Degradation of 3-Phenylbutyric Acid by Pseudomonas sp. - Journal of

Vol. 152, No. 1

JOURNAL OF BACTERIOLOGY, Oct. 1982, p. 411-421 0021-9193/82/100411-11$02.0O/0 Copyright 0 1982, American Society for Microbiology

Degradation of 3-Phenylbutyric Acid by Pseudomonas sp. F. SIMA SARIASLANI,'t J. L. SUDMEIER,2 AND D.D. FOCHT'* Department of Soil and Environmental Sciences,' and Department of Chemistry,2 University of California, Riverside, California 92521

Received 23 April 1982/Accepted 24 June 1982

Pseudomonas sp. isolated by selective culture with 3-phenylbutyrate (3-PB) as the sole carbon source metabolized the compound through two different pathways by initial oxidation of the benzene ring and by initial oxidation of the side chain. During early exponential growth, a catechol substance identified as 3-(2,3dihydroxyphenyl)butyrate (2,3-DHPB) and its meta-cleavage product 2-hydroxy7-methyl-6-oxononadioic-2,4-dienoic acid were produced. These products disappeared during late exponential growth, and considerable amounts of 2,3-DHPB reacted to form brownish polymeric substances. The catechol intermediate 2,3DHPB could not be isolated, but cell-free extracts were able only to oxidize 3(2,3-dihydroxyphenyl)propionate of all dihydroxy aromatic acids tested. Moreover, a reaction product caused by dehydration of 2,3-DHPB on silica gel was isolated and identified by spectral analysis as (-)-8-hydroxy-4-methyl-3,4-dihydrocoumarin. 3-Phenylpropionate and a hydroxycinnamate were found in supernatants of cultures grown on 3-PB; phenylacetate and benzoate were found in supernatants of cultures grown on 3-phenylpropionate; and phenylacetate was found in cultures grown on cinnamate. Cells grown on 3-PB rapidly oxidized 3phenylpropionate, cinnamate, catechol, and 3-(2,3-dihydroxyphenyl)propionate, whereas 2-phenylpropionate, 2,3-dihydroxycinnamate, benzoate, phenylacetate, and salicylate were oxidized at much slower rates. Phenylsuccinate was not utilized for growth nor was it oxidized by washed cell suspensions grown on 3-PB. However, dual axenic cultures of Pseudomonas acidovorans and Klebsiella pneumoniae, which could not grow on phenylsuccinate alone, could grow syntrophically and produced the same metabolites found during catabolism of 3PB by Pseudomonas sp. Washed cell suspensions of dual axenic cultures also immediately oxidized phenylsuccinate, 3-phenylpropionate, cinnamate, phenylacetate, and benzoate. The microbial degradation of detergents by soil and sewage microorganisms has received considerable attention in recent years. Although the domestic detergents deposited on soil are rapidly transformed by aerobic microorganisms, their complete degradation is considerably slower. A classic study by Swisher (23) shows that the molecular architecture of the surfactant largely determines its biological lability. Bird and Cain (Biochem. J. 127:46p, 1972) reported the accumulation of short-chain alkylsulfophenyl-alkanoates during the degradation of C1l and C12 alkyl chains of 1-phenylalkane-sulfonates by a Vibrio sp. However, 2-(p-sulfophenyl)dodecane, rather than the 1-phenyl isomer used in most biodegradation studies, comprises the major fraction of isomers in commercial detergent preparations. Consequently, sulfonat-

ed 3-phenylbutyrate (3-PB) would be the focal point in the metabolism of linear alkylbenzenesulfonates since further degradation of p-oxidation would be precluded. Indeed, there is evidence that relatively stable intermediates may be formed which then persist in soil and water (16). These intermediates presumably contribute to the much higher incidence of sulfonate groups in sludge-derived fulvic acids than in fulvic acid polymers normally found in nature (21, 22). Previous studies on phenylalkanes and alkyl-

benzenesulfonates have indicated that these compounds are degraded via two different routes by soil and sewage microflora depending on the length of the alkyl sidechain. Bird and Cain (3) reported the degradation of alkylbenzenesulfonates with an alkyl chain of less than C4 by oxidation to orthodiols followed by their without prior side chain attack by an t Present address: Department of Medicinal Chemistry and cleavage sp. Baggi et al. (2) also showed that Alcaligenes of Natural Products, College of Pharmacy, The University the aromatic ring of 3-phenylpentane can be Iowa, Iowa City, IA 52242. 411

412

SARIASLANI, SUDMEIER, AND FOCHT

J. BACTERIOL.

Ultrasonics, Inc., Plainview, N.Y.), and the resulting suspensions were centrifuged at 35,000 x g for 20 min at 4°C. For studies of dioxygenases, 10% (vol/vol) ethanol was added to the cell suspension before sonication. Protein was determined by the method of Lowry et al. (17), with bovine serum albumin as the standard. Catechol 2,3-oxidoreductase (EC 1.13.11.2) and catechol 1,2-oxidoreductase (EC 1.13.11.1) were assayed as described by Bird and Cain (3). 3-(2,3-Dihydroxyphenyl)propionate dioxygenase was assayed as deMATERIALS AND METHODS scribed by Dagley et al. (8). Other dioxygenases were Cultural methods. A bacterium was isolated from assayed in a Gilson respirometer. Flasks contained, in California soil by selective culture on 3-PB, main- a total volume of 3 ml: 2.5 ml of sodium phosphate tained on agar slants containing 3-PB (3 g liter-') as buffer (pH 7), cell extracts (3 to 5 mg of protein), the sole carbon source, and grown at 28°C in liquid glutathione (1 ,ug), and FeSO4 (1 ,g) in the main culture in a 6-liter-capacity stiffed fermentor (Micro- compartment; 100 ~Lmol of substrate in the side arm; ferm; New Brunswick Scientific Co., New Brunswick, and 0.2 ml of NaOH, 20% (wt/vol), in the center well. Separation and isolation of metaboites. Cells were N.J.) with an air flow rate of 4 liters min 1 or on a gyratory shaker in 2-liter conical flasks, plugged with grown in 10 liters of media described previously and cotton wool and containing 1 liter of growth medium. removed during mid-exponential growth by centrifugaGrowth medium contained (in grams per liter): tion (10,000 x g, 10 min, 25°C). The supernatant was NaH2PO4, 4.77; MgSO4, 0.12; K2HPO4, 1.36; acidified with HCI to pH 2 and extracted three times (NH4)2SO4, 1.32; Ca(NO3)2, 0.0005; FeSO4, 0.0005; with 0.5 volumes of diethyl ether. The extracts were CuS04, 0.0005, NaMo04, 0.0005; and yeast extract, 50 combined and dried over anhydrous MgSO4 before removal of the ether under reduced pressure. The mg. A mixed culture was obtained from sewage by residue was then taken up in 5 ml of diethyl ether. selective enrichment with phenylsuccinate as the sole Small amounts were applied to silica gel (100-,m carbon and energy source, using the growth medium thickness) GF 254 thin-layer chromatography plates mentioned above. After several serial transfers, at- (Eastman-Kodak, Rochester, N.Y.), whereas the matempts to isolate colonies on a defined agar medium jority was saved for large-scale separation by column containing phenylsuccinate were futile because of the chromatography. Plates were developed in a solvent apparent failure of a single strain to utilize the sub- system of benzene-dioxane-acetic acid (45:12:1, vol/ strate for growth. Two distinct colony forms, howev- vol/vol). When sprayed with 1% (vol/vol) ethanolic er, were noted when the mixed culture was streaked Gibbs reagent (2,6-dichloro-p-benzoquinone4chlorionto nutrient agar. Inoculation from both colony types mine) followed by saturated aqueous NaHCO3, two produced a dual axenic culture capable of growth in monohydroxylated aromatic compounds (blue color) defined liquid media containing phenylsuccinate as the with Rf values of 0.9 and 0.5 and one dihydroxylated sole carbon source. Pure cultures were lyophilized for derivative (brown color) at Rf 0.3 could be detected. The ether extracts were applied to a column (66 by long-term preservation. They were maintaind for current experimentation by transfer on nutrient agar 4.5 cm) of silica gel (100 mesh; Mallinckrodt Co., Paris, Ky.). Elution with benzene-dioxane-acetic acid slants and in defined liquid media. Cell density was determined on a Bausch & Lomb (40:1:0.1 vol/vol/vol) resulted in the separation of Spectronic 20 spectrophotometer at 510 nm and relat- hydroxylated aromatic acids from polymeric subed to dry weight by using a standard curve prepared stances. Fractions containing the hydroxylated acids with 3-PB-grown organisms. To avoid interference were combined and concentrated by removing the from the colored substances in the medium, portions solvent under reduced pressure. Trace amounts of a (50 ml) of the culture were centrifuged, washed once monohydroxylated compound (Rf 0.9 on thin-layer with 10 mM phosphate buffer (pH 7.2), and suspended chromatography) were present early in the elution. in 50 ml of the same buffer. These suspensions were These fractions were combined, solvent was evaporated under reduced pressure, and the residue was ethylused for optical density determination. Reo ated and subjected to gas chromatographic-mass spec. Washed cell suspensions were prepared from cells harvested during log phase by centrif- trometric (GC-MS) analysis. 8-Hydroxy4-methylugation (10,000 x g, 10 min, 25°C), washed twice with 3,4-dihydrocoumarin (Rf 0.5 on thin-layer chromatog10 mM sodium phosphate buffer (pH 7.2), and sus- raphy), which resulted from decomposition of the pended in the same buffer. Oxygen consumption was dihydroxylated compound (Rf 0.3 on thin-layer chrorecorded at 30°C with a Gilson respirometer (Gilson matography), was completely separated from the other Medical Electronics, Middleton, Wis.). Unless other- hydroxylated compounds by several passages through wise stated, incubation mixtures contained 0.1 M the silica gel column, and fractions were combined, sodium phosphate buffer, pH 7.2 (2 ml), cell suspen- evaporated to dryness, and recrystallized from hexsion (3 to 10 mg, dry weight), and substrate (100 FLmol) ane. Presumed polymers and oligomers were eluted in a total volume of 3.0 ml. Respirometer flasks from the silica gel column by increasing the proporcontained 0.2 ml of 20% (wt/vol) NaOH in the center tions of dioxane and acetic acid in the solvent system. The fractions were set aside for spectral analysis. well to absorb CO2. Phenolic and aromatic acids were converted to their Enzyme asays. Extracts were prepared by sonication of washed cell suspensions (3 x 20-s pulses) in a ethyl esters by treatment with diazoethane prepared Sonifier cell disruptor (model W 140; Heat Systems- by the method of Fales et al. (10) and were separated

degraded without prior side chain attack. Metabolism of long-chain alkylbenzenesulfonates and phenylalkane is initiated by w-oxidation of the terminal methyl group in both hydrocarbons (24) and surfactants (14). After w-oxidation, the alkanoate side chain is degraded by a- or 1oxidation. This communication describes the degradation of 3-PB via both pathways by a Pseudomonas sp. isolated from soil.

VOL. 152, 1982

413

3-PB DEGRADATION BY PSEUDOMONAS SP.

by using a prepacked, coiled stainless-steel column (180 cm) containing 3% OV101 on Gas-Chrom Q (100 to 120 mesh). A Varian gas chromatograph (series 2700) with a flame ionization detector was used under the following conditions: N2 carrier gas flow rate, 30 to 50 ml min-'; oven temperature, 130 to 180°C; injector temperature, 190°C; detector temperature, 225°C. To obtain a sufficient quantity of the ring fission product, the isolate was grown in 200 ml of nutrient broth for 24 h, harvested aseptically by centrifugation, washed aseptically, and suspended in 200 ml of sterile basal salts medium containing 0.6 g of 3-PB as the sole carbon source. The procedure was necessary to prolong the catabolism of the ring fission product. Cells were removed after 48 h by centrifugation, and the supernatant was acidified to pH 3.5 and extracted three times with 0.5 volumes of ethyl acetate. The solvent was removed under vacuum, and the oily residue obtained was dissolved in a small volume of methanol and applied to a glass column (21 by 0.5 cm) containing DEAE-cellulose (DE52; Whatman Biochemicals Ltd., Springfield Mill, Maidstone, Kent, U.K.). The column was eluted with 0.1 M sodium phosphate buffer, pH 7.2. Fractions (10 ml) were collected, and those containing the yellow product were combined, acidified to pH 3.5, and extracted three times with 0.5 volumes of ethyl acetate. The solvent was removed under vacuum, and the residue obtained was converted to a pyridine derivative by the method of Adachi et al. (1). For mass spectral analysis, the pyridine derivative was methylated with diazomethane prepared by the method of deBoer and B$acker (9). Analytical methods. Free aldehyde groups were assayed with 2% p-phenylenediamine, o-dianisidine, and 1,2-dianilinoethane as described by Feigel (11). The presence of keto groups was determined by the method of Friedman and Haugen (12). Catechols were assayed by the method of Evans as modified by Bird and Cain (3). Infrared spectra were recorded as KBr disks (400 mg) with a Perkin-Elmer 621 grating spectrophotometer. UV spectra were obtained with a recording Beckman DB-G spectrophotometer. Mass spectra were obtained from a Finnigan 4000 mass spectrometer by using either the directsolids probe inlet for pure samples or interfaced GCMS with mixtures of volatile derivatives, using GC conditions identical to those described previously, except that He was used as the carrier gas. Proton nuclear magnetic resonance (NMR) spectra were obtained at 500 MHz on a Bruker WM500 FTNMR spectrometer and at 90 MHz on a Varian EM-390 NMR spectrometer, using 5-mm-outer diameter sample tubes with CDCl3 as the solvent. Natural abundance 13C-NMR spectra were obtained on a Bruker WH90 FTNMR spectrometer, using a 10-mm-outer diameter sample tube with CDC13 as the solvent. Chemicals. 3-(3,4-Dihydroxyphenyl)propionic acid, o-hydroxycinnamic acid, N-ethyl-N'-nitro-N-nitrosoguanidine, 1,2-dianilinoethane, and phenylsuccinic, 3phenylbutyric, cinnamic, and 3-phenylpropionic acids were obtained from Aldrich Chemical Co., Milwaukee, Wis. 3,4-Dihydroxycinnamic acid and 2,6-dichloro-quinone-4-chlorimide were purchased from Sigma Chemical Co., St. Louis, Mo. Sodium benzoate, salicylic acid, and phenol were obtained from Mallinckrodt Co.

3-(2,3-Dihydroxyphenyl)propionate and 2,3-dihydroxycinnamic acids were gifts from P. J. Chapman. RESULTS

Growth and characteristics of the isolate utilizing 3-PB. The isolate obtained by selective culture with 3-PB was a short, gram-negative rod with polar flagella that was both oxidase and catalase positive and reduced both nitrate and nitrite. Colonies on nutrient agar were smooth, off-white, and opaque. Neither acid nor gas was produced from glucose. The isolate was thus identified as a Pseudomonas sp. (4). During the lag and early exponential phases on 3-PB, a yellow color appeared in the medium. Chemical tests also indicated the transient formation of a catechol substance, which reached its maximum concentration (750 ,ug ml-l) during the exponential growth phase (Fig. 1). At this stage, a brown color developed in the medium. The catechol substance then decreased rapidly and disappeared after 48 h, and the culture maintained its brown color. This suggested that the catechol was being polymerized into colored complexes. Identification of metabolites produced during growth of Pseudomonas sp. 3-Phenylpropionate, phenylacetate, and a hydroxycinnamate were produced when Pseudomonas sp. was grown on

4 0

T

-t

w

L()

cn

z

-i w

m cn 0

-J

0 I w

3. a Cr

C..

0

10

20

30

40

50

HR FIG. 1. Growth of Pseudomonas sp. formation of the catechol substance.

60

on

3-PB and

414


TABLE 1. Ethyl esters of metabolites identified by GC-MS in culture supernatants of Pseudomonas sp. Growth substrate

Compound identified

Rete(ntio

6 3-Phenylpropionate 3.3 Phenylacetate 9.8 2,3-DHPB Not deter3-(Hydroxymined phenyl)butyrate Hydroxycinnamate Not determined

3-PB

3-Phenylpropionate Phenylacetate Benzoate

3.3 2.5

Cinnamate

3.3

Phenylacetate

3-PB (Table 1). Mass spectral data of ethylated derivatives of these compounds were identical to those of authentic derivatives. However, the three isomers of hydroxycinnamate are identical; thus, the position of the hydroxyl group could not be resolved by mass spectrometry. GC-MS analysis also indicated the presence of another compound having an apparent molecular ion peak at mle 280. The major fragmentation pattern was as follows: mle 266, -CH2; mle 265, -CH3; mle 249, -C2H5; mle 224, -2C2H4; mle 206, -C2H5, -C2H50, or -COOC2H5, -H; mle 164, -C3H6COOC2H5, -H; mle 105 (base peak), quinotropylium rearrangement ion (C7H50). The spectrum is indicative of an aromatic acid containing a C4 side chain with two hydroxyl groups attached to the benzene ring. The compound was tentatively identified as 1ethyl-3-(O,O'-diethylphenyl)butanoate. GC-MS analysis of the ethylated monohydroxy derivative obtained from the silica gel column gave a molecular ion peak at mle 236 (7% of the base peak intensity). Major fragment ions and relative intensities were as follows: mle 206 (14%), -C2H5, -H; mle 178 (42%),

J. BACTERIOL.

-2C2H5; mle 163 (14%), -COOC2H5; mle 162 (100%), -C2H5, -C2H50; mle 146 (71%), -2C2H5O; mle 136 (85%), rearrangement ion, -C2H4, -COOC2H2; mle 91 (57%), tropylium rearrjingement ion. The molecular ion and fragmentation pattern suggests that the compound is 1-ethyl-3-(O-ethylphenyl)butanoate. In 3-phenyipropionate cultures, phenylacetate and benzoate were detected. These compounds had mass spectra identical to those of authentic ethylated samples. Growth on cinnamate as the sole carbon source resulted in the formation of phenylacetate (Table 1). Isolation, crystaflization, and characterization of 8-hydroxy-4-methyl-3,4-dihydrocoumarin. Two-dimensional thin-layer chromatography indicated that the dihydroxylated 3-PB product was unstable on silica gel and dehydrated t6o produce 8-hydroxy-4-methyl-3,4-dihydrocoumarin. White feathery crystals (70 mg) of this compound, obtained by the procedure described in Materials and Methods, had an mp of 77 to 79°C and [a]D26 = -32° in ethanol. Elemental analysis gave the following results: C, 67.44; H, 5.71, (calculated for C1oH1003: C, 67.41; H, 5.66). The mass spectrum gave a molecular ion peak at mle 178 (62% relative intensity of base peak). The major fiagmentation pattern was as follows: mle 163 (4.9%6), -CH3; mle 149 (9.9o), -CHO; mle 136 (100%o), -CH2CO; mle 107 (16.8%), quinone ion; mle 78 (7.9%), phenyl ion. Infrared spectroscopy of this compound (Fig. 2) indicated the presence of a hydroxyl group with an OH stretch at 3,400 tQ 3,500 cm' , a saturated lactone at 1,750 cm-l, a methyl group C-H stretch at 2,940 to 3,000 cm_1, a CH2 stretch at 2,840 and 2,920 cm-, and an aromatic nng with H-scissoring vibrations at 690 to 880 cm1. NMR spectra of 8-hydroxy-4-methyl-3,4-dihydrocoumarin are shown in Fig. 3, 4, and 5. Figure 3B shows an expanded view of the aromatic region of the 500-MHz proton spectrum

z

w

0

H

cr 4-

Cl) 20-

z

I*o

WAVENUMBER (cm-1) FIG. 2. Infrared spectrum of 8-hydroxy-4methyl-3,4-dihydrocoumarin.

VOL. 152, 1982 VOL. 1982 DEGRADATION BY PSEUDOMONAS SP. ~3-PB 152,

A

*

* _______________________________ 8 7 6 5 4 3 2 6

_________________________________

415

theoretical calculations, as are the negative ~~~~~~~~~~~by

signs (15).

Proton nuclear Overhauser enhancement difference spectra at 500 MHz, in which proton C is irradiated, produced a greater increase in the multiplet at 2.6 ppm than the one at 2.9 ppm, that the intramolecular A to C distance is smaller than the B to C distance. This enables us to assign the stereochemistry of protons A and B relative to C as shown in the structural formula in Table 2, although the absolute configuration is still unknown. Elementary consideration of the Karplus equation in relation to molecular models indicates dihedral angles of 140 and 200 for AC and AB, respectively. Figure 4B shows an expanded view of the aromatic region at 90 MHz. When the multiplet at 3.2 ppm is strongly irradiated, fine splittings collapse (Fig. 4C), proving the existence of long-

~~~~~~~~~~showing

A

C

3.4

3.2

3.0

2.8

2.6

*

8, ppm FIG. 3. Proton NMR spectra of 8-hydroxy-4methyl-3,4-dihydrocoumarin at 500 MHz. Peaks marked with asterisks arose from extraneous sources. (B) and (C) are expanded views of (A) together with peak integration.

together with peak integrations. There are three distinct aromatic protons. Two are roughly doublets and the other is roughly a triplet, all possessing the ~-8-Hz spacing characteristic of ortho spin couplings. The only arrangement of

B

three aromatic protons which satisfies these .ow wow. observations is that in which all three are contiguous as shown in the structural formula (E, F, and G) of Table 2. (Moving the hydroxyl group C to C-S is excluded by the nuclear Overhauser effects described below.) Proton nuclear Overhauser enhancement difference spectra at 500 MHz reveal a pronounced increase in the intensity of the most upfield of the aromatic multiplets relative to the other two upon irradiation of proton C. This observation proves the location of the hydroxyl group to be at C-8, rather than at C-S. Although it may appear curious that the six-bond spin-coupling constant JCG (-0.6 Hz) is greater in absolute FIG. 4. Proton NMR of 8-hydroxy-4-methvalue than the five-bond coupling constant JCF yl-3,4-dihydrocoumarin atspectra 90 MHz. (B) (aromatic re(:S0.1 Hz), it has been observed in other mem- gion) and (C) (spin-decoupled spectrum) are expanded bers of the coumarin family (18) and is supported views of (A).

416


J. BACTERIOL.

TABLE 2. Chemical shifts and spin-spin coupling constants of 8-hydroxy4 methyl-3,4-dihydrocoumarin.

OH (K)

H Chemical Shifts (ppm vs. tns)

A

B

2.609

2.867

C

3.190

D

E

1.345

6.751

F 7.012

G

K

6.904

6

H- H Spin Coupling Constants (Hz) AB -15.9

AC 7.3

3C Chemical

CF

BC

CD

CE

5.5

W

-0.8~ ,0.1

Shifts (ppm

CG -0. 6

EF

7

EG

r.5

FG

V8s. tms)

2

3

4

4-CH3

167.5

36.9

29.6

19.8

5, 6, 7 115.1,117.4,125.0

range couplings of proton C (Table 2) to the two upfield aromatic resonances centered at 6.9 and 7.0 ppm. Figure 5 shows the 13C-FTNMR spectrum with broad-broad proton decoupling (in Fig. 5B). The carbonyl C-2 at 167.5 ppm is easy to assign because of its characteristic downfield shift and small relative intensity. The six aromatic carbons fall into two groups: those having no protons directly attached, and those having a proton. The two groups are easily distinguished because of the nuclear Overhauser enhancement enjoyed by the latter group in Fig. 5A and the splitting into doublets of the latter group in Fig. 5B. Respirometric studies with washed cell suspensions. Comparative oxidation rates of possible intermediates by washed cells grown on 3-PB or on glucose (Table 3) are considerably higher for 3-PB-grown cells than for glucose-grown cells. Organisms grown on 3-PB could not oxidize 3-(3,4-dihydroxyphenyl)propionic acid, phenylsuccinate, or 3,4-dihydroxycinnamate. When mhydroxybenzoate, p-hydroxycinnamate, and phydroxybenzoate served as substrates for washed cell suspensions, the oxygen uptake rates were very low. Ring fissuon product of 2,3-DHPB. When Pseudomonas sp. was grown with 3-PB as the sole carbon source, the medium became yellow after

8, 9, 10 128.5,138.7,143.9

24 h. Spectral analysis of the culture supernatant indicated maximum absorption between 390 and 420 nm with 0.1 M NaOH (pH 7.1 to 13) (Fig. 6). Upon acidification of the supematant with 0.1 M HCl (pH 1 to 6), the maximum absorption was shifted to 310 to 330 nm, indicating a keto-enol tautomerism. This is consistent with observations of ring cleavage products of 2-phenylbutane (2) and 3-phenylpropionate (8). The yellow product contained at least one keto group (12) but no aldehyde groups (11). These data indicate that 3-(2,3-dihydroxyphenyl)butyric acid (2,3DHPB) is cleaved between aryl C-1 and C-2 to produce 2-hydroxy-7-methyl-6-oxononadioic2,4-dienoic acid (2-hydroxy-7-methyl-6-oxo-2,4dienoic azelaic acid). Because of the unstable nature of this product, a stable pyridine derivative was made since meta-ring fission products are known to produce pyridine derivatives in aqueous ammonia solutions (8). When the absorption of this compound was monitored in the region of 240 to 400 nm, it showed a single peak having Xkn at 270 nm with a shoulder at 264 nm, which was unaffected by changes in pH: this is characteristic of a-picolinate derivatives (25). Further derivatization of the pyridine derivative was necessary for mass spectral analysis. This was achieved by methylation of the compound with diazomethane as described previously. The mass spectrum of this derivative


VOL. 152, 1982

417

A

II

I

B

*

200

1-"; L

III I

I

180

JJJIJi,

AILILLL

-

I

160

140

120

00

8,

80

60

40

20

0

ppm

FIG. 5. 13C-FTNMR spectra of 8-hydroxy-4-methyl-3,4-dihydrocoumarin, with (A) broad-broad proton decoupling and (B) off-resonance continuous-wave proton decoupling. The asterisked peaks at 0 and 789 ppm represent resonances from trimethylsilyl internal standard and solvent CDCl3, respectively.

indicated an apparent molecular ion peak at mle 237 (4.4%) of the base peak intensity. The major fragmentation pattern was as follows: mle 206 (11.8%), -OCH3; mle 178 (67%), -COOCH3; mle 164 (22.6%), -CH2COOCH3; mle 132 (8.7%), -H, -CH2COOCH3, -OCH3; mle 105 (100%), 2-vinylpyridine ion; mle 91 (34%), -CH2 from 2-vinylpyridine ion. All chemical and spectral data obtained on the ring fission product support the compound having the structure indicated in Fig. 7. Metabolism of phenylsuccinate by dual axenic cultures of Pseudomonas acidovorans and Kkebsiella pneumoniae. Dual axenic cultures isolated from sewage were capable of using phenylsuccinate as the sole source of carbon and energy, yet were unable to utilize the substrate when each was incubated with it alone. One isolate was an obligately aerobic short gram-negative rod that was motile by polar flagella, was oxidase and catalase positive, produced acid but no gas from glucose, did not reduce nitrates, and did not hydrolyze gelatin, starch, or urea. Production of indole, sulfide, and acetylmethylcarbinol was negative. The isolate grew at pH 4.4 (although not as well as at pH 7.0) and utilized

TABLE 3. Oxidation of possible intermediates by washed cell suspensions of Pseudomonas sp. grown on 3-PB or glucose Oxidation (1J of 02 mg, dry Wt-1 h-1)a Substrate by cells grown on: 3-PB

Phenylsuccinate 3-Phenylpropionate 2-Phenylpropionate Cinnamate Catechol 2,3-Dihydroxyphenylpropionate 3,4-Dihydroxyphenylpropionate 2,3-Dihydroxycinnamate 3,4-Dihydroxycinnamate Benzoate

Phenylacetate

3-PB

Glucose

104 0 148 32 114 112 84 0 24 0 34 41 5 45 4.5 17

25 0 14 24 33 36 17 0 0 3.2 12 28 0.5 40 0 0

p-Hydroxycinnamate Salicylate m-Hydroxybenzoate p-Hydroxybenzoate a Oxidation was measured for 1 h in both cases and corrected for endogenous respiration (10 and 19 when grown on glucose and 3-PB, respectively).

418


J. BACTERIOL.

TABLE 4. Oxidation of possible intermediates by mixed cultures of P. acidovorans and K. pneumoniae grown in phenylsuccinate or succinate Oxidation (jji of 02 mg, dry

wt!1 h-1)a

Substrate

Phenylsuccinate

Succinate

0.6 41 Phenylsuccinate 5 32 Cinnamate 3 250 Benzoate 4 36 3-Phenylpropionate a Oxidation was measured for 1 h in both cases and corrected for endogenous respiration. w 0 z

0.5-~~~~~~~1.

im

cr0.5 0

3.2 ~~6

300

340

380

420

460

WAVELENGTH (nm) FIG. 6. Absorption spectra of meta-ring cleavage product. Numbers associated with curves indicate different pH values. Shift in Xa. between 390 and 420 nm from alkaline conditions (pH 7.1 to 13) to X,,,, between 310 and 330 nm upon acidification (pH 1 to 6) indicates keto-enol tautomerism.

maltose and lactate for growth. Colonies on nutrient agar were round, entire, smooth, offwhite, and opaque; greenish-yellow diffusible pigment was produced in the medium. These characteristics are consistent with P. acidovorans (4). The other isolate was a gram-negative, nonmotile rod that was oxidase negative and catalase positive and produced acid and gas from glucose, lactose, and maltose. Nitrate but not nitrite was reduced. Neither gelatin nor starch was hydrolyzed. Indole, sulfide, and acetylmethylcarbinol were produced. Urea was hydrolyzed, and growth occurred in the presence of KCN. Citrate, but not malonate or phenylalanine, were utilized for growth. The isolate was methyl red positive and grew either aerobically or fermentatively. Colonies on nutrient agar were smooth, off-white, round, and entire.

These characteristics are consistent with K. pneumoniae (4). Production of 3-phenylpropionate during growth of Pseudomonas sp. on 3-PB suggests woxidation to produce phenylsuccinate, which is then decarboxylated to 3-phenylpropionate. Although it was not possible to detect any phenylsuccinate during 3-PB degradation, an indirect approach was used to examine the possibility of phenylsuccinate involvement in the degradation of 3-PB. GC-MS analysis of the dual axenic culture supematant indicated the presence of 3-phenylpropionate, cinnamate, and benzoate in these cultures. When 2-phenylpropionate served as the sole carbon source for these organisms, phenylacetate, benzoate, and benzaldehyde were detected in the culture supernatants. Respirometric studies with the dual axenic culture grown on phenylsuccinate indicated that they were readily adapted to oxidize phenylsuccinate, cinnamate, benzoate, and 3-phenylpropionate (Table 4). Enzymatic activities in organisms grown on different substrates. Cells of Pseudomonas sp. grown on 3-PB possessed higher m-pyrocatechase activity than those grown on cinnamate, whereas growth on phenylacetate or benzoate as sole carbon source did not result in the induction of m-pyrocatechase activity (Table 5). Organisms grown on benzoate had no m-pyrocateTABLE 5. Enzymatic activities with catechol in cell-free extracts of Pseudomonas sp. grown on differtnt substrates Activity (nmol mg of protein-' min-') Growth m-Pyrocatechase o-Pyrocatechase substrate

3-PB Cinnamate Phenylacetate Benzoate Glucose

(catechol 2,3-oxidoreductase) 90

(catechol 1,2-oxidoreductase) 0

39 0 0 0

0 0 10 0


VOL. 152, 1982 H

419

H

H3C-C -C H2-COOH

H

H3C-C-CH2-COOH C=O COOH

VIA

"META-CLEAVAGE PATH WAY

OH

I I 0~~~~~~~1 CH2-CH2-COOH

I CH=CH-COOH

CH2COOH

|POLYMERIZATION]

COOH

-t FIG. Proposed route

H . .

OH

C Ho

the degra

,___VIA

META-CLEAVAGE

PATHWAY

OH

FIG. 7. Proposed route for the degradation of 3-PB by Pseudomonas sp.

chase activity and very low o-pyrocatechase activity. A search in cell extracts for any activity with dihydroxy-phenylacetates, dihydroxybenzoates, and dihydroxycinnamates was not successful, but when 3-(2,3-dihydroxyphenyl)propionate was used as the substrate for the enzyme, very low levels of activity were observed (7 nmol mg of protein-1 min-1). No activity was observed with 3-(3,4-dihydroxyphenyl)propionic acid. Mass spectrometry of oligomer. An oligomer from the culture medium of Pseudomonas sp. grown on 3-PB was isolated as a separate band from silica gel columns. When subjected to direct-probe mass spectrometry, an apparent molecular ion peak of mle 568 (1%) was noted: this would correspond to a trimer consisting of two dihydroxy-3-PB units and one monohydroxy-3-PB unit. The major fragmentation pattern was as follows: mle 536 (5%), loss of two 0; mle 389 (5%), loss of hydroxy-3-PB; mle 164 (93%), loss of CH3 from hydroxy-3-PB; mle 163 (100%), loss of 0 or CH3 and H from hydroxy-3PB; mle 115 (66%), indene rearrangement ion. DISCUSSION On the basis of our data, Pseudomonas sp. appears to degrade 3-PB via two different routes (Fig. 7): (i) dioxygenation of 3-PB to form 2,3DHPB, which is cleaved by an m-pyrocatechase to form 2-hydroxy-7-methyl-6-oxononadioic2,4-dienoic acid; (ii) w-oxidation and subsequent decarboxylation to produce 3-phenyl-

propionate and a-oxidations to produce phenylacetate and benzoate. Further degradation of benzoate probably proceeds through catechol in light of respirometric and enzymatic data. Since this organism has both m- and o-pyrocatechases, a trait common to pseudomonads (7), the exact catabolic pathway from catechol is unclear; however, the higher m-pyrocatechase activity induced by growth on 3-PB versus the low o-pyrocatechase activity induced by growth on benzoate would be more in accordance with the scheme shown in Fig. 7. The initial steps in the dioxygenase pathway probably involve peroxygenation and subsequent reduction to form 3-PB dihydrodiol. Formation of dihydrodiol during the degradation of 2-phenylbutane by two Pseudomonad strains has been reported by Baggi et al. (2). In our studies, isolation of 3-PB dihydrodiol was not possible, but detection of a monohydroxy-3-PB, which is formed from acid-catalyzed dehydration of the dihydrodiol (13), strongly supports the hypothesis that the step between 3-PB to 2,3DHPB is through its dihydrodiol. Since 2,3DHPB was unstable and could not be obtained in its pure form, it could not be used as a substrate in experiments with cell-free extracts. Nevertheless, extracts of cells grown on 3-PB were active only with catechol and, to a lesser extent, 3-(2,3-dihydroxyphenyl)propionate, whereas they were inactive with 3-(3,4-dihydroxyphenyl)propionic acid and all of the other dihydroxylated aromatics tested.

420


Some of the 2,3-DHPB polymerizes into higher-molecular-weight complexes. Oxidative coupling reactions with aromatic compounds resulting in the formation of polymers have been reported previously (5, 20). This type of polymerization is not only catalyzed by iron- or copper-containing phenol oxidases, but also by a wide range of oxidizing chemicals present in the growth medium. Since growth of Pseudomonas sp. on 3-phenylpropionate or cinnamate did not result in the formation of brownish-colored polymers, or in the detectable formation of catechol products, polymer formation apparently arises from the formation of 2,3-DHPB. Prolonged incubation of the cultures did not result in the disappearance of the oligomers and polymers formed. Formation of these stable polymers is of considerable importance since biodegradation of 2-phenyldodecanesulfonate by soil microflora could lead to the formation of recalcitrant complexes which persist in the environment. This hypothesis is also strongly supported by the finding that incubation of soil mixtures with 3PB but not with other phenylalkanoic acids leads to the formation of recalcitrant brown "fulvic like" polymers (Focht, unpublished data). Degradation of 3-PB through 3-phenylpropionate, phenylacetate, and benzoate suggests the involvement of an a-oxidation mechanism operating from 3-phenylpropionate to benzoate. A similar mechanism has been reported by Sariaslani et al. (19) during studies on the degradation of phenylalkanes by Nocardia salmonicolor. Formation of 3-phenylpropionate from 3-PB probably involves w-hydroxylation of the methyl group of the latter compound to phenylsuccinate. Production of phenylsuccinate from 4- or 5-phenyldecane-p-sulfonates has been discussed by Cain (6), who suggests that further degradation of phenylsuccinate could result in the formation of 3-phenylpropionate which could then undergo a- or "-oxidation to produce phenylacetate or benzoate, respectively. Despite the inability to find phenylsuccinate in culture supernatants of Pseudomonas sp. grown in 3-PB and despite the failure of 3-PB cells to oxidize it, phenylsuccinate cannot be precluded as a metabolite since it may be impermeable to the cell. Dual axenic cultures of P. acidovorans and K. pneumoniae, when grown on phenylsuccinate, produced the same products (i.e., 3-phenylpropionate, cinnamate, phenylacetate, and benzoate) that Pseudomonas sp. did when grown on 3PB. Moreover, washed cells of the dual axenic culture grown on phenylsuccinate readily oxidized these compounds. The 8-hydroxy-4-methyl-3,4-dihydrocoumarin that was formed from 2,3-DHPB during the extraction and isolation procedure was optically active, [a]D26 = -32°, whereas the 3-PB used as

J. BACTERIOL.

the growth substrate was a mixture of both

optical isomers. Analysis of the culture supernatants at the end of the growth phase indicated the complete metabolism of the starting material. Thus, either both optical isomers were degraded via the 2,3-DHPB pathway with subsequent accumulation of the (-) isomer because of

resistance to further degradation, or only the (-) isomer was degraded through the 2,3-DHPB pathway while the (+) isomer was attacked by the w-hydroxylase system through the 3-phenylpropionate route. Further research on the whydroxylase in the Pseudomonas sp. would be required to prove one of these hypotheses unambiguously. Such information would be useful in assessing the importance of stereospecificity in pathways of aromatic catabolism. ACKNOWLEDGMENTS This work was supported by a grant from the Kearney Foundation of Soil Science. The NMR investigation was supported by Public Health Service grant 1RO1-GM25877 from the National Institutes of Health. The Bruker WM500 NMR spectrometer belongs to the Southern California Regional NMR Facility at the California Institute of Technology. Experimental assistance from William Croasmun, California Institute of Technology, for NMR studies is gratefully acknowledged. F.S.S. thanks D. T. Gibson for helpful discussions, J. P. Rosazza for allowing her to complete part of this work in his laboratory, and P. J. Chapman for helpful discussions and for the gift of 3-(2,3-dihydroxyphenyl)propionate and 2,3-dihydroxycinnamate. LITERATURE CITED 1. Adachl, K., Y. Takeda, S. Senoh, and H. Kita. 1964. Metabolism of p-hydroxyphenylacetate in P. ovalis. Biochim. Biophys. Acta 93:483-493. 2. Baggl, G., D. Catelani, E. Gall, and V. Treccani. 1972. The microbial degradation of phenylalkanes:2-phenylbutane, 3-phenylpentane, 3-phenyldodecane and 4-phenylheptane. Biochem. J. 126:1091-1097. 3. Bfrd, J. A., and R. B. Cain. 1974. Microbial degradation of alkylbenzene-sulfonates: metabolism of homologues of short alkyl-chain length by an Alcaligenes sp. Biochem. J. 140:121-134. 4. Buchanan, R. E., and N. E. Gibbons (ed.). 1974. Bergey's manual of determinative bacteriology, 8th ed. The Williams & Wilkins Co., Baltimore. 5. Bollag, J. M., R. D. Sjoblad, and R. D. MInard. 1977. Polymerization of phenolic intermediates of pesticides by a fungal enzyme. Experientia 33:1564-1566. 6. Cain, R. B. 1976. Surfactants biodegradation in waste waters, p. 283-327. In A. G. Callely, C. F. Forster, and D. A. Stafford (ed.), Treatment of industrial effluents. John Wiley & Sons, Inc., New York. 7. Dagley, S. 1971. Catabolism of aromatic compounds by microorganisms. Adv. Microb. Physiol. 6:1-46. 8. Dagley, S., P. J. Chapman, and D. T. Gibson. 1965. The metabolism of 13-phenylpropionic acid by an Achromobacter. Biochem. J. 97:643-649. 9. deBoer, T. J., and H. J. Backer. 1963. Diazomethane, p. 250-253. In N. Rabjohn (ed.), Organic synthesis coll, vol. 4. John Wiley & Sons, Inc., New York. 10. Fales, H. M., T. M. Jaovni, and J. F. Babshak. 1973. Simple device for preparing ethereal diazomethane without resorting to codistillation. Anal. Chem. 45:2302-2303. 11. Feigel, F. 1943. Laboratory manual of spot tests. Academic Press, Inc., New York.

VOL. 152, 1982


12. Friedman, T. E., and G. E. Haugen. 1943. Pyruvic acid. H. The determination of keto acids in blood and urine. J. Biol. Chem. 147:415-442. 13. Gibson, D. T., V. Mahadevan, and J. F. Davey. 1974. Bacterial metabolism of para- and meta-xylene: oxidation of the aromatic ring. J. Bacteriol. 119:930-936. 14. Huddleston, R. L., and R. C. Allred. 1963. Microbial oxidation of sulfonated alkylbenzenes. Dev. Ind. Microbiol. 4:24-38. 15. Joklaarl, J. 1974. INDO- and CNDO-approximated molecular orbital calculations on the spin-spin coupling constants in some 2-substituted oxetanes. Z. Naturforsch. Teil A 29:1907-1913. 16. Leldner, H., R. Gloor, and K. Wuhrmann. 1976. Abbankinetik linear alkylbenzolsulfonate. Tenside Deterg. 13:122-130. 17. Lowry, 0. H., N. J. Rosebrough, A. L. Farr, and R. J. Randall. 1951. Protein measurement with the Folin phenol reagent. J. Biol. Chem. 193:265-275. 18. Rowbotham, J. B., and T. Schaefer. 1973. Experimental and theoretical estimates of sigma and pi electron contributions to long-range spin-spin constants in coumarin and

421

its methyl derivatives. Can. J. Chem. 51:953-960. 19. Sariaslani, F. S., D. B. Harper, and I. J. Higgins. 1974. Microbial degradation of hydrocarbons: catabolism of 1phenylalkanes by Nocardia salmonicolor. Biochem. J. 140:31-45. 20. Sjoblad, R. D., and J. M. Bollag. 1977. Oxidative coupling of aromatic pesticide intermediates by a fungal phenol oxidase. Appl. Environ. Microbiol. 33:906-910. 21. Sommers, L. E., M. A. Tabatabal, and D. W. Nelson. 1977. Forms of sulfur in sewage sludge. J. Environ. Qual. 6:42-46. 22. Sposito, G., K. M. Holtzdaw, and J. Baham. 1976. Analytical properties of the soluble, metal-complexing fractions in sludge-soil mixtures. II. Comparative structural chemistry of fulvic acid. Soil Sci. Soc. Am. J. 40:691-697. 23. Swisher, R. D. 1963. The chemistry of surfactant biodegradation. J. Am. Oil Chem. Soc. 40:648-656. 24. van der Linden, A. C., and G. J. E. ThJsse. 1965. The mechanism of microbial oxidations of petroleum hydrocarbons. Adv. Enzymol. 27:469-546. 25. Weast, R. C. (ed.). 1975. Handbook of chemistry and physics, 56th ed. CRC Press, Cleveland, Ohio.