INTRODUCTION

Top Abstract Introduction Materials & Methods Results Discussion References

In 1985, it was discovered that the white fungus Phanerochaete chrysosporium ATCC 24725 has a unique ability to mineralize free and lignin-bound chloroanilines (1). The fungus produced about the same high yield of ¹⁴CO₂ regardless of whether the ¹⁴C label was in the lignin-bound chloroaniline, the natural lignin subunit (coniferyl alcohol), or the free chlorinated anilines. The unique capability of P. chrysosporium for mineralization of xenobiotics was also discovered for polychlorinated biphenyls (7) as well as TCDD, DDT, benzo[a]pyrene, and chlorinated biphenyls (6). Numerous reports on the successful mineralization of additional xenobiotics have since appeared (for reviews, see references 4, 12, and 26). The initial report on the mineralization of free and lignin-bound chloroanilines was further confirmed by studies on the mineralization of a chloroaniline-lignin metabolite fraction from wheat (2). A fungal lignin peroxidase preparation was found to react rapidly with chlorinated anilines, but the main reaction was oligomerization rather than ring opening. The highly toxic 3,4,3',4'-tetrachloroazobenzene and other condensation products were formed (23, 23a). More complex condensation products of 3,4-dichloroaniline (DCA) have been characterized from another white rot fungus (16). In addition, fungi are known to transform anilines to the N-formyl, N-malonyl, N-acetyl, N-succinyl, N-hydroxyglutaryl, and N-glucosyl conjugates (11, 27, 28) as well as to unusual cyclic succinimides (3). More recently, chlorinated anilines and herbicides tightly complexed with native plant cell wall components were found to be mineralized in high yield (20). P. chrysosporium was successfully used in a two-stage fermentation system to clean soil heavily contaminated with polycyclic hydrocarbons (19).

Our initial results have been confirmed with the same strain 24725 (21) as well as with P. chrysosporium 32629 (10). However, our earlier work has also been attributed to experimental artifacts, and P. chrysosporium ATCC 34541 was reported to be unable to mineralize DCA (11). It has been postulated that P. chrysosporium is no different from other soil fungi, which generally have only a low capability for the mineralization of chloroanilines (24, 27).

The initial purpose of the present work was to define optimal fermentation conditions. In addition, a new intermediate of mineralization was discovered when the kinetics of mineralization were studied. This metabolite, termed DCAX, is now identified as N-(3,4-dichlorophenyl)--ketoglutaryl--amide by isolation from C- and N-rich medium and by standard spectroscopic techniques. A hypothetical mechanism for the formation of this metabolite and its conversion to the succinimide is given. Some of the results have been described briefly (26, 29).

	MATERIALS AND METHODS

Top Abstract Introduction Materials & Methods Results Discussion References

Chemicals. All chemicals used in this study were of analytical grade. The high-performance liquid chromatography (HPLC) solvents used were from Riedel-de-Haën. 3,4-Dichloro[U-¹⁴C]aniline was obtained and purified as previously described (1, 20). Reference substances were generally from our laboratory collection (3, 23, 23a, 28).

Fungal strains and kinetic fermentation studies. P. chrysosporium ATCC 24725 and 34541 were maintained on malt extract agar (1) and grown under the previously described conditions in either static culture (2.0-liter Fernbach flask with 100 ml of medium and no detergent or veratryl alcohol added [1, 11]) or agitated cultures (250-ml Erlenmeyer flask with 0.1% [wt/vol] Tween 80 and 1.5 mM veratryl alcohol added [cf. reference 20]). The composition of the N-limited, C-limited, and C- and N-rich growth media were as described previously (1, 14). A fungal spore suspension was used for inoculation with 1.3 · 10⁶ spores/100 ml of medium. The agitated cultures were shaken (150 rpm, 37°C) for 3 days to allow the fungal mycelium to develop. The culture medium was then reduced from 100 to 40 ml, and the agitation speed was set to 60 rpm. The ¹⁴C-labelled test chemical was added, and incubation was continued, with flushing with pure oxygen twice per week. ¹⁴CO₂ development was monitored as described previously (1, 20). All kinetic experiments were performed with 100 µM [¹⁴C]DCA (1.5 · 10⁶ dpm). Immediately following flushing with oxygen, 1 ml of growth medium was removed. A sample of 500 µl was directly employed for HPLC analysis of metabolites. Radioactivity (1, 20) and lignin peroxidase activity (20) were also determined. One unit of lignin peroxidase activity is defined as catalyzing the oxidation of 1.0 µmol of veratryl alcohol/min. The growth media were directly extracted with ethyl acetate to obtain DCASI as a major product (cf. reference 3). Prior acidification with KH₂PO₄ also allowed extraction of DCAX with ethyl acetate (28). General fermentation conditions were as previously reported (1, 20). Total radioactivity associated with the mycelium was determined by combustion.

Metabolite isolation. Because of the rapid decline in N-limited cultures (26, 28), C- and N-rich medium (40 ml) was selected for isolation of DCAX. For this purpose, [U-¹⁴C]DCA (38.5 kBq, 4.0 µmol, 23 days of incubation), nonradioactive DCA (4.0 µmol, 10 day of incubation), or nonradioactive DCASA (53.4 µmol, 10 days of incubation) was applied to 3-day-old cultures of P. chrysosporium. Incubation was at 39°C and 70 rpm, with oxygen flushing every 2 days. Medium samples of 500 µl were analyzed by HPLC to monitor the formation of DCAX.

Thin-layer chromatography (TLC). Precoated fluorescent silica gel G 60F₂₅₄ plates (Merck no. 5554) were used with an ethyl acetate-2-propanol-water solvent system (6:2:1 [by volume]). R_f values were 0.11 for DCAX, 0.50 for DCASA, 0.57 for N-glucosyl-DCA, 0.76 for DCASI, and 0.80 for DCA.

HPLC. Samples (500 µl each) from the incubation medium of the kinetic experiments were analyzed by reversed-phase HPLC (see reference 20) on a Lichrosphere RP8 column (16 by 250 mm). The gradient for the analysis of metabolites was formed between solvent system A (water containing 100 µl · 1¹ trifluoroacetic acid) and solvent B (acetonitrile-water containing 100 µl · liter of trifluoroacetic acid¹ [9:1 by vol]). Over a total time period of 24 min, a stepwise elution program from 100% solvent system A to 100% solvent B was applied. UV detection was at 210 and 245 nm. Radioactivity was monitored with an LB503 radiomonitor (Berthold, Wildbach, Germany). The column was calibrated with authentic reference compounds. The following retention times were obtained: for N-glucosyl-DCA, 12.2 min; for N-malonyl-DCA, 16.2 min; for DCA-succinimide, 17.6 min; for N-acetyl-DCA, 18.2 min; and for DCA, 18.3 min. DCAX appeared at a retention time of 11.6 min.

Metabolite purification. The lyophilized culture medium was dissolved in 5 ml of water and passed through an RP18 cartridge (500 mg; Merck no. 1.19849.001). DCAX remained stable in this solution when stored at 4°C. Reversed-phase HPLC was carried with a Spherisorb C₈ column (125 by 4.6 mm in diameter) connected to a C₈ guard column (28 by 4.6 mm in diameter) (particle size, 5 µm; Bischoff). The eluent was formed between solvent A (water) and solvent B (89.4% acetonitrile and 0.6% methanol in water) according to the following program (1 min 10% B, 2 min 10% B, 5 min 20% B, 8 min 40% B, 13 min 45% B, 15 min 50% B, 20 min 60% B, 24 min 100% B, 28 min 100% B, and 30 min 0% B) at a flow rate of 0.6 ml/min. The column had been calibrated for analysis with the following authentic samples: N-(3,4-dichlorophenyl)glucoside (retention time [R_t], 10.4 min), N-(3,4-dichlorophenyl)--ketoglutaryl--amide (DCAKGA; R_t, 12.1 to 13.7 min), N-(3,4-dichlorophenyl)-succinamide (DCASA; R_t, 15.7 to 16.2 min), N-(3,4-dichlorophenyl)-succinimide (DCASI; R_t, 19.5 to 21.2 min), N-(3,4-dichlorophenyl)-acetanilide (R_t, 22.1 to 22.2 min), and free 3,4-dichloroaniline (DCA; R_t, 22.6 to 22.9 min). For comparison, the R_t values of the authentic samples on a Spherisorb ODSII C₁₈ column (125 by 4.6 mm in diameter; particle size, 5 µm; Knauer) without a precolumn were also determined for the same gradient as follows: for N-(3,4-dichlorophenyl)-glucoside, 11.3 min; for DCAKGA, 13.3 min; for DCASA, 14.6 min; for DCASI, 19.8 min; for N-(3,4-dichlorophenyl)-acetanilide, 20.7 min; and for free DCA, 22.0 min. The UV signal was detected at 250 nm (UV detector model 757; Applied Biosystems), and the radioactivity signal was monitored on-line with a solid yttrium glass scintillator (HPLC radioactivity detector model LB 505; Berthold, Wildbad, Germany). Samples of 500 µl of undiluted culture medium or the enriched DCAX solution (see above) were injected into the HPLC system.

Chemical syntheses. N-(3,4-Dichlorophenyl)-acetanilide, DCASA, and N-(3,4-dichlorophenyl)-succinimide were synthesized essentially as described previously (2, 11, 28).

Synthesis of N-(3,4-dichlorophenyl)--ketoglutaryl--amide. The synthesis of DCAKGA was performed in two steps. First, -ketoglutaric anhydride was synthesized in a manner analogous to succinic anhydride synthesis (9, 22). In the second step, DCAKGA was synthesized from -ketoglutaric anhydride and DCA. In a closed system, 12 mmol of -ketoglutaric acid was suspended in 10 ml of dry benzene and 24 mmol of acetic acid anhydride was slowly added. The -ketoglutaric acid crystals slowly dissolved under warming and occasional shaking. The solution was then transferred to an ice bath. Dry petrol ether (40 to 60°C) was added in small amounts until an amorphous solid mass precipitated. The latter was washed several times with dry petrol ether (40 to 60°C), with dry benzene added in order to remove excess acetic acid and acetic acid anhydride. The amorphous solid mass was dried in vacuo at ambient temperature and finally dried under a stream of nitrogen. In the second step, the crude -ketoglutaric anhydride was dissolved in 45 ml of dry dioxane and 6.2 mmol of DCA was added portionwise. The solution was stirred at ambient temperature for 3 days. After 3 days, the solution was concentrated in vacuo at ambient temperature to ca. 5 ml. Precipitation of DCAKGA was induced by the addition of dry diethyl ether-petrol ether (40 to 60°C), 1:2 (vol/vol). The precipitate was washed with dry diethyl ether in order to remove -ketoglutaric anhydride as well as DCA and DCA-acetanilide. The crystalline solid mass was dried under nitrogen and stored under nitrogen at 18°C. The yield of DCAKGA (melting point, 204 to 205°C) was about 62%, as shown by HPLC. In addition, DCA-acetanilide (27%) was detected by HPLC and ¹H-nuclear magnetic resonance (NMR) spectroscopy. This compound appeared to be formed from residual acetic anhydride. The investigation of the reaction product with 2D-HH-COSY-NMR spectroscopy allowed us to separate the signals of the main product (DCAKGA) from those of the by-product (DCA-acetanilide). The authenticity of DCAKGA was further demonstrated by HPLC-mass spectrometry-mass spectrometry (MS-MS) (see Results).

Combined HPLC-MS. Combined HPLC-MS was conducted with a single quadrupole (Fisons, Mainz, Germany) as well as a triple-stage quadrupole MS system (TSQ 7000; Finnigan MAT, Bremen, Germany). When coupled with HPLC, an acid-free acetonitrile-water gradient on a reversed-phase C₁₈ column (250 by 4.6 mm in diameter [Fisons] and 150 by 4.5 mm in diameter [Finnigan MAT]) was used. The column was connected to the MS system via an electrospray interface. All mass spectra were acquired in the negative ion mode at a typical scan rate of 110 to 167 amu/s. By triple-quadrupole MS, the fragmentation of DCAX and DCASA was achieved with argon as a collision gas (1 mTorr) and atmospheric pressure ionization-collision-induced dissociation at the upfront, whereas synthetic DCAKGA was fragmented by atmospheric pressure ionization-chemical ionization. The collision-induced dissociation spectra of DCASA, DCAKGA, and DCAX were obtained at energies of 35, 50, and 50 eV, respectively. The atmospheric pressure ionization-chemical ionization mass spectrum of synthetic DCAKGA was available at 5 eV energy, and the daughter ion (m/z, 288) mass spectrum of DCAKGA was available at 18 eV.

Combined GC-MS. The experimental conditions of the combined GC-MS were as follows. GC was done with a 30 m by 0.25 mm MS select (fused silica ITP-Pro, DB-5) column, an injector temperature of 250°C, an initial oven temperature of 80°C and an initial hold of 3 min, a final temperature of 260°C at an increment of 30°C per minute, and a final hold of 10 min. MS was done with a source temperature of 150°C, a source emission current of 400 mA, 70 eV of electron energy, and CI-reagent gas methane at 10⁶ Torr in the source. The mass spectrometer (TSQ 7000; Finnigan MAT) was run in the negative mode.

¹H-NMR spectroscopy. ¹H-NMR spectra were obtained with a Bruker-AC 400 NMR spectrometer (400.13 MHz) at 30°C in acetone-d₆ ( = 2.04 ppm) and/or methanol-d₄ ( = 3.30 ppm) at 303 K with a 5-mm inverse geometry probe (90° = 8.5 µs).

	RESULTS

Top Abstract Introduction Materials & Methods Results Discussion References

Optimal fermentation conditions. The original work of the efficient mineralization of DCA was carried out with P. chrysosporium ATCC 24725 (1). However, P. chrysosporium strain ATCC 34541 was reported to be unable to mineralize DCA (11). Hallinger et al. (11) attributed the earlier mineralization results to experimental artifacts. No data that could resolve the controversy have been published since. ATCC strains 24725 and 34541 have now been compared in the two reported fermentation regimes which both use the standard N-limited growth medium (14). The time courses of ¹⁴CO₂ development from 3,4-dichloro[U-¹⁴C]aniline are shown in Fig. 1. Both fungal strains were effective mineralizers under the original conditions (1), i.e., 37°C and gassing with pure oxygen in 2.0-liter flasks. It made no difference whether ¹⁴CO₂ was collected after passage through a polyurethane plug in 2-aminoethanol (1, 2) or in 1 M NaOH (11). Both strains were inactive under the conditions described by Hallinger et al. (11), i.e., at 27°C and gassing with air in 0.2-liter flasks. When an inactive culture was transferred to 37°C and aerated with oxygen, there was a rapid development of mineralization activity. A similar observation was made with static C-limited cultures that exhibited a low mineralization rate. When C limitation was ended by the addition of 5 mM D-glucose, there was an immediate increase in ¹⁴CO₂ development (28). Therefore, inactive fungal cultures remain competent for mineralization when transferred to optimal conditions.

View larger version (17K): [in this window] [in a new window]   FIG. 1.   Time courses of 14CO2 development from 3,4-dichloro[U-14C]aniline by P. chrysosporium 24725 (-) and 34541 (-) in static culture (conditions described in reference 1). Results of a repetition of the same experiments under the conditions used by Hallinger et al. (11) with both strains are also shown (×-×). The latter curve was also obtained with heat-inactivated mycelium tested in N-limited medium (1, 14). Mean values ± standard deviations (n = 4) are shown.

Lignin peroxidase production. Lignin peroxidase production in low-N cultures began between fermentation days 2 and 4 (cf. reference 20). Optimal induction in shake cultures has been reported to require C limitation and the addition of veratryl alcohol as well as the detergent Tween 80 (18). This could be confirmed in our study (28). Lignin peroxidase activity (units · liter¹ in agitated cultures) were 40 ± 20 (mean ± standard deviation) under N limitation, 150 ± 30 under C limitation, and <10 in C- and N-rich medium. Total lignin peroxidase activity did not correlate with the extent of mineralization in the three different growth media, since mineralization was high under N limitation and very low under C limitation (see below). These results did not exclude the possibility that mineralization required a low amount of basal lignin peroxidase activity.

Intermediates of mineralization. Incubation of an N-limited static culture with [U-¹⁴C]DCA led upon ethyl acetate extraction to the succinimide conjugate as the main soluble metabolite (3). A systematic comparison of N-limited, C-limited, and C- and N-rich growth media has now been performed with HPLC metabolite analysis as shown in Fig. 2. In all three growth media, DCA disappeared within 2 days of application, with the concomitant appearance of a highly polar metabolite that was termed DCAX (R_t, 11.6 min). The time courses determined in the three growth media are shown in Fig. 3. The HPLC profiles obtained with N-limited growth medium showed that DCAX was high after 2 days (Fig. 2). This was followed by a rapid decline (Fig. 3) with a concomitant rise of DCASI, which reached a maximum at day 10. DCASI then declined in favor of CO₂ development. In C-limited and even more in C- and N-rich medium, DCAX remained high as a dead-end metabolite up to day 21. There were much lower amounts of DCASI, and CO₂ formation was very low (C limitation) or not detectable (C- and N-rich medium). Smaller amounts (<20%) of additional metabolites were detected (unidentified products as well as N-glucosyl-DCA [cf. reference 30] and DCASA [cf. reference 3]). These minor metabolites could not be correlated with ¹⁴CO₂ development and are not documented here. Mean total recoveries of ¹⁴C over the nine time points per kinetic experiment were 99.9% ± 11.3% (N limitation), 98.6% ± 8% (C limitation), and 100.0% ± 5.1% (C- and N-rich medium). The representative test series shown in Fig. 3 was reproduced three times under slightly different experimental conditions. Treatment of incubation medium with ethyl acetate led to the selective extraction of DCASI (cf. reference 3). After acidification with KH₂PO₄, DCAX could also be extracted with ethyl acetate, but it decomposed slowly to material chromatographing near the DCASI standard (R_f, 0.7; TLC analysis [28]). Fungal mycelium never contained more than 2 to 5% of applied radioactivity. In the case of N limitation, DCASI was the main mycelial component (TLC analysis [28]).

View larger version (19K): [in this window] [in a new window]   FIG. 2.   HPLC profiles of aliquots from N-limited growth medium 2 days (upper panel) and 7 days (lower panel) after addition of 100 µM [U-14C]DCA. The column had been calibrated with the authentic reference metabolites (see Materials and Methods). The positions of DCAX (Rt, 11.6 min) and of DCASI (Rt, 17.6 min) are shown. The initial [U-14C]DCA (Rt, 18.3 min) had largely disappeared after 2 days.

View larger version (18K): [in this window] [in a new window]   FIG. 3.   Kinetics of [U-14C]DCA metabolism. (A) N-limited medium. (B) C-limited medium. (C) C- and N-rich medium. Fermentation was carried out at 37°C, with oxygen gassing every 2 days. The substrate concentration was 100 µM. Aliquots (500 µl each) of the growth medium were directly employed for HPLC analysis (Fig. 2). The distributions of DCA (-) and the major metabolites, DCAX (-) and DCASI (-), are plotted. 14CO2 (-) was trapped and quantitated as described previously (1, 20).

Isolation of DCAX. The new intermediate DCAX was labile upon solvent extraction, even though it was quite stable in C- and N-rich growth medium at pH 4 to 5 (26, 28). Therefore, no solvent extraction of DCAX from the C- and N-rich culture medium with organic solvents was performed. Instead, DCAX from the C- and N-rich cultures was purified by lyophilization and HPLC with a step-wise gradient from water to organic solvent. The collected fraction (R_t, 12.1 min) was dried under a stream of N₂ and could be stored under N₂ at 18°C for several weeks without significant decomposition.

Chemical structure of DCAX. DCAX and synthetic DCAKGA had the same R_ts of 12.1 and 13.3 min upon C₈ and C₁₈ reversed-phase HPLC, respectively. The purified DCAX fraction was analyzed by combined HPLC-MS after being dissolved in CH₃CN-MeOH (1:1). A main product at an R_t of 10.9 min and a by-product at an R_t of 17.45 min appeared upon combined HPLC-MS (electrospray ionization, negative mode). The UV analog signals (at 250 nm) agreed with the mass scan maxima at the calculated values of m/z 288 ([M-H]; DCAKGA; R_t, 10.9 min] and m/z 260 ([M-H]; DCASA; R_t, 17.45 min). Both peaks showed the isotope cluster for two chlorine atoms. This result was confirmed by three independent HPLC-MS systems (data not shown). The mass spectra obtained with the [¹⁴C]DCAX sample and with synthetic DCAKGA are shown in Fig. 4. The proposed fragmentation pattern is indicated. The fragment with an m/z of 288 led to major daughter ions at m/z of 160, 216 and 244. The latter fragments corresponded to the 3,4-dichloroaniline fragment and to fragments due to the loss of the -COOH and -CO-COOH groups. Incubation of the fungus with DCASA (instead of DCA) in C- and N-rich medium also led to the formation of DCAX, which was identified by direct HPLC-MS analysis showing a mole peak at m/z 288 ([M-H]) with an isotope cluster for two chlorine atoms. Fungal mycelium was found to contain arylacylamidase activity for DCASA (0.73 pkat/mg of soluble protein) (data not shown). The presence of a COOH group in DCAX was further demonstrated by methylation and HPLC on the reversed-phase C₈ column (R_t, 11.2 min). The mass spectra of methylated DCAX and DCAKGA were identical and indicated that the -NH- and the -CO₂H groups had both been methylated (Fig. 5). Synthetic DCAKGA was further characterized by its ¹H-NMR spectrum (Fig. 6 and Table 1). The spectrum showed an intact 3,4-dichloroaniline ring system as well as the expected side-chain protons.

View larger version (18K): [in this window] [in a new window]   FIG. 4.   HPLC-MS-MS electrospray-negative mode mass spectra of fungal DCAX (right panel) and of synthetic DCAKGA (left panel). An MS-MS system (TSQ7000; Finnigan MAT) was used with a collision energy of 50 eV. Anions are detected. The proposed fragmentation pattern of N-(3,4-dichlorophenyl)--ketoglutaryl--amide is shown.

View larger version (22K): [in this window] [in a new window]   FIG. 5.   GC-MS of methylated DCAKGA (left panel) and methylated DCAX (right panel). Methylation was performed with a TMS-diazomethane reagent (5). The methylated compounds both had Rts of 11.2 min.

View larger version (14K): [in this window] [in a new window]   FIG. 6.   1H-NMR spectrum of synthetic DCAKGA with assignment of proton signals (acetone-d6, 2.04 ppm at 303 K).

Decomposition products of DCAKGA. During purification by HPLC, DCAX as well as DCAKGA partially decomposed into DCASA and smaller amounts of DCASI. DCASA and DCASI were identified by HPLC retention times on C₈ and C₁₈ columns and by ¹H-NMR spectroscopy (Table 1). When DCAX was transferred from a pH of 4 to 5 to a pH of 5.5, 6.0, or 6.5, complete conversion to a distinctly more polar, unidentified product occurred. It seems likely that the known reversible isomerization of -ketoglutaric acid to 2-hydroxy-5-oxo-tetrahydrofuran-2-carboxylic acid (8) had occurred.

	DISCUSSION

Top Abstract Introduction Materials & Methods Results Discussion References

Optimal fermentation conditions. The previous failure to obtain mineralization of DCA with P. chrysosporium strain 34541 can now be explained by the inadequate fermentation conditions used. The use of smaller flasks and gassing with air rather than oxygen both reduced the oxygen partial pressure. The latter is well known to be crucial for mineralization of lignin (13, 17). In addition, room temperature was used instead of the optimal growth temperature of 37°C. Inactive fungal cultures are shown here to gain full mineralization activity upon an increase of oxygen pressure and a shift in temperature to 37°C or upon the addition of D-glucose to a C-limited culture. Mineralization activity is known to appear in the transition from exponential to stationary growth phase and to be optimal under N limitation and after the addition of an inducer such as veratryl alcohol (13, 17). These various ways of optimizing mineralization may be useful tools in studies of differential gene expression. The enzymes and transcripts that are involved in ring cleavage and mineralization reactions of lignin or of xenobiotics are still largely unknown.

Detection of a new intermediate. Addition of [U-¹⁴C]DCA to cultures of P. chrysosporium is shown here to lead within 2 days to a yield of up to 90% of a novel, highly polar intermediate termed DCAX. In agreement with the previous comparison of N-limited, C-limited, and C- and N-rich growth media (1), significant mineralization occurred only under N-limited growth conditions. These conditions also led to a near-stoichiometric conversion of DCAX to DCASI (Fig. 2 and 3). Extraction of DCAX with ethyl acetate required prior acidification. This observation and the low R_f value upon TLC led to the proposal of DCAX being the anilide of a dicarboxylic acid (28). The chemical lability of DCAX was consistent with a -keto-acyl structure. -Ketoadipic acid is a known intermediate of aromatic ring cleavage in P. chrysosporium (cf. reference 25), and its terminal anilide could conceivably rearrange to yield DCASI. However, the above-mentioned detailed structural studies have shown that DCAX is the -anilide of -ketoglutaric acid.

Structure of the new intermediate. The comparison between DCAX and synthetic DCAKGA by HPLC retention times, HPLC-MS, and GC-MS gave unequivocal evidence for the identity of both compounds. Analogous to the spontaneous formation of DCASA from DCA and succinyl-S-coenzyme A (3), one would expect the new compound to form from DCA and -ketoglutaryl--S-coenzyme A. The latter could be formed by a broad-specificity coenzyme A ligase. The derived structure of DCAX contrasts with that of -hydroxy-glutaryl-S-coenzyme A, which is activated at the vicinal rather than the distal carboxyl group (22). Hallinger et al. (11) also placed DCA at the vicinal carboxyl group of their -hydroxyglutaryl conjugate, but no direct experimental evidence was presented.

View larger version (12K): [in this window] [in a new window]   FIG. 7.   Proposed metabolic scheme. After initial conjugation of DCA, the new intermediate DCAKGA is converted to the coenzyme A derivative of DCASA by an -ketoglutarate dehydrogenase-type reaction. This is followed by spontaneous cyclization to give DCASI, which is the closest intermediate of mineralization that has been identified so far. CoA, coenzyme A.