(aerobic/anoxic) conditions

2 downloads 0 Views 317KB Size Report
sources at different levels, was studied in a chemostat with a 48-h hydraulic residence time under cyclic aerobic and anoxic (denitrifying) conditions. The cyclic ...
FEMS Microbiology Letters 231 (2004) 59^65

www.fems-microbiology.org

The impact of feed composition on biodegradation of benzoate under cyclic (aerobic/anoxic) conditions º zer C|nar O



Kahramanmaras Su«tc u« Imam University, Department of Environmental Engineering, Karacasu, Kahramanmaras 46160, Turkey Received 27 May 2003; received in revised form 14 August 2003; accepted 3 December 2003 First published online 23 December 2003

Abstract The response of a mixed microbial culture to different feed compositions, that is, containing benzoate and pyruvate as sole carbon sources at different levels, was studied in a chemostat with a 48-h hydraulic residence time under cyclic aerobic and anoxic (denitrifying) conditions. The cyclic bacterial culture was well adapted to different feed compositions as evidenced by the lack of accumulation of benzoate or pyruvate in the chemostat. Both the benzoate-degrading capabilities and the in vitro catechol 2,3-dioxygenase (C23DO) activities of the cyclic bacterial cultures were in direct proportion to the flux through the chemostat of the substrate degraded by the pathway containing C23DO, with some exceptions. The quantity of C23DO showed a transient decrease during the initial portion of the aerobic period before returning to the level present during the anoxic period. That decrease was most likely caused by the production of H2 O2 by the cells upon being returned to aerobic conditions. 5 2003 Federation of European Microbiological Societies. Published by Elsevier B.V. All rights reserved. Keywords : Biodegradation; Aromatic compound ; Multicomponent substrate; Aerobic; Anoxic ; Cyclic ; Denitri¢cation

1. Introduction The aromatic hydrocarbons include a number of toxic, carcinogenic, and otherwise hazardous compounds that are frequently found as contaminants of soil and ground water. These compounds can be transformed by hydrocarbon-degrading microorganisms which occur naturally in the environment. Biological nitrogen removal (BNR) processes are common for the treatment of municipal wastewaters and are ¢nding increased application in the treatment of industrial wastewaters containing aromatic compounds. In their simplest form, they contain two bioreactors in series, with the

* Tel. : +90 (344) 251 2315/278; Fax: +90 (344) 2512342. E-mail address : [email protected] (O. C|nar). Abbreviations : AeBUR, speci¢c aerobic benzoate uptake rate; AePUR, speci¢c aerobic pyruvate uptake rate; AnBUR, speci¢c anoxic benzoate uptake rate; AnPUR, speci¢c anoxic pyruvate uptake rate; BNR, biological nutrient removal; C12DO, catechol 1,2-dioxygenase; C23DO, catechol 2,3-dioxygenase; CFE, cell-free extract; COD, chemical oxygen demand; DO, dissolved oxygen; G12DO, gentisate 1,2-dioxygenase; ORP, oxygen reduction potential ; P34DO, protocatechuate 3,4dioxygenase ; P45DO, protocatechuate 4,5-dioxygenase ; TEA, terminal electron acceptor; TOC, total organic carbon

¢rst anoxic (denitrifying) and the second aerobic. Under anoxic conditions, the bacteria in a BNR process use oxidized forms of nitrogen (nitrate, nitrite, nitrous oxide, and nitric oxide) as their terminal electron acceptors (TEAs), whereas under aerobic conditions they use molecular oxygen. This means that they must use di¡erent terminal electron transport enzymes (e.g., nitrogen oxide reductases and cytochrome oxidases) under the two conditions. The non-aromatic organic substrates typically present in municipal wastewaters are degraded by the same pathways under both anoxic and aerobic conditions, so the major regulatory challenge facing the bacteria in municipal BNR processes concerns the activity of the terminal electron transport enzymes [1^3]. On the other hand, aromatic organic substrates typically present in industrial wastewaters are degraded by totally di¡erent pathways under anoxic and aerobic conditions. Under aerobic conditions bacteria use molecular oxygen as a reactant in oxidative peripheral and ring cleavage reactions to achieve biodegradation of aromatic compounds [4^6]. On the other hand, under anaerobic conditions bacteria use di¡erent sets of enzymes to achieve biodegradation via the central intermediates benzoyl-CoA, resorcinol, and phloroglucinol using reductive reactions [7]. This means that when wastewaters containing aromatic compounds are treated in BNR systems, the

0378-1097 / 03 / $22.00 5 2003 Federation of European Microbiological Societies. Published by Elsevier B.V. All rights reserved. doi:10.1016/S0378-1097(03)00920-0

FEMSLE 11358 27-1-04

60

O. C|nar / FEMS Microbiology Letters 231 (2004) 59^65

biomass must regulate not only their terminal electron transport enzymes, but also their degradative enzymes. In addition to mixed electron acceptor phenomena, mixed substrate phenomena are also of great microbiological and engineering interest in several man-made environments (i.e., BNR) [8^10]. Studies on mixed substrate growth could play an important role in determining the relative importance of the various regulatory mechanisms involved in mixed substrate growth, thereby enhancing our understanding of microbial ecology, and enabling the rational design of bioreactors in environmental engineering (e.g., BNR). Understanding of how critical enzyme levels within bacterial cultures are regulated is lacking when the microorganisms are growing on a multicomponent substrate, which is normally encountered in BNR. This is a serious obstacle during the design of bioreactors in environmental engineering because enzyme levels are linked to the kinetic properties of the microorganisms, which are considered very important design parameters. To be able to remove that kind of obstacle, continuous culture experiment were run with mixed bacterial cultures cycled between aerobic and anoxic conditions growing on binary substrate mixtures which are pyruvate and benzoate. The activities of the ring cleavage enzymes and whole cell kinetic activities were measured. Benzoate was chosen as the substrate for these studies because it can be degraded aerobically through any of the three central aerobic intermediates (i.e., catechol, gentisate, and protocatechuate) [11], depending on the organisms involved, and is degraded anoxically through benzoyl-CoA, the major anoxic central intermediate [12]. Consequently, the response of benzoate-degrading enzymes is likely to be representative of the response of many aromatic-degrading enzyme systems. Pyruvate was provided as a co-substrate in the feed to allow us to determine whether the electron transport system was in£uenced by the cyclic conditions. Pyruvate is a non-aromatic substrate and its degradative pathway is unchanged by the nature of the TEA, suggesting that any change in the pyruvate biodegradative capability would be due to changes in the activity of the enzymes involved in the aerobic and/or anoxic electron transport chain in the cells. Because pyruvate goes directly into the TCA cycle after one enzymatic step, its catabolic degradation mechanism is not in£uenced by the presence or absence of oxygen unlike benzoate. Consequently, the ability of the bacteria to use pyruvate in batch assays under both aerobic and anoxic conditions was used as an indicator of changes in the ability of the bacteria to transfer electrons to the TEA provided.

2. Materials and methods 2.1. Experimental plan A mixed microbial culture was grown in a cyclic chemo-

stat on a feed that contained di¡erent concentrations of benzoate and pyruvate on a chemical oxygen demand (COD) basis. In the ¢rst study, the feed contained 50% benzoate and 50% pyruvate of the total COD. In the second study, this ratio was changed to 70% benzoate and 30% pyruvate. In the last study, the ratio was kept at 90% benzoate and 10% pyruvate. The chemostat had a 48-h hydraulic retention time (dilution rate: 0.021) and a 12-h cycle time. Each 12-h cycle had an anoxic period and an aerobic period. The length of the aerobic period was 3 h whereas the length of the anoxic period was 9 h. General performance of the chemostats was assessed by measurement of total organic carbon (TOC), benzoate, pyruvate, nitrate N, nitrite N, and biomass concentrations. Aerobic enzyme assays were also performed throughout a cycle to measure the variation in the levels of catechol 1,2-dioxygenase (C12DO), catechol 2,3-dioxygenase (C23DO), protocatechuate 3,4-dioxygenase (P34DO), protocatechuate 4,5-dioxygenase (P45DO), and gentisate 1,2-dioxygenase (G12DO) within the bacteria. Finally, measurements of the ability of the bacteria to degrade benzoate and pyruvate under both aerobic and anoxic conditions were made throughout a cycle. 2.2. Reactor setup and operation The mixed microbial culture was obtained from a chemostat originally started with activated sludge from a municipal wastewater treatment plant (Mauldin Road Wastewater Treatment Plant, South Carolina, USA) that was fed with benzoate and pyruvate under totally anoxic conditions. The chemostat was a closed 14-l Microferm vessel (New Brunswick Scienti¢c, Edison, NJ, USA) with a working volume of 8 l. The chemostat was covered to exclude light and was mixed by a single shaft impeller system at a speed of 720 rpm. A pH controller (Model 5997, Horizon Ecology, Chicago, IL, USA) with an Ingold pH electrode (Chicago, IL, USA) was used in each chemostat to control the pH between 7.3 and 7.5. The pH was adjusted by adding 1.0 M NaOH when the pH dropped below 7.3 or adding 1.75 M o-phosphoric acid when pH exceeded 7.5. The chemostat was operated on mineral salts medium containing benzoate and pyruvate as carbon and energy source. All inorganic nutrients were provided in excess, making benzoate and pyruvate the growth-limiting substances. Each liter of medium contained 55.13 mg CaCl2 W2H2 O, 0.38 mg CoCl2 W6H2 O, 106.5 mg MgSO4 W 7H2 O, 48 mg FeSO4 W7H2 O, 20.53 mg ZnSO4 W7H2 O, 0.1525 mg Na2 MoO4 W2H2 O, 0.04 mg H3 BO3 , 2.89 mg MnSO4 WH2 O, 0.7375 mg CuSO4 W5H2 O, 13.9 mg Na2 EDTAW2H2 O, 179.3 mg (NH4 )2 CO3 , 340 mg KH2 PO4 , and 680 mg K2 HPO4 . The total concentration of organic compounds (i.e., benzoate and pyruvate) in the feed was 2000 mg l31 as COD (8.33 mM). Nitrate was provided as NaNO3 at a concentration of 610 mg l31 as N (43.6 mM; equivalent

FEMSLE 11358 27-1-04

O. C|nar / FEMS Microbiology Letters 231 (2004) 59^65

to 1744 mg l31 of O2 in electron-accepting capacity) to the chemostat, which was in excess of the quantity needed to accept the electrons from benzoate and pyruvate biodegradation. The feed was delivered to the chemostats from two 21-l carboys. One contained the mono- and dibasic phosphates, as well as ammonium carbonate; the other contained benzoate, pyruvate, sodium nitrate, and the inorganic nutrients. The feed was divided in this way to minimize the potential for contamination. Prior to a change in the feed bottles, they were purged with N2 gas (National Welders, Charlotte, NC, USA) for 30 min to ensure that the feed components were free of oxygen. In addition, to eliminate any potential oxygen £ux in the system during chemostat operation, the feed jars and chemostats were continually purged with N2 gas, which was passed through a hot copper column to eliminate trace quantities of oxygen. The temperature in the chemostats was maintained at 25‡C by the room air conditioning system. 2.3. Analytical methods Routine analyses used in operation of the chemostats were run according to standard methods [13]. Benzoate was measured using high-pressure liquid chromatography (HPLC) [14]. The £ow rate of the mobile phase, containing 60% distilled-deionized water, 38% methanol, and 2% acetic acid by volume, was 0.4 ml min31 . A Supelco (Bellefonte, PA, USA) ABZ+ (150 mmU2.1 mm) reverse phase column was used with an injection volume of 50 Wl, giving a lower detection limit of 0.1 mg l31 as COD (0.42 WM). Nitrate and nitrite were determined by ion chromatography (Dionex 100, Sunnyvale, CA, USA). The eluent was a 3.0 mM carbonate/bicarbonate solution with a £ow rate of 2.0 ml min31 [13]. The lower detection limit for nitrate was 0.1 mg l31 as N (7.1 WM) and the limit for nitrite was 0.15 mg l31 as N (11 WM). TOC was measured with a Shimadzu TOC-5000 analyzer, equipped with an ASI-5000 autosampler. The lower detection limit was 0.02 mg l31 as C (1.67 WM). Dissolved oxygen (DO) in the experimental chemostat was determined by an oxygen meter (YSI Model 57, Yellow Spring, OH, USA) with a lower detection limit of 0.05 mg l31 . The oxygen reduction potential (ORP) in the experimental chemostat was measured with a meter (Model 5997, Horizon Ecology) and probe (Mettler-Toledo, Millersville, MD, USA). The biomass concentration was measured by absorbance at 600 nm using a Spectronic 20 (Milton Roy, Rochester, NY, USA) and converted to mass units with the experimentally determined conversion factor of 638 mg l31 biomass/absorbance unit. Protein in the biomass was determined by the bicinchoninic acid^copper reaction [15]. This method is based on the bicinchoninic acid^copper reaction in which proteins reduce Cu2þ to Cu1þ in alkaline solution, giving a purple color. The change in color was measured with a UV-spectrometer at a wavelength of 562 nm to

61

determine the quantity of total protein in the cell-free extract (CFE). 2.4. Aerobic and anoxic substrate uptake capabilities Anoxic and aerobic organic substrate (benzoate and pyruvate) uptake rates were measured to quantify the metabolic capabilities of the bacteria in the chemostat. The ability to remove benzoate was measured under each environmental condition to establish the maximum capability of the bacteria under whole cell conditions. The ability to remove pyruvate was measured as an indication of the relative abilities of the bacteria to transfer electrons to oxygen or nitrate. Approximately 30 ml of cell suspension was removed from the chemostat and placed in a gas-tight serum bottle for anoxic substrate uptake measurements. To inhibit microbial growth and new enzyme synthesis, kanamycin was added at a concentration of 100 mg l31 [16]. The pH was adjusted to 7.5 with 10% o-phosphoric acid. There was no need to add nitrate since the residual nitrate concentration 31 in the chemostat was 70^80 mg NO3 3 N l , which was su⁄cient to meet all electron-accepting needs during the assay. As soon as a sample was placed in a serum bottle it was purged with N2 . After 15 min of purging, either benzoate or pyruvate was added to the serum bottle to provide a concentration of 250 mg COD l31 . Purging was continued for an additional 15 min, but was then stopped and the ¢rst sample was collected. Two additional samples were taken at 10-min intervals. The 1.5-ml samples were added to mini centrifuge tubes containing 20 Wl 10% o-phosphoric acid. The resulting drop in pH stopped the microbial activity immediately. Samples were centrifuged to remove the cells and the supernatant was analyzed for pyruvate or benzoate with HPLC. The rate was determined from the linear region of the substrate versus time plot and divided by the biomass concentration to give the speci¢c anoxic benzoate or pyruvate uptake rate (AnBUR or AnPUR, respectively). Since both substrate and nitrate were provided in excess, the measured rates represented the maximum values at which the cells could remove the substrates. A similar procedure was used to measure the speci¢c aerobic benzoate or pyruvate uptake rates (AeBUR or AePUR, respectively), except that the samples were purged with air rather than with N2 . Su⁄cient DO was present at all times to prevent its concentration from limiting the substrate uptake rate. 2.5. Aerobic enzyme assays Aliquots (60 ml) of biomass were taken from the experimental chemostats, centrifuged, resuspended, and mechanically disrupted with a Mini-beadbeater1 (Biospec Products, Bartlesville, OK, USA) using 0.1-mm glass beads. The disrupted cells were removed by centrifugation

FEMSLE 11358 27-1-04

62

O. C|nar / FEMS Microbiology Letters 231 (2004) 59^65

(11 000Ug for 15 min) and the CFE was collected and tested for protein concentration and enzyme activities. Protein concentration was measured by the bicinchoninic acid^copper reaction [15]. The activities of all of the enzymes were quanti¢ed by measuring changes in absorbance with time in a Beckman DU-640 (Fullerton, CA, USA) UV recording spectrophotometer. The products of C12DO and C23DO were monitored when catechol was provided as the substrate [17]. They were cis,cis-muconic acid (257 nm) and 2-hydroxy-cis,cis-muconic semialdehyde (375 nm), respectively. The activity of G12DO was quanti¢ed by measuring the formation of its product, maleylpyruvate, at 337 nm [18]. The activity of P34DO was quanti¢ed by measuring the consumption of protocatechuate at 290 nm [19], whereas the activity of P45DO was quanti¢ed by measuring the generation of 4-carboxy-2-hydroxymuconic semialdehyde at 410 nm [20]. Because oxygen was a reactant for all of the enzymes, all solutions added, as well as the CFEs, were saturated with oxygen prior to being mixed together. Various solutions were required for each assay, as listed in the cited references. All enzyme activities were normalized to the total protein concentrations to obtain speci¢c enzyme activities. Appropriate positive and negative controls were periodically run to ensure that the assays were working properly. Given the amount of biomass used to obtain the CFEs, the extinction coe⁄cients of the measured chemicals, and the sensitivity of the spectrophotometer used, the lowest detectable enzyme activities were: C12DO, 0.52 pmol catechol (Wg protein min)31 ; C23DO, 0.38 pmol catechol (Wg protein min)31 ; G12DO, 0.58 pmol gentisate (Wg protein min)31 ; P34DO, 1.37 pmol protocatechuate (Wg protein min)31 ; and P45DO, 0.47 pmol protocatechuate (Wg protein min)31 . Since 1 mol of benzoate can be converted into 1 mol of catechol, gentisate, or protocatechuate, depending on the pathway employed, each of these activities is equivalent to the activity expressed as pmol benzoate (Wg protein min)31 . Based on the average protein concentration in the chemostat and the mass £ow rate of benzoate in the feed, the speci¢c removal rate of benzoate was approximately 16.5 pmol benzoate (Wg protein min)31 , suggesting that the aerobic enzyme assays were su⁄ciently sensitive to detect the enzymes should they be expressed.

The lack of any detectable benzoate or pyruvate during a cycle indicates that the bacteria were well adapted to the cyclic condition (the lower detection limits for benzoate and pyruvate were 0.1 mg COD l31 and 0.02 mg COD l31 , respectively). The chemostat performance was also monitored by measuring TOC, nitrate, and nitrite concentrations during several complete cycles. The baseline TOC concentration was approximately 24 mg C l31 when the feed contained 50% benzoate and around 17 mg C l31 when the feed contained 70% benzoate and 90% benzoate. Consequently, these TOC levels in the chemostat did not change signi¢cantly during a cycle, indicating that metabolic intermediates did not transiently accumulate. During anoxic periods, the residual TEA was present in the form of nitrate, suggesting that the bacterial community was composed of complete denitri¢ers. Overall, performance of the chemostats was excellent, demonstrating that bacterial cultures can e¡ectively remove high concentrations of aromatic compounds while being grown under alternating aerobic/anoxic conditions. To learn more about the response of the cultures under di¡erent feed compositions, samples were taken to establish the metabolic capabilities of the bacteria for benzoate and pyruvate throughout a cycle and to quantify the level of key enzymes present. The benzoate uptake capabilities of the whole cells from the chemostat fed three di¡erent feed compositions were measured under anoxic and aerobic conditions to better understand the metabolism of benzoate, and the data obtained are shown in Fig. 1. There it can be seen that the AnBUR at the end of the anoxic period (or the very beginning of the aerobic period) was increased signi¢cantly as the ratio of benzoate in the feed was increased from 50% to 90% as expected. In addition, after the mixed cultures fed with di¡erent benzoate concentrations were exposed to the 3-h aerobic period, the AnBUR of all cultures decreased to around 13 pmol benzoate (Wg protein min)31 .

3. Results and discussion When the chemostat was operated with a 3-h aerobic and a 9-h anoxic period under di¡erent feed compositions, the redox conditions via ORP levels and DO concentration were monitored throughout the cycle. Since the chemostat was purged with N2 immediately upon initiation of the anoxic period the DO concentration dropped rapidly to zero. However, the ORP changed more slowly because it re£ects the net condition established by all oxidized and reduced chemical species in the environment.

Fig. 1. AnBUR and AeBUR pro¢les in the cyclic chemostat operated with three di¡erent feed compositions. The error bars denote the standard deviations.

FEMSLE 11358 27-1-04

O. C|nar / FEMS Microbiology Letters 231 (2004) 59^65

Fig. 2. AnPUR and AePUR pro¢les in the cyclic chemostat operated with three di¡erent feed compositions. The error bars denote the standard deviations.

When these cultures were re-exposed to anoxic conditions, they resumed their AnBUR. There it can also be seen that the AeBUR at the end of the aerobic period (or the very beginning of the anoxic period) increased signi¢cantly as the ratio of benzoate in the feed was increased from 50% to 90% as expected. Moreover, the AeBUR of the culture fed with 50% benzoate increased slightly during the aerobic period and declined slightly during the anoxic period. However, the AeBUR of the cultures fed with higher benzoate concentrations increased more sharply during the aerobic period and declined more sharply during the anoxic period. As mentioned earlier, bacteria must use di¡erent terminal electron transport enzymes under the two redox conditions. Thus, the possibility exists that changes in the levels of those enzymes might in£uence the observed benzoate uptake capability. To check this possibility, the uptake capabilities for pyruvate under anoxic and aerobic conditions were also measured. Since pyruvate is not an intermediate in the anoxic and aerobic biodegradation of benzoate [12], this suggests that pyruvate was used constitutively by the culture. The levels of constitutive enzymes are generally stable in bacterial cells, being required for a variety of purposes. This suggests that any decrease in the ability of the cells to take up pyruvate in batch assays would most likely be due to loss of the nitrogen oxide reductases or cytochrome oxidases. Conversely, a stable AnPUR or AePUR would be suggestive of stable nitrogen oxide reductases and cytochrome oxidases, respectively. Therefore, assays for AnPUR and AePUR were also run and presented in Fig. 2. The results presented in Fig. 2 show that both AnPUR and AePUR were relatively constant throughout a cycle, suggesting that the cells maintained their cytochrome oxidases during anoxic conditions and their nitrogen oxide reductases during aerobic conditions although the AnPUR and AePUR for the cultures fed with smaller amount of pyruvate increased slightly as

63

not expected. The reason for this is unknown, although it was most likely due to di¡erences in the microbial communities present in the chemostat. Kaewpipat and Grady [21] have shown that mixed microbial communities are highly dynamic, and the observed changes are consistent with that characteristic. This maintenance of both terminal electron transport enzyme systems was one factor that allowed the bacteria to switch rapidly from aerobic to anoxic conditions, and vice versa. Thus, the lags suggested by others [1,3] were not signi¢cant within the sampling interval employed herein. In addition, the observed AePUR and AnPUR values were relatively higher than the speci¢c benzoate removal rate in the chemostat and the AeBUR and AnBUR values in Fig. 1, suggesting that transfer of electrons to the terminal acceptors did not affect the AeBUR and AnBUR data. To better understand the metabolism of benzoate, aerobic ring cleavage enzymes involved in the metabolism of benzoate were screened. C23DO was the only aerobic ring cleavage enzyme expressed. The data obtained are shown in Fig. 3. Since the culture was a mixed microbial community I expected to have more than one pathway present, as Ma and Love [22] had experienced. Thus, to be certain that the enzyme assays were working properly and with su⁄cient sensitivity, tests were performed with pure cultures known to express each of the key enzymes. These tests con¢rmed that signi¢cant levels of the other aerobic ring cleavage enzymes were indeed absent from the culture. The level of C23DO present in the culture throughout a cycle as measured under three di¡erent feed compositions is presented in Fig. 3. The level of C23DO at the end of the aerobic period (or the very beginning of the anoxic period) increased signi¢cantly as the ratio of benzoate in the feed was increased from 50% to 90% as expected, but the increase in the level of C23DO for 70% benzoate was higher than that for 90% benzoate. The reason for this is unknown, although it was most likely due to di¡erences in the microbial communities present in the chemostat [21]. The other thing of note is that the cells signi¢cantly lost their enzyme throughout the 9-h anoxic period although this is not seen well for the culture receiving 50% benzoate in Fig. 3 due to the wider scale on the

Fig. 3. C23DO activity pro¢le in the cyclic chemostat operated with three di¡erent feed compositions. The enzyme activity is expressed per unit of protein in the CFE. The error bars denote the standard deviations.

FEMSLE 11358 27-1-04

64

O. C|nar / FEMS Microbiology Letters 231 (2004) 59^65

y-axis. This is consistent with the ¢nding of Ma and Love [22], who observed a loss of C23DO during anoxic periods. It should be emphasized that C23DO was not active during the anoxic period because there was no oxygen which is a reactant for C23DO. It is signi¢cant that the C23DO was retained because it allowed the cells to respond rapidly upon being returned to aerobic conditions. As Kovarova-Kovar and Egli [9] postulated three strategies for the synthesis of catabolic enzymes by bacteria growing on a multicomponent substrate in continuous culture, the bacterial cultures in this study employed the second postulated strategy in dealing with mixtures of benzoate and pyruvate; that is, both the benzoate-degrading capabilities and the in vitro activities of C23DO were in direct proportion to the £ux through the chemostat of the substrate degraded by the pathway containing C23DO with some exceptions. Similar observations were made by Rudolph and Grady [23]. The most surprising thing about the data in Fig. 3 is the drop in the level of C23DO upon return of the culture to aerobic conditions. A possible reason for this response is the production of H2 O2 by the bacteria. Bacteria commonly produce H2 O2 and superoxide while using oxygen as an electron acceptor, but are protected from them by catalase and superoxide dismutase, which are enzymes that destroy them. It is possible that the bacteria in the culture stopped producing catalase and/or superoxide dismutase during the 9-h anoxic period, making them susceptible to the oxidative e¡ects of H2 O2 or superoxide upon their return to aerobic conditions. Since oxygen is a reactant for C23DO, one might expect it to be particularly susceptible to oxidative attack, and indeed, Hasset et al. [24] have shown that C23DO is very susceptible to inactivation by H2 O2 through oxidation of the essential active site ferrous iron. Studies with our CFE (data not shown) found that C23DO activity was completely stopped when it was exposed to 200 WM H2 O2 . After 1.0^1.5 h the cells started increasing their levels of C23DO, suggesting that they were again synthesizing the enzymes to protect them from the toxic products of oxygen. It should be noted that even the lowest level of C23DO observed was signi¢cantly greater than the amount required to metabolize the benzoate entering the chemostat. This could explain why no benzoate was observed in the chemostat, even though the C23DO level dropped.

4. Summary and conclusions Both the benzoate-degrading capabilities and the in vitro activities of C23DO were in direct proportion to the £ux through the chemostat of the substrate degraded by the pathway containing C23DO. Cyclic bacterial cultures grown under various feed compositions can e¡ectively remove benzoate and pyruvate without the accumulation of metabolic products.

It can be concluded that variation in the amounts of aromatic compounds in the feed is not likely to be a problem because the data clearly demonstrate the robustness of bacterial cultures grown in cyclic aerobic/anoxic environments under di¡erent feed compositions. Acknowledgements º .C. was Research Assistant At the time of this study, O Professor in the Department of Environmental Engineering and Science at Clemson University, Clemson, SC, USA. References [1] Baumann, B., Snozzi, M., Zehnder, A.J.B. and van Der Meer, J.R. (1996) Dynamics of denitri¢cation activity of Paracoccus denitri¢cans in continuous culture during aerobic-anaerobic changes. J. Bacteriol. 178, 4367^4374. [2] Kornaros, M. and Lyberatos, G. (1998) Kinetic modelling of Pseudomonas denitri¢cans growth and denitri¢cation under aerobic, anoxic and transient operating conditions. Water Res. 32, 1912^1922. [3] Liu, P.H., Svoronos, S.A. and Koopman, B. (1998) Experimental and modelling study of diauxic lag of Pseudomonas denitri¢cans switching from oxic to anoxic conditions. Biotechnol. Bioeng. 60, 649^655. [4] Assinder, S.J. and Williams, P.A. (1990) The TOL plasmids ^ Determinants of the catabolism of toluene and the xylenes. Adv. Microb. Physiol. 31, 1^69. [5] Altenschmidt, U., Oswald, B., Steiner, E., Herrmann, H. and Fuchs, G. (1993) New aerobic benzoate oxidation pathway via benzoyl-coenzyme and 3 hydroxybenzoyl-coenzyme A in a denitrifying Pseudomonas sp. J. Bacteriol. 175, 4851^4858. [6] Harwood, C.S. and Parales, R.E. (1996) The L-ketoadipate pathway and biology of self-identity. Annu. Rev. Microbiol. 50, 553^590. [7] Heider, J. and Fuchs, G. (1997) Anaerobic metabolism of aromatic compounds. Eur. J. Biochem. 243, 577^596. [8] Egli, T. (1995) The ecological and physiological signi¢cance of the growth of heterotrophic microorganisms with mixtures of substrates. Adv. Microb. Ecol. 14, 305^386. [9] Kovarova-Kovar, K. and Egli, T. (1998) Growth kinetics of suspended microbial cells: From single-substrate-controlled growth to mixed-substrate kinetics. Microbiol. Mol. Biol. Rev. 62, 646^666. [10] Arora, P., Kumar, R.A. and Venkatesh, K.V. (1999) Analysis of the optimal model for substrate substitutability in continuous microbial cultures. Chem. Eng. Sci. 54, 987^997. [11] Gottschalk, G. (1986) Bacterial Metabolism, 2nd edn. Springer, New York. [12] Harwood, C.S., Burchhardt, G., Herrmann, H. and Fuchs, G. (1999) Anaerobic metabolism of aromatic compounds via benzoyl-CoA pathway. FEMS Microbiol. Rev. 22, 439^458. [13] Clesceri, L.S., Greenberg, A.E. and Eaton, A.D. (Eds.) (1998) Standard Methods for the Examination of Water and Wastewater, 20th edn. APHA, AWWA, and WEF, Washington, DC. [14] Coschigano, P.W., Haggblom, M.M. and Young, L.Y. (1994) Metabolism of both 4-chlorobenzoate and toluene under denitrifying conditions by a constructed bacterial strain. Appl. Environ. Microbiol. 60, 989^995. [15] Daniels, L., Hanson, R.S. and Philps, J.A. (1994) Chemical Analysis. In: Methods for General and Molecular Bacteriology (Gerhardt, P., Ed.), pp. 512^554. American Society for Microbiology, Washington, DC. º . and Grady Jr., C.P.L. (2001) Aerobic and anoxic biodeg[16] C|nar, O radation of benzoate : Stability of biodegradative capability under endogenous conditions. Water Res. 35, 1015^1021.

FEMSLE 11358 27-1-04

O. C|nar / FEMS Microbiology Letters 231 (2004) 59^65 [17] Dorn, E. and Knackmuss, H.J. (1978) Chemical structure and biodegradability of halogenated aromatic compounds. Two catechol 1,2dioxygenases from a 3-chlorobenzoate-grown Pseudomonad. Biochem. J. 174, 73^84. [18] Crawford, R.L., Hutton, S.W. and Chapman, P.J. (1975) Puri¢cation and properties of gentisate 1,2-dioxygenase from Moraxella osloensis. J. Bacteriol. 121, 794^799. [19] Fujisawa, H. and Hayaishi, O. (1968) Protocatechuate 3,4-dioxygenase. I. Crystallization and characterization. J. Biol. Chem. 243, 2673^ 2681. [20] Ono, K., Nozaki, M. and Hayaishi, O. (1970) Puri¢cation and some properties of protocatechuate 4,5-dioxygenase. Biochim. Biophys. Acta 220, 224^238.

65

[21] Kaewpipat, K. and Grady, C.P.L. Jr. (2001) Population dynamics in laboratory-scale activated sludge reactors. Proceedings of the 3rd IWA International Specialized Conference on Microorganisms in Activated Sludge and Bio¢lm Process, Rome, Italy. [22] Ma, G. and Love, N. (2001) BTX biodegradation in activated sludge under multiple redox conditions. J. Environ. Eng. 127, 509^516. [23] Rudolph, J.M. and Grady Jr., C.P.L. (2002) Catabolic enzyme levels in bacteria grown on binary and ternary substrate mixtures in continuous culture. Biotechnol. Bioeng. 79, 188^199. [24] Hassett, D.J., Urs, A.O., Groce, S.L., Parvatiyar, K., Ma, J. and Lipscomb, J.D. (2000) Hydrogen peroxide sensitivity of catechol 2,3-dioxygenase : a cautionary note on use of xylE reporter fusions under aerobic conditions. Appl. Environ. Microbiol. 66, 4119^4123.

FEMSLE 11358 27-1-04