FEMS Microbiology Letters 50 (1988) 233-239 233 Published by Elsevier FEM 03127 fl-methylmuconolactone, a key intermediate in the dissimilation of methylaromatic compounds by a modified 3-oxoadipate pathway evolved in nocardioform actinomycetes Neil C. Bruce and Ronald B. Cain Department of Agricultural and Environmental Science, The University, Newcastle upon Tyne, NE1 7RU, U.K. Received 23 October 1987 Revision received and accepted 3 December 1987 Key words: fl-methylmuconolactone; Actinomycete; Rhodococcus sp. 1. SUMMARY p-toluate-grown cells of Rhodococcus ruber N75, R. corallinus N657, R. rhodochrous N5 and Rhodococcus strains BCN1, BCN2 and 4PH1 metabolized 4-methylcatechol by a modified 3- oxoadipate pathway. Steps in the conversion of this compound to 4-methyl-3-oxoadipic acid were investigated. The conversion of 4-carboxymethyl- 4-methylbut-2-en-l,4-olide to 4-carboxymethyl-3- methylbut-2-en-l,4-olide by a new enzyme is described. 2. INTRODUCTION Alkyl-substituted aromatic nuclei are commonly found in petroleum, coal tar and in the liquors of coking oven wastes, as well as in many synthetic compounds. The microbial dissimilation of such compounds generally proceeds via oxida- Correspondence to." Dr. N.C. Bruce at his present address: The Biotechnology Centre, University of Cambridge, Downing Street, Cambridge, CB2 3EF, U.K. 0378-1097/88/$03.50 1988 Federation of European Microbiological Societies tion to an alkylcatechol as the key diphenolic intermediate followed by meta-fission reactions which lead to the appropriate alkanals and pyruvate as products [1-5]. Metabolism of methyl-substituted benzoates and phenols by an ortho-fission route analogous to the 3-oxoadipate pathway, has been considered unlikely in bacteria, due not to any substrate specifity of the catechol dioxygenase [6-14], or the muconate cycloisomerase [15,18,24] which in numerous genera can be shown to accept substrates with a methyl-substituent, but to the next enzyme in the pathway, muconolactone isomerase. If the methylmuconolactone substrate is methyl-substituted on the C-4 carbon [15-16], there is no free proton to undergo the shift of the muconate isomerase reaction. The product thus becomes a 'dead-end' metabolite. The complete catabolism of methyl-substituted benzoates by an ortho-fission pathway was first demonstrated by Miller [12] in the nocardioform actinomycete, Gordona rubra (now R. corallinus) N657, capable of utilizing 4-methylbenzoate (ptoluate) and 3-methylbenzoate (m-toluate). Miller found the metab01ite /3-methylmuconolactone (4- carboxymethyl-3-methylbut-2-en-l,4-olide; triv-
234 ially 3-methyl-2-enelactone) (VI) to accumulate from 4-methylcatechol in incubation mixtures with small amounts of extract from cells grown at the expense of benzoate or p-toluate [12]. With a methyl substituent in the C-3 position, further degradation of this lactone substrate is possible in a manner entirely analogous to that by which (+)-muconolactone is converted to 3-oxoadipate [17]. Recently the operation of a modified 3- oxoadipate pathway for the dissimilation of methyl-substituted aromatics has been found both in the eukaryotic fungus Trichosporon cutaneum [18] and described in a laboratory-constructed strain of Alcaligenes eutrophus JMP134 [19]. In this investigation we show that such a pathway, has evolved widely in the nocardioform actinomycetes which circumvent accumulation of the 'dead-end' metabolite, -{-methylmuconolactone (4-carboxymethyl-4-methylbut-2-en- 1,4-olide; 4- methyl-2-enelactone) (V), by first isomerising it to 3-methyl-2-enelactone and degrading this metabolite further to 4-methyl-3-oxoadipate (VIII). 3. MATERIALS AND METHODS The principal organism used in this investigation was R. ruber N75 (LA1069). This and the additional bacterial strains R. corallinus N657 (NCTC1068; ATCC25593 type strain), and R. rhodochrous (Nocardia salmonicolor) N5 (NCIB 9701), were originally supplied by Dr. M. Goodfellow. The Rhodococcus strains BCN1, BCN2 and 4PH1 were isolated from garden soil in Whitstable, Kent by enrichment with p-toluate. Growth conditions for bulk preparation of the bacteria and preparations of cell-free extracts were as described by Miller [12]. (-)-3-Methyl-2-enelactone (VI) was prepared biologically from 4-methylcatechol according to the method of Miller [12] using crude extracts from cells of R. ruber N75 or R. corallinus N657 grown at the expense of 10 mm benzoate. Samples of the isomeric 2-methyl and 4-methyl-2-enelactones and of synthetic (racemic) 3-methyl-2-enelactone were kindly supplied by Dr. K.-H. Engesser and Professor S. Dagley, respectively, while 4-methyl-3-oxoadipic acid was a gift from Dr. J.B. Powlowski. p-toluate cis-dihydrodiol (II) was prepared by the method of Whited et al. [21] using a mutant of Pseudomonas putida BGXM1 kindly supplied by Professor D.T. Gibson. The resolution and identification of 2-methyl- 3-methyl- and 4-methyl-2-enelactones in experimental samples was determined by HPLC analysis at 215 nm, on an LDC/Milton Roy 3000 system, linked to an LDC C1/10B integrator-printer plotter. The 25-cm length column contained 5 t~ Spherisorb-ODS (C18) reverse-phase packing. The solvent system was that described by Pieper et al. [19]. The colorimetric determination of 3-oxoacids was adapted from the method of Walker [20] in which 3-oxoacids are coupled with diazotized p- nitroaniline. Dihydrodiol dehydrogenase, was measured by the method of Whited et al. [21]; catechol 1,2-dioxygenase (EC 1.13.11.1) and muconate cycloisomerase (EC 5.5.1.1) and 3-oxoadipate; succinyl- CoA transferase (EC 2.8.3.6) by the methods described by Ornston [17]. The assay for 4-methyl- 3-enelactone methyl isomerase (no EC number yet assigned) depended upon the separation of the two lactone isomers by high-performance liquid chromatography (HPLC). The reaction mixture contained 4-methyl-2-enelactone, 0.3 /~mol and enzyme in 1 ml of 50 mm potassium phosphate buffer, ph 7.0. From the assay mixture, incubated at 30 C in a shaking water bath, 100-/zl aliquots were removed at intervals and immediately discharged into an eppendorf tube containing 10 /zl of conc. H3PO 4 to stop the reaction. The solution was then made up to 1 ml with 890/~1 of distilled water and the protein precipitate removed by centrifugation in a Microfuge before samples (20 #1) of the supernatant were injected into the column and analysed. All incubations were performed in duplicate. Standard curves were all linear in the range investigated (0 to 2 mm) and recoveries were reproducible and quantitative when H3PO 4 was used to precipitate the protein. The unit of enzyme activity is defined as the amount of enzyme necessary to convert 1/zmol of substrate into product in 1 min under the assay conditions described. Protein in extracts used in enzyme assays was measured by the method of Bradford using bovine serum albumin as standard [24].
235 4. RESULTS AND DISCUSSION Growth of R. ruber N75 at the expense of 10 mm p-toluate, but not glucose or succinate, elicited in extracts high activities (over 600 nmol/min mg protein- 1) of (i) a dehydrogenase active against p-toluate dihydrodiol and benzoate dihydrodiol; (ii) a catechol 1,2-dioxygenase; (iii) a muconate cycloisomerase and (iv) a 3-oxoadipate : succinyl-coa transferase active with 3-oxoadipate and 4-methyl-3-oxoadipate. The catechol 1,2-dioxygenase showed activity against catechol, 3-methyl- and 4-methylcatechols (specific activities in excess of 120 nmol/min.mg protein-i), while the cycloisomerase was active against cis, cis-muconate, 2-methyl and 3-methylmuconates (241, 155 and 283 nmol/min-mg protein -a, respectively). Incubation of crude extracts of p- toluate or benzoate-grown cells of R. ruber N75 with 4-methylcatechol or 3-methyl-cis, cismuconate (IV) accumulated a product which formed an hydroxamate, gave a positive test with nitroprusside for/~-unsaturated lactones and presented a UV spectrum with ~kma x at 210 nm (E = 12090). Isolation of this product yielded an offwhite crystalline compound, m.p. 77-78 o C (found C, 53.79%; H, 5.08%; C7HsO 4 requires C, 53.85%; H, 5.16%). The high resolution mass spectrum gave a mass ion at 156.0416; calculated mass for C7H804 requires 156.0422. The biological product was optically active [a]~-35.8 + 0.1 (c = 10.88 mg/ml in water) (mean of 6 determinations on 2 separate biological preparations). The ah-nmr spectrum of the product showed a multiplet at 2.12 ppm attributable to a methyl group on a double bond. A proton multiplet centred at 5.26 ppm represented the X part of an ABX system and was attributable to a proton on an oxygenbearing C-atom. Other NMR assignments coincided well with reference data for 3-methyl-2-enelactone [19], but not for the 4-methyl isomer [15]. This product must therefore be 3-methyl-2-enelactone (VI). We originally suspected it to be the direct lactonisation product of the 3-methylmuconate precursor but when samples of the two other isomeric methylmuconolactones became available to us, we observed that growth at the expense of p-toluate, but not benzoate, glucose or succinate, conferred on whole cells of R. ruber N75, the ability to degrade the 3-methyl- and 4-methyl-2-enelactones at the same rate, whereas the 2-methyl isomer was inert. Favourable rates of disappearance of both lactones by extracts were obtained only with high concentrations of extract protein, but whereas the biologically synthesised (-)-3-methyl-2-enelactone (VI) was completely utilised, only half of a synthetic (racemic) preparation, m.p. 130 C, was consumed by the same extracts. No delactonisation occurred with extracts of glucose- or benzoate-grown cells, or when boiled extracts from p-toluate-grown cells was used in the reaction mixture. The absence of activity in extracts of benzoate-grown cells indicated that classical muconolactone isomerase (EC 5.3.3.4) and 3-oxoadipate enol lactone hydrolase (EC 3.1.1.24) were not responsible for metabolism of the methyl-2-enelactones. In order to identify the immediate product of 4-methyl-2-enelactone (V) metabolism by R. ruber, the lactone (30 /~mol) in 10 ml of 50 mm Tris-HzSO4, ph 7.5 was incubated for 1 h with extract (5.7 mg of protein) from cells grown at the expense of p-toluate. The mixture was acidified with a few drops of conc. H3PO 4 and the precipitated protein removed by centrifugation. Analysis of the supernatant by HPLC showed that the 4-methyl-2-enelactone (V) substrate had been converted almost quantitatively to a compound whose HPLC retention time and other properties coincided with those for 3-methyl-2-enelactone (VI). This conversion was not observed when boiled extract replaced fresh crude extract in the reaction mixture. As the responsible enzyme catalysed the interconversion of the 4-methyl-2-enelactone to the 3-methyl isomer, it was termed lactone methyl isomerase. This observation suggested that the 4-methyl-2-enelactone was an intermediate between 3-methylmuconate and 3-methyl-2-enelactone and thus the immediate lactonisation product of 3-methyl-cis, cis-muconate (IV). It was possible to separate catechol 1,2-dioxygenase, muconate cycloisomerase and 4- methyl-2-enelactone isomerase by anion exchange chromatography (Fig. 1) from which it was possible to select fractions that contained one of the three enzymes negligibly contaminated with the
236 2'5-0'3~0-3 A280 2"0-- 1.5- i 0.2-0-2-1.0-- 0'2 0.1-0-1/ 0"5-- 4, 20 40 60 Fraction number Fig. 1. Separation of the initial enymes of the modified 3-oxoadipate pathway. Crude extract was prepared from 10 g (wet weight) of cells grown with p-toluate and applied to a DEAE-Sephacel column (2.5 11 cm), which had been previously equilibrated with 50 mm Tris-HC1 buffer ph 7.0. After the sample had been loaded, the column was washed with 200 ml of the same buffer mixture and the enzymes were then eluted with 400 ml of buffer containing a linear gradient of 0-0.6 M NaC1. Fractions of 8 ml were collected at a flow rate of 20 ml/h. Each fraction was assayed for catechol 1,2-dioxygenase ( ), muconate cycloisomerase (t3), 4-methyl-2-enelactone isomerase (e) and (solid line, no symbols) protein content (A280). others. The reaction sequence of the metabolic intermediates concerned was then determined by repetitive scanning of the UV spectral in a Beckman DU7 spectrophotometer and by reverse-phase HPLC of aliquots from the reaction mixture. 3-Methyl-cis, cis-muconate (IV) was synthesised from 4-methylcatechol (III) using a partially purified catechol 1,2-dioxygenase preparation (fraction 54, Fig. 1) that contained no muconate cycloisomerase activity. The reaction mixtures contained Tris-HC1 buffer, ph 7.5, 300 /~mol; 4-methylcatechol, 0.5 #tool; and 50 m-units of catechol 1,2-dioxygenase in a total volume of 3 ml. After 1 h the enzymic conversion approached completion. The incubation was continued for a further 30 rain, after which the solution was passed through a PM10 Diaflow ultrafiltration membrane M r 10000 cut-off) to remove the protein. The enzymically prepared 3-methyl-cis,cis-muconate (0.5/~mol) was then incubated with 60 m-units of partially purified muconate cycloisomerase (fraction 69, Fig. 1) and 10 mm MgC12 in a final volume of 3 ml of 50 mm Tris-HC1 buffer, ph 7.5. The UV spectral changes accompanying enzymic cycloisomerisation of 3-methylmuconate showed that with the decrease in A260 there was a coincident increase in A200. At intervals, 50-/~1 aliquots were removed from the reaction mixture and analysed by HPLC. The enzymic cycloisomerisation product was identified by reference to authentic standard compounds as 4-methyl-2- enelactone (V). In a study of substituted muconic acids, Schmidt et al. [24] reported that 3-methylcis, cis-muconate was the most unstable compound examined, chemically cycloisomerising to 3- methyl-2-enelactone almost completely at ph 6.5 within 40 min. Under the assay conditions, however, the use of boiled rather than active enzyme led to less than 5% chemical cyclisation of the 3-methyl-cis, cis-muconate substrate to 3-methyl- 2-enelactone. Enzymic lactonisation was also negligible when 3-methyl-cis,cis-muconate was replaced in the reaction mixture with 3-methylcis, trans-muconate.
237 2 0'5-- 0 --0'5 o 0"4-- E ::L =* o 0"3-- --0"4 -~ o E --0'3 o o I 0.2--? J 0.1-- -0-2 + --0-1 ' I 40 I 80 Time(min) 1~0 160 ~ ' Fig. 2. Lactonisation of 3-methyl-cis, cis-muconate to 4-methyl-2-enelactone and its further conversion to 3-methyl-2-enelactone by the new isomerase. Partially purified muconate cycloisomerase was used to accumulate 4-methyl-2-enelactone ( ) from 3-methylcis, cis-muconate (monitored by HPLC). Addition of the partially purified isomerase enzyme was then shown to be necessary for the conversion of the 4-methyl- to the 3-methyl-2-enelactone (e). The reaction mixture contained: 0.5/~mol of 3-methyl-cis, cis-muconate in 3 ml of 50 mm Tris-HC1 buffer, ph 7.5. At the first arrow, 60 m-units of mueonate cycloisomerase were added, followed by 50 m-units of the lactone methyl isomerase enzyme at the second arrow. After the enzymic cyclisation of 3-methylcis, cis-muconate to 4-methyl-2-enelactone had reached completion, 50 m-units of partially purified 4-methyl-2-enelactone isomerase from the column effluent (fraction 61, Fig. 1) were added to the reaction mixture and analysis by HPLC was resumed. This fraction converted the accumulated 4-methyl-2-enelactone to 3-methyl-2-enelactone quantitatively. A time course of these reaction sequences is presented in Fig. 2. Thus it appears that the novel isomerase is an integral part of the catabolic sequence for the dissimilation of 4-methylcatechol in p-toluate grown cells. In this respect, R. ruber differs markedly from the eukaryotes T. cutaneum [18] and Aspergillus niger (G.W. Cameron and R.B. Cain, unpublished observations) in which direct enzymic cyclisation of 3- methyl-cis, cis-muconate to 3-methyl-2-enelactone occurs and the 4-isomer is metabolically inert. In view of the novel property displayed by R. ruber, several other nocardioform actinomycetes were screened for the new lactone isomerase activity after growth with p-toluate and benzoate. In each case, growth with the methyl aromatic sub- strate led to greatly enhanced (> 10-fold) activities of this lactone methyl isomerase compared with those found in extracts of succinate- or benzoate-grown cells. Absolute activities (nmol/min mg protein-i) among strains varied considerably (N75,410; N657,175; N5,205; BCN1,25; BCN2,18) but were generally related to growth rates of the organism on p toluate. No activities for any recta-pathway enzymes were ever detected other than trace amounts of catechol 2,3-dioxygenase which we have shown [12,25] is a peripheral expression of the catechol 1,2-dioxygenase. With the findings of Catelani et al. [15], Schmidt et al. [24] and Pieper et al. [19], our observations suggest that catabolism of methyl-substituted aromatics by the ortho-cleavage pathway in bacteria always results in the initial formation of 4-methyl-2-enelactone. 4.1. Identification and recovery of products of 3-methyl-2-enelactone degradation Extracts of p-toluate-grown R. ruber N75, catalysed the degradation of 4-methylcatechol and all the subsequent intermediates of the modified
238 3-oxoadipate pathway to an end-product which reacted in the Walker test [20] for 3-oxoacids but unlike 3-oxoadipate gave only a weak Rothera reaction [26]. In order to confirm the identity of the ketoacid produced, 3-methyl-2-enelactone (6.5 #mol) was incubated at 30 C with 7 mg of crude extract protein from p-toluate-grown cells in 2 ml of potassium phosphate buffer, ph 7.3 and the reaction monitored by HPLC at 215 rim. After protein precipitation with conc. H3PO 4, part of the reaction mixture was treated with 2,4-dinitrophenylhydrazine (0.2% w/v in 1 N HC1) and the subsequent dinitrophenylhydrazones were resolved by t.l.c, on silica gel plates (0.2 mm thickness) with solvent A (propan-2-ol-conc. NH4OH-water; 20:1:2, by vol.) and solvent B (butan-l-ol-conc.nh4oh-ethanol; 7 : 1 : 2, by vol.). Incubations of crude extract with authentic 3-oxoadipate and 4-methyl-3-oxoadipate in buffer served as controls which were subjected to the same isolation procedures and used as standards to assist in the identification of the reaction products. The derivatised sample of the Walker-positive reaction product readily resolved in both solvent systems into two components, with RU values of 0.53 and 0.64 in solvent A and 0.41 and 0.52 in solvent B, respectively. The compound giving R/values 0.53 and 0.41 was identical in R/ values and colour reaction after an NaOH spray with authentic 4-methyl-3-oxoadipate 2,4-di- ~i TM OH ~, ' IH ), nitrophenylhydrazone. No such material was produced by extracts from benzoate-grown cells though they did convert (+)-muconolactone to 3-oxoadipate, the dinitrophenylhydrazone derivatives of which had R/ values of 0.49 and 0.34 in solvents A and B, respectively. The second reaction product (Rf values 0.65 and 0.52 in solvents A and B, respectively) was identified as the 2,4-dinitrophenylhydrazone of 3-methyllaevulinic acid, the decarboxylation product of 4-methyl-3- oxoadipate, by catalytically decarboxylating an authentic specimen of the latter with 4-aminoantipyrene [27], reacting the product with acidic 2,4-dinitrophenylhydrazine and resolving it by t.l.c. The remainder of the acidified reaction mixture was extracted with ethyl acetate, and extract dried over anhyd. Na2SO 4 and then evaporated to dryness in vacuo. The residue film and a small speciment (1.2 nag) of authentic 4-methyl-3-oxoadipic acid were separately treated with excess diazomethane in dry diethyl ether overnight, after which the decomposition products of the reagent and the ether were removed in vacuo. The products were examined by fast atom bombardment mass spectrometry to yield in each case a molecular ion (M +) of both the biologically produced and the authentic material at m/e 202, which corresponds to the formula C9Ha405 for the dimethylester of methyloxoadipic acid. The proposed modified ortho-cleavage pathway /COOH h COOH CH3 I CH3 II CH 3 III (,..-COOH F/COOH O" ~.COOH o OO...<o.. CH a VIII II CH 3 V] " C~H~ [ IV l /COOH H 3 C ~ O V Fig. 3. Proposed pathway for the dissimilation of p-toluate in nocardioform actinomycetes. I. p-toluate; II. cis-p-toluate diol; III. 4-methylcatechol; IV. 3-methyl-cis, cis-muconic acid; V. 4-methyl-2-enelactone; VI. 3-methyl-2-enelactone; VII. 4-carboxymethyl-3- methylbut-3-en-l,4-olide (postulated as an intermediate); VIII. 4-methyl-3-oxoadipic acid.
239 for the degradation of 4-methylcatechol in R. ruber is illustrated in Fig. 3.. The results establish unequivocally the intermediate roles of the 4- methyl- and 3-methyl-2-enelactones in the conversion of 4-methylcatechol to 4-methyl-3-oxoadipate and demonstrate the widespread natural occurrence of this pathway, and the new lactone methyl isomerase enzyme, in nocardioform actinomycetes. ACKNOWLEDGEMENTS N.C.B. acknowledges the award of an SERC Studentship for this study which was commenced by R.B.C. at the Institut fiar Mikrobiologie, Universit~it GiSttingen during the tenure of a CIBA- Geigy Senior Visiting Fellowship. We warmly acknowledge gifts of chemicals from, and exchange of information and discussion with, Prof. H.J. Knackmuss, Dr. K.-H. Engesser and Mr. D.H. Pieper. REFERENCES [1] Bayly, R.C., Dagley, S. and Gibson, D.T. (1966) Biochem. J. 101, 293-301. [2] Hegeman, G.D. and Rosenberg, S.L. (1970) Annu. Rev. Microbiol. 24, 429-462. [3] Murray, K., Duggleby, C.J., Sala-Trepat, J.M. and Williams, P.A. (1972) Eur. J. Biochem. 28, 301-310. [4] Sala-Trepat, J.M., Murray, K. and Williams, P.A. (1972) Eur. J. Biochem. 28, 347-356. [5] Bayly, R.C. and Barbour, M.G. (1984) in Microbial Degradation of Organic Compounds (Gibson, D.T., Ed.) pp. 253-294. Marcel Dekker, New York. [6] Nakazawa, H, Inou, H. and Takeda, Y. (1963). J. Biochem. Tokyo 54, 65-74. [7] Kojima, Y., Kujisawa, H., Nakazawa, A., Nakazawa, T., Kanetsuna, F., Taniuchi, H., Nozaki, M. and Hayaishi, O. (1967) J. Biol. Chem. 242, 3270-3278. [8] Varga, J.M. and Neujahr, H.Y. (1969) Biochem. J. 12, 427-434. [9] Patel, R.N., Hou, C.T., Felix, A. and Lillard, M.O. (1976) J. Bacteriol. 127, 536-544. [10] Dorn, E. and Knackmuss, H.-J. (1978) Biochem. J. 174, 85-94. [11] Engelhardt, G., Rast, H.G. and Wallnofer, P.R. (1979) FEMS Microbiol. Lett. 5, 377-383. [12] Miller, D.J. (1981) Zbl. Bakteriol. Suppl. 11, 353-360. [13] Aoki, K., Konohara, T., Shinke, R. and Nishira, H. (1984) Agric. Biol. Chem. 48, 2097-2104. [14] Chen, Y.P., Glenn, A.R. and Dilworth, M.J. (1985) Arch. Microbiol. 141, 225-226. [15] Catelani, D., Fiecchi, A. and Galli, E. (1971) Biochem. J. 121, 89-92. [16] Knackmuss, H.-J., Hellwig, M., Lackner, H. and Otting, W. (1976) Eur. J. Appl. Microbiol. 2, 267-276. [17] Ornston, L.N. (1966) J. Biol. Chem. 241, 3795-3810. [18] Powlowski, J.B. and Dagley, S. (1985) J. Bacteriol. 163, 1126-1135. [19] Pieper, D.H., Engesser, K.-H., Don, R.H., Timmis, K.N. and Knackmuss, H.-J. (1985) FEMS Microbiol. Lett. 29, 63-67. [20] Walker, P.G. (1954) Biochem. J. 58, 699-704. [21] Whited, G.M., McCombie, W.R., Kwart, L.D. and Gibson, D.T. (1986) J. Bacteriol. 166, 1028-1039. [22] Bradford, M.M. (1976) Analyt. Biochem. 72, 248-254. [23] Sistrom, W.R. and Stanier, R.Y. (1954) J. Biol. Chem. 210, 821-836. [24] Schmidt, E., Remberg, G. and Knackmuss, H.-J. (1980) Biochem. J. 192, 331-337. [25] Bruce, N.C., Cain, R.B. and Miller, D.J. (1987) Metabolism of methyl substituted aromatics by a modified /3-ketoadipate pathway; intermediates and enzymology. Submitted for publication. [26] Rothera, A.C.H. (1908) J. Physiol. 37, 491-494. [27] Anderson, J.J. and Dagley, S. (1980) J. Bacteriol. 141, 534-543.