Previous Article | Next Article ![]()
Applied and Environmental Microbiology, October 2002, p. 4965-4970, Vol. 68, No. 10
0099-2240/02/$04.00+0 DOI: 10.1128/AEM.68.10.4965-4970.2002
Copyright © 2002, American Society for Microbiology. All Rights Reserved.
School of Biological Sciences,1 Department of Chemistry, University of Wales Bangor, Bangor, Gwynedd LL57 2UW, Wales, United Kingdom2
Received 22 January 2002/ Accepted 1 July 2002
|
|
|---|
max of 446 nm. The compound producing this color has been isolated and identified as a new phenoxazinone, 2-aminophenoxazin-3-one-7-carboxylate (APOC). This compound is formed more rapidly and in greater quantity when 4-amino-3-hydroxybenzoate (4A3HB) is added to growing cultures of strain TW3 and is also formed nonbiologically when 4A3HB is shaken in mineral salts medium but not in distilled water. It is postulated that APOC is formed by the oxidative dimerization of 4A3HB, although 4A3HB has not been reported to be a metabolite of 4NT or a product of 4NB catabolism by strain TW3. Using the cloned pnb structural genes from TW3, we demonstrated that the formation of the phenoxazinone requires 4-hydroxylaminobenzoate lyase (PnbB) activity, which converts 4-hydroxylaminobenzoate (4HAB) to 3,4-dihydroxybenzoate (protocatechuate) and that 4-nitrobenzoate reductase (PnbA) activity, which causes the accumulation of 4HAB from 4NB, does not on its own result in the formation of APOC. This rules out the possibility that 4A3HB is formed abiotically from 4HAB by a Bamberger rearrangement but suggests that PnbB first acts to effect a Bamberger-like rearrangement of 4HAB to 4A3HB followed by the replacement of the 4-amino group by a hydroxyl to form protocatechuate and that the phenoxazinone is produced as a result of some misrouting of the intermediate 4A3HB from its active site. |
|
|---|
![]() View larger version (24K): [in a new window] |
FIG. 1. Metabolic pathway and genes for conversion of 4NB to PCA in P. putida TW3 (8). (A) Metabolic pathway involving 4-nitrobenzoate reductase (PnbA) and 4-hydroxylaminobenzoate lyase (PnbB). The brackets around 4-nitrosobenzoate imply that it is an intermediate, but there is no experimental evidence to support its presence. (B) Arrangement of pnb genes in TW3. pnbR is a regulator gene, and orf1 and orf2 have no known function. The inserts in plasmids pRK3.14 and pRK3.15 are shown at the bottom of the diagram.
|
|
|
|---|
|
View this table: [in a new window] |
TABLE 1. Bacterial strains and plasmids
|
Triparental matings for transfer of DNA into P. putida PaW340.
The donor strain, the recipient strain, and E. coli HB101 carrying pRK2013 as a helper plasmid were grown in LB medium until they reached an A600 of 0.6. Five hundred microliters of each culture was mixed and centrifuged, and the pellets were washed in minimal medium. The pellets were finally resuspended in 50 µl of minimal medium and dispensed onto a sterile nylon membrane (Bio-Rad) laid on the surface of an LB plate. Following incubation overnight at 30°C, the cells were washed off the filter into 2 ml of minimal medium, and appropriate dilutions were spread onto selective medium. Donor-only and recipient-only controls were treated in the same way.
Spectrophotometric analysis of the orange compound.
To analyze bacteria containing recombinant plasmids for the accumulation of the orange compound, they were grown in minimal salts medium containing 0.01% yeast extract, 0.01% Casamino Acids, and 10 mM glucose supplemented with 5 mM 4NB. Cultures of TW3 were grown in minimal medium containing 5 mM 4NB. Aliquots of the cultures were centrifuged, the supernatants were diluted 1:5 with minimal medium, and the absorption spectra between 400 and 600 nm were measured.
Isolation of the orange compound.
The yellow-orange supernatant with the characteristic absorption spectrum (
max, 430 and 446 nm) (Fig. 2A) was extracted from cultures (2 liters) with ethyl acetate (four extractions with 400 ml each time), which removed none of the colored material. Upon acidification of the material to pH 1 with concentrated HCl, the color changed to bronze, the characteristic peak became broader, and the
max shifted to a 5- to 10-nm-higher wavelength. This material was extracted again with ethyl acetate (four extractions with 400 ml each time), which produced a bright-orange organic layer with an absorption peak slightly displaced to a lower wavelength (
max, 416 and 436 nm) and which left a colorless residual aqueous layer with no major absorption peak. The solvent was evaporated from the organic layer at 14 mm Hg to give a brownish-black solid. A 1H nuclear magnetic resonance (NMR) spectrum of this solid in d6-dimethyl sulfoxide (d6-DMSO) showed one major set of signals in the aromatic region but more complex signals between
-1 and
-3, indicative of impurities. The solid was washed first with 10 ml of ethyl acetate and then with 5 ml of methanol. Comparisons of the visible spectra of the two organic washings at standard dilution showed that neither contained more than 5% of the compound, with a
max of 436 nm. The remaining solid (
20 mg) was thoroughly freed of residual solvent under a stream of N2 and then dissolved in d6-DMSO (1 ml). Dilution of the solution in methanol showed that it contained the bulk (
80%) of the product, with a
max of 436 nm and an overall absorbance matching that of the original ethyl acetate extracts. The 1H and 13C NMR spectra of the compound indicated that it was pure by both criteria and that the washings had removed the signals between
-1 and
-3: the 13C spectrum showed only the 13 carbons, and the 1H spectrum contained only signals assigned to the molecule and these were appropriately correlated to the 13C spectrum in a 1H-13C two-dimensional (2D) experiment. Column chromatography on silica, with methanol as eluting solvent, or ion-exchange chromatography on Dowex 50W-X8(H) did not lead to an increase in purity. A similar extraction and washing procedure performed on the culture supernatants of a 4NT-grown culture or on solutions of 4-amino-3-hydroxybenzoate (4A3HB) which had been shaken at 30°C gave products with the same 1H NMR spectrum as that of a 4NB-grown culture.
![]() View larger version (15K): [in a new window] |
FIG. 2. (A) Absorption spectrum of supernatants from P. putida TW3 grown on 4NT or 4NB minimal medium. (B) Measurements of turbidity (, OD600), indicating the accumulation of yellow color ( , A446), and the calculated concentration ( , concentration) of APOC causing the yellow color during growth of TW3 on 4NB minimal medium.
|
Determination of the molar extinction coefficient of APOC.
Solutions containing from 0.1 to 0.5 mM commercial 4A3HB in minimal salts medium were shaken at 30°C for 280 h until the A446 ceased to increase. The molar extinction coefficient was calculated as 22,000 M-1, assuming complete conversion of 1 mol of APOC from 2 mol of 4A3HB.
|
|
|---|
Identification of the orange by-product.
In addition to broad signals at 7.1 and 4.1 (probably arising from HO and H2N subgroups), the 1H NMR spectrum of the purified compound in d6-DMSO showed just seven resonances. The spectrum showed
H signals at 7.9 (1H, broad singlet), 7.87 (1H, d, J = 8.8 Hz), 7.75 (1H, d, J = 8.8 Hz), 6.4 (1H, s), and 6.35 (1H, s). The signals at 7.87 and 7.75 were tented towards each other, as is typical of an AB pattern. The 13C NMR spectrum showed eight quaternary and five tertiary carbons which ranged from a
c of 98 to 180 and showed two carbonyl carbons at 180 and 166 (Table 2): the proton and carbon signals were correlated by 1H-13C 2D spectroscopy (Table 2). This is extremely close to the spectrum of 2-aminophenoxazin-3-one (Table 2) (7), apart from the additional carbon at a
c of 166 and the replacement of a tertiary carbon by a quaternary carbon at a
c of 129/130. The mass spectrum of the compound was rather difficult to obtain and suggested that some impurities were present. The major molecular ion that appeared in the electron impact spectrum was measured as m/z 256.0484 (C13H8N2O4 requires 256.0484). The spectra and mass are consistent with the structure of the compound APOC (Fig. 3, compound II). The same compound was isolated from both 4NB- and 4NT-grown cells and from the abiotic dimerization of 4A3HB (see below).
|
View this table: [in a new window] |
TABLE 2. Comparison of 13C NMR signals
|
![]() View larger version (21K): [in a new window] |
FIG. 3. Proposed reactions leading to the formation of APOC (compound II) during the oxidative dimerization of 4A3HB. For comparison, the structures of 2-aminophenoxazin-3-one (compound III) formed from dimerization of 2-aminophenol (7, 20) and of 2-amino-4a,7-dimethyl-4,4a-dihydrophenoxazin-3-one (compound IV) formed by the condensation of 4-amino-3-hydroxytoluene (19) are shown.
|
![]() View larger version (16K): [in a new window] |
FIG. 4. Biological formation of APOC from 4A3HB by 4NB-induced cells of P. putida TW3. , cells grown on 5 mM 4NB; ----, cells grown on 5 mM 4NB plus 0.1 mM 4A3HB; x, cells grown on 5 mM 4NB plus 0.5 mM 4A3HB; , cells incubated with 0.5 mM 4A3HB alone.
|
![]() View larger version (18K): [in a new window] |
FIG. 5. Abiotic formation of APOC from 4A3HB by shaking 0.5 mM 4A3HB in minimal medium for 0 h ( ), 24 h (----), 48 h (x), and 72 h ().
|
max of 262 nm (data not shown). Further incubation of the culture supernatants for periods of up to 72 h caused no detectable change in their spectra from one resembling 4HAB to one characteristic of APOC. By contrast, the supernatants of PaW340(pRK3.14) did acquire the characteristic spectrum of APOC (Fig. 6). This occurs not only in the P. putida host, but also in E. coli(pRK3.14), although in both cases the peaks are broader due to the simultaneous accumulation of PCA, which was detected by its lower absorption maxima, 254 and 290 nm, and by its reaction with p-toluidine (8). E. coli, unlike P. putida, cannot further metabolize PCA, which therefore accumulates. In E. coli(pRK3.15) containing pnbA but without pnbB, no characteristic APOC spectrum was observed (Fig. 6) and only the accumulation of 4HAB was observed (data not shown), as occurs with P. putida. This suggests that 4A3HB is not being formed from an abiotic rearrangement of the hydroxylamino compound but that it is being formed as a direct result of the activity of PnbB.
![]() View larger version (12K): [in a new window] |
FIG. 6. Absorption spectrum of medium after the growth of heterologous hosts carrying cloned pnb genes in the presence of 4NB. , P. putida PaW340(pRK3.14); ----, E. coli(pRK3.14); x, E. coli(pRK3.15).
|
|
|
|---|
max, 425 and 466 nm) has been reported in cultures of other strains grown on 4NT (13) or 4NB (14).
The question we have tried to address is how 4A3HB is formed in strain TW3. The obvious answer appeared to be that 4HAB, which is a metabolite of the pathway (Fig. 1), accumulates in the medium and a proportion of it undergoes an abiotic Bamberger-type rearrangement (1, 18) to form 4A3HB, which then condenses to form APOC. If this were so, then a mutation inactivating pnbB, the gene for 4-hydroxylaminobenzoate lyase, should cause stoichiometric accumulation of 4HAB, thereby leading indirectly to the formation of APOC. We had already cloned the genes for 4NB catabolism from TW3 (Fig. 1) (8) and demonstrated that in a heterologous P. putida host (strain PaW340), they would support the complete catabolism of 4NB by providing a metabolic route to PCA which the host could utilize via its ß-ketoadipate pathway. It was therefore more practical to use the cloned genes in the heterologous host and simply delete pnbB rather than try to work with the native genes in the TW3 genome. Expecting that PaW340(pRK3.15), carrying pnbA alone (Fig. 1), would thus produce APOC when grown in the presence of 4NT or 4NB, we were surprised to find that although the medium turned yellow, the color was due to 4HAB and not APOC. Furthermore, prolonged shaking of the medium under growth conditions did not cause detectable phenoxazinone production. However, when the same experiment was repeated with the pnbA+ pnbB+ plasmid pRK3.14, there was production of the phenoxazinone during growth on 4NB; this occurred even during growth of E. coli(pRK3.14) in the presence of 4NB, although in this case, there was also considerable accumulation of PCA, which complicated the spectrum and the color of the medium. This demonstrated (i) that under the conditions of incubation, an abiotic Bamberger rearrangement of 4HAB does not take place and (ii) that PnbB is an essential factor in the production of the phenoxazinone. The most likely explanation of this is that PnbB is itself responsible for the conversion of its substrate 4HAB to 4A3HB by a mechanism similar or analogous to a Bamberger rearrangement. If some of the 4A3HB escapes from the enzyme into the medium, where we have shown that it can form APOC abiotically, this would account for the production of APOC. The final A446 of a culture grown on 5 mM 4NB (Fig. 2B) indicates that about 0.15 mM APOC is produced, accounting for a significant amount (
6%) of the substrate being misrouted in this way. The low rate of the abiotic conversion of 4A3HB to APOC (half-life, 40 to 50 h) also explains the continued production of APOC in wild-type cultures grown on 4NT or 4NB well after growth of the culture ceased (Fig. 2B) or when the cells were removed from the culture after growth has ceased if 4A3HB is released into solution during the growth phase and is then abiotically converted to APOC. An alternative explanation involving the oxidative condensation of two molecules of 4A3HB on the PnbB active site during the formation of PCA would fit the data less well, since APOC formation in this case would be growth dependent. In terms of the mechanism of action of PnbB, 4A3HB is a possible reaction intermediate between 4HAB and PCA, and the second part of the reaction would simply involve the replacement of the 4-amino group of 4A3HB with a hydroxyl (Fig. 7). There are precedents in nitroaromatic catabolism for the first part of this reaction, since Bamberger-like mutase reactions on hydroxylaminobenzenes have been described previously (3, 6, 14, 16, 19), although recent work has suggested that the mechanism of the conversion to aminophenols might be different from that of the abiotic Bamberger rearrangement (6). Recent results by J. C. Spain and his coworkers have made a similar suggestion, based on 18O labeling experiments, i.e., that PnbB initially causes the migration of the OH from the 4-hydroxylamino group to become the 3-hydroxyl group en route to PCA (L. J. Nadeau, Z. He, and J. C. Spain, Abstr. 102nd Gen. Meet. Am. Soc. Microbiol. 2002, abstr. O-31, p. 352, 2002). In their mechanism, however, the intermediate is not 4A3HB directly. These differences might be reconciled if there was a common reaction intermediate both in the PnbB-catalyzed reaction and in the abiotic formation of APOC from 4H3AB.
![]() View larger version (16K): [in a new window] |
FIG. 7. Proposed mechanism for the action of 4-hydroxylaminobenzoate lyase (PnbB) through 4A3HB as an intermediate leading to the accumulation of APOC in the medium.
|
|
|
|---|
This article has been cited by other articles:
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Copyright © 2009 by the American Society for Microbiology. For an alternate route to Journals.ASM.org, visit: http://intl-journals.asm.org | More Info»