Murphy, T. Belhocine, J. F. Malone, S. L. James, C. C. R. Allen and J. T. G. Hamilton, Chem. Commun., 2008, 5535–5537.
29. M. Kwit, J. Gawronski, D. R. Boyd, N. D. Sharma, M. Kaik, R. A. More O'Ferrall and J. S. Kudavalli, Chem.–Eur. J., 2008, 14, 11500–11511.

---

Footnote

† CCDC reference numbers 815425–815426. For crystallographic data in CIF or other electronic format see DOI: 10.1039/c1ra00706h

---