Microbial Ecology

A Bioreactor for Growth of Sulfate-Reducing Bacteria: Online Estimation of Specific Growth Rate and Biomass for the Deep-Sea Hydrothermal Vent Thermophile Thermodesulfatator indicus
Joost Hoek1,2, Donald Caneld3, Anna-Louise Reysenbach2 and Lnsmann Iversen4
(1) (2) (3) (4) Department of Earth and Environmental Science, University of Pennsylvania, Philadelphia, PA, USA Department of Biology, Portland State University, Portland, OR, USA Nordic Center for Earth Evolution (NordCEE) and Institute of Biology, University of Southern Denmark, Odense, Denmark Department of Biochemistry and Molecular Biology, University of Southern Denmark, Odense, Denmark

Received: 19 January 2005 / Accepted: 22 January 2005 / Online publication: 28 April 2006

Introduction

The accurate and reliable measurement of biomass is an essential requirement during growth experiments. With balanced exponential growth there is a rectilinear relationship between biomass concentration and the rate of change of substrate and product concentrations in the culture medium, which allows for indirect growth measurements. The uptake of substrates and formation of products are often associated with changes in the proton concentration in the culture medium, and the rate of acid or base titrant addition to maintain constant pH is a measure of the rate of biomass formation. Recording titrant addition therefore provides a simple method for measuring specic growth rates, changes in biomass concentration, and other aspects of energetics and metabolic function [8]. Titrant addition has been used for many years [9], to assess the growth parameters of a variety of different microorganisms [7, 8, 17, 18, 21]. For example, titrant addition has been used to estimate biomass concentrations and specic growth rates in cultures of Escherichia coli, denitrifying bacteria [4, 8], fungi [7, 14, 21], and green algae [22], and for estimation of acidic fermentation products in batch cultures of lactic acid bacteria [1, 10]. To our knowledge, titrating metabolic proton exchange has not been used for online growth measurements in experiments for sulfate-reducing bacteria. Nagpal et al. [12], correlated the production of H2S, _ a metabolic end product of SO42 reduction, to biomass production. The authors channeled the efuent gas of a bioreactor containing biogenic H2S purged from the liquid medium through a ZnSO4 solution that trapped H2S as ZnS(solid). We have modied this approach by
Correspondence to: Joost Hoek; E-mail: joost.hoek@gmail.com

noting that the formation of ZnS in a ZnCl solution results in acidication of the solution by the following reaction: ZnCl2 H2 Sgas $ ZnS 2HCl 1

Sodium hydroxide is titrated into the H2S trap to maintain a constant pH, and a computer program controlling the process records the NaOH additions, which allows for continuous online estimation of the change in biomass concentration and the specic growth rate of the culture, assuming balanced growth. In principle, this system can be used to measure growth and quantify the energetics of any H2S-producing organism and is particularly suited for experiments where H2S is required for further analyses. For example, this system is ideal for experiments that measure the controls on the fractionation of sulfur isotopes under controlled environmental conditions. We have used this experimental conguration to determine the growth kinetics of Thermodesulfatator indicus strain CIR29812, a chemolithotrophic, thermophilic, sulfate-reducing bacterium recently isolated from a deep-sea hydrothermal vent on the Central Indian Ridge [11]. T. indicus was grown in a reactor under fed-batch culture with H2 as electron donor and CO2 as primary carbon source. These experiments were designed to measure growth kinetics under high and low H2 ux.

Methods
Culture Conditions. T. indicus was maintained in anaerobic saltwater carbonate buffered medium with H2 as the electron donor and CO2 as the primary carbon source. The medium contained (per liter of doubleSpringer Science+Business Media, Inc. 2006

470

DOI: 10.1007/s00248-006-9046-8

& Volume 51, 470478 (2006) & *

J. HOEK

ET AL.:

A BIOREACTOR

FOR

GROWTH

OF

SULFATE-REDUCING BACTERIA: ONLINE ESTIMATION

OF

SPECIFIC GROWTH RATE

AND

BIOMASS

FOR THE

DEEP-SEA

471

deionized H2O) 20 g NaCl, 3 g MgClI6H2O, 0.15 g CaCl2I2H2O, 4.0 g Na2SO4, 0.25 g NH4Cl, 0.2 g KH2PO4, 0.5 g KCl, 1 mL of trace element mixture [11], 1 mL vitamin solution [11], 1 mL 10% yeast extract, and 1 mL of 2 M sodium acetate solution. The medium was prepared anaerobically, and 50 mL was transferred into 100-mL serum bottles. The headspace was lled with a H2/CO2 mixture (80:20 v/v) to a pressure of 2 atm. The pH of the medium was adjusted to 6.25 with 5 N HCl. Media for the growth experiments in the bioreactor were prepared as described above except that a CO2/N2 mixture (80:20 v/v) and different concentrations of H2 were continuously bubbled into the bioreactor. Fifty milliliters of exponentially grown (70-C) culture was aseptically transferred to the bioreactor at the start of each experiment.
Bioreactor. Growth experiments were carried out using a 1.2-L bioreactor with a 1-L working volume. The bioreactor consisted of two sections: a glass container (Schott, Mainz, Germany) and a custom-designed glass lid (Schott) built with access ports (Fig. 1). The two sections were joined with a butyl rubber O-ring and clamped together. The bioreactor was equipped with a Pt 100 temperature sensor (Kika Labortechnik, Staufen, Germany) and an autoclavable glass pH electrode (Metler, Toledo, OH, USA). The temperature was kept constant at 70-C by means of a thermostat (Kika Labortechnik) that controlled the temperature of a hot plate positioned under the bioreactor. Multiple temC C

perature sensors placed in different positions in the bioreactor measured no temperature differences within the culture medium, which is attributed to the rapid mixing of the reactor volume. The pH was maintained at 6.25 by means of a pH controller (pH/ORP controller 3672, Jenco, Hainault, UK), which activated a peristaltic pump (Ole Dich, Hvidovre, Denmark) delivering 1 M HCl. Mixing was performed by a Teon-coated magnetic stir bar spinning at 900 rpm. Mixing time was assumed to be on the order of a few seconds based on mixing times that have been previously characterized for bioreactors using a similar stirring mechanism and having a similar or larger working volume [3]. Hydrogen gas was transferred to the medium via a membrane bubbler system (Sanitaire EPDM rubber, Water Pollution Control Corp., Milwaukee, WI, USA) designed for controlled gas transfer at a range of gas ow rates [5]. The ow rates of H2 into the reactor were changed for different _ experiments and ranged from 2 to 10 mL min 1. The CO2/N2 mixture was transferred to the culture medium through a glass orice bubbler. The ow rate of CO2/N _2 through the reactor was held constant at 200 mL min 1 for all experiments. Gas ow rates were regularly calibrated by an ADM 1000 ow meter (Agilent, Palo Alto, CA, USA). The efuent gas, containing a mixture of N2, CO2, H2, and biogenic H2S, was passed through a condenser at 4-C to reduce evaporation from the culture medium and subsequently bubbled through a sulde trap containing 2 L of anaerobic 5% ZnCl2. The H2S reacted with ZnCl2 and formed ZnS(solid). The H2S trap

C D D N2/ CO2 HCl pH control C A C

C pH control A B A B

NaOH

H2

Bioreactor

ZnCl2 Trap

Secondary ZnCl2 Trap

Figure 1. Schematic diagram of the reactor setup. (A) Peristaltic pump, (B) clamp, (C) 0.2-mm lter, (D) low-pressure gas regulator/ gas ow meter.

472

J. HOEK

ET AL.:

A BIOREACTOR

FOR

GROWTH

OF

SULFATE-REDUCING BACTERIA: ONLINE ESTIMATION

OF

SPECIFIC GROWTH RATE

AND

BIOMASS

FOR THE

DEEP-SEA

was equipped with a glass pH electrode (Metler) connected to a pH meter (Radiometer, Copenhagen, Denmark). To ensure that there was no loss of H2S from the trap, the efuent gas from the rst trap was transferred to a second H2S trap. This trap was visually monitored for ZnS precipitation; however, no precipitation was observed in the second trap during the course of the experiments.
Titration System. The titration system of the primary H2S trap is shown in Fig. 1. When the pH drops below a set-point value (5.0), due to the accumulation of HCl, pulses of 1 M NaOH are titrated into the trap until the pH returns to the set-point value. Sodium hydroxide is added from burettes through Viton tubing via a peristaltic pump (Ole Dich) controlled by a computer. The titrant was held under a N2 atmosphere to a pressure of 1.2 atm by using a low-pressure regulator. This prevented the formation of gas bubbles in the tubing and maintained the titrant under anaerobic conditions. The mean volume delivered per pulse was determined by dividing the total volume dispensed from the burette at the end of each experiment by the total number of pulses. With this titration system, ZnCl2 can react with NaOH to form a ZnOH2 sludge, and an unbuffered solution is therefore easy to overtitrate. Nevertheless, the success of the titration system was shown in the reliability of the titration curve, i.e., in how well it matched sulde accumulation, indicating that overtitration was not a signicant problem. Data Recording and Processing. A computer program for data recording and processing and for control of pH and titrant pulse addition in the H2S trap was written in Microsoft QuickBasic 4.5 [4]. The pH and the number of pulses of acidic and basic titrant added since the beginning of the experiment were recorded at 10-s intervals. Determination of Specic Growth Rates. The log of slope (LOS) plot was used to calculate specic growth rates of the culture. LOS plots were implemented for our experimental data by using a specially written program [14], which takes the natural log of the rst derivative (slope) of an exponential function. This method is preferred because experimental data described by an exponential function often contain an offset (C):

By taking the natural logarithm, loge, to the slope:

loge dy =dx loge y 0 k k t t 0 ;

the rst-order rate constant (k) is found as the slope of any rectilinear section in a plot of the natural logarithm of the slope, loge(dy/dx), against t, dened as a loge of slope (LOS) plot. This is possible without knowing which part of the data set is exponential, and any exponential section of the curve can be used to nd the rst-order rate constant. Alternatively, direct natural log transformation of the titration data can be used.
Off-line Growth Measurements. Samples for biomass measurements, SO2 4 , and H2S concentration were collected in sterile anaerobic 15-mL Hungate tubes. Growth was followed by measuring the turbidity of the culture and by SO2 4 and H2S analysis. Culture turbidity was measured spectrophotometrically at 595 nm in the sampling tubes immediately after sample collection. After absorbance was determined, samples were immediately frozen to stop all biological activity and stored for SO2 4 and H2S analysis. The SO2 4 concentration of the culture was measured by suppressed ion chromatography. A Sykam LAC A14 column (3 250 mm, 70-C) was used, with 5 mM NaCO3 with 1% hydroxybenzonitrile (v/v) as eluent at a ow rate of 1.5 mL min1. The concentration of H2S was determined spectrophotometrically at 670 nm on ZnS samples with the methylene blue technique [2]. The variability in H2S measurements is estimated to be 5%, resulting mainly from sampling error because ZnS is a solid and is difcult to completely homogenize. Mathematical Models. Van Houten et al. [20] developed kinetic models for hydrogen-utilizing sulfate reducers. The models assume that growth of biomass proceeds according to Monod kinetics with simultaneous inhibition by biogenic H2S. Product formation is directly coupled to biomass production, and substrate consumption for maintenance is incorporated into the overall biomass yield. Therefore, during exponential growth the reaction equations are as follows:

dX =dt 2X ; dSO4 =dt Y S=X 2X ; dH2 =dt Y H2 =X 2X ; dH2 S=dt Y H2 S=X 2X ;

5 6 7 8

y y0 exp k t t 0 C :

By differentiating Eq. (2) the offset is removed and the slope is obtained:

dy =dx y0 k exp k t t 0 :

where X (cells mL1) is biomass concentration, m (h1) is specic growth rate, SO4 (mM) is sulfate concentration, H2 (mM) is hydrogen concentration, H2S (mM) is hydrogen sulde concentration, and YS/X is a stoichiometric constant specic to the substrate.

J. HOEK

ET AL.:

A BIOREACTOR

FOR

GROWTH

OF

SULFATE-REDUCING BACTERIA: ONLINE ESTIMATION

OF

SPECIFIC GROWTH RATE

AND

BIOMASS

FOR THE

DEEP-SEA

473

A linear relationship between biomass concentration and the rate of change of the concentration of substrates or products during exponential growth makes indirect growth measurements possible. The amount of base titrant required to maintain constant pH in the H2S trap is proportional to the changes in substrate (SO2 4 ) or product (H2S) concentrations and, hence, is a measure of the biomass production, whereas the rate of titrant addition reects the specic growth rates. Recording titrant addition in the H2S trap therefore provides a simple method for indirect growth measurements. The rate of proton formation in the H2S trap is given by: dH =dt Y H=X 2X ; 9
where H is the proton concentration and YH/X is a stoichiometric constant.

The amount of proton production from time t0 to time t is given by: H t H 0 exp 2t ;
where

10 11

H 0 Y H=X X 0 :

Results
High H2 Flux Growth Experiments. Three sets of experiments were conducted in which titrant addition to the H2S trap was recorded during fed-batch growth of T. indicus under nonlimiting substrate conditions. In each of the experiments, the _gas ow of the CO2/N2 gas mixture was 200 mL min 1, and the gas ow of H2 was

Figure 2. Growth of T. indicus under high H2 ux conditions. (A) () NaOH titration response curves, () LOS data for titration response curve. 2 (B) Off-line growth measurements; () SOj 4 concentration, ()) H2S concentration, and (q) biomass. Arrows indicate time of inoculation of starting culture. Vertical dashed line indicates termination of exponential growth. Error bars indicate standard deviation.

474

J. HOEK

ET AL.:

A BIOREACTOR

FOR

GROWTH

OF

SULFATE-REDUCING BACTERIA: ONLINE ESTIMATION

OF

SPECIFIC GROWTH RATE

AND

BIOMASS

FOR THE

DEEP-SEA

10 mL min 1. To ensure steady-state conditions, the system was allowed to equilibrate for approximately 24 h prior to inoculation. The formation of ZnS caused acidication of the H2S trap, which was compensated for by the addition of basic titrant. Cumulative titrant addition (Fig. 2A, results shown for one of the experiments) for all the experiments was exponential with the same kinetics as the biomass accumulation rate, SO2 4 uptake rate, and H2S production rate (Fig. 2B, results shown for one of the experiments). Sulde concentrations in the culture media remained below 40 mM during the course of the experiment, indicating that H2S was effectively purged by the relatively high CO2/N2 ow rates. These low concentrations of H2S are not in a range where they would be expected to inhibit growth [13, 16, 20]. The base titration rate was used to estimate the specic growth rate by the LOS method [14] for all the experiments (Table 1). To minimize the variability resulting from sample noise, we calculated the slope for every 10 data points generated in the base titration data set. Nonlinear sections of the LOS plot prior to the initiation of exponential growth were removed. The LOS data indicate that exponential growth initiated soon after inoculation of the culture and continued for approximately 15 h. The scatter in the LOS data during the initial growth phase results from the low number of NaOH additions in this period. Best ts to rectilinear sections of LOS data for all experiments yielded specic _ growth rates of (0.30.36) T 0.02 h 1 (error reported as standard error) (Table 1), which is in agreement with previous reports of the specic growth rate of T. indicus [11]. This is one of the highest growth rates recorded for sulfate-reducing bacteria [15]. An abrupt break in the slope 15 h after inoculation indicates that exponential growth ended. In one of the experiments (Fig. 2), the LOS data levels off and becomes positive again soon after the initial break in slope, suggesting that exponential growth has initiated once more, albeit at a much lower specic growth rate (0.015 h1, r 2 = 0.056, P = 0.275). The cause for this transition remains to be determined, although it may have resulted from a switch to the utilization of an unidentied metabolic intermediate. This secondary exponential growth phase was not observed in any other growth experiments. Cessation of growth, which is seen as a near stop in base titration after approximately 40 h, is associated with a rapid decrease in the LOS data. A prerequisite for estimating biomass concentration and specic growth rate from base titrations is constant stoichiometry between proton exchange and biomass formation, substrate consumption, and product formation. For all of our batch culture experiments, culture turbidity, SO2 4 utilization, and H2S production were compared to corresponding base titration. The results for

one of the experiments are shown in a concordance plot (Fig. 3), and the linear relationships conrm that the growth kinetics follows titrant addition, SO2 4 consumption, and H2S production. As dissimilatory sulfate reduction is thought to reduce SO2 4 completely to H2S, a 1:1 stoichiometry is expected between the amount of SO2 4 consumed and the amount of H2S produced. Figure 4 shows the relationship between SO2 4 consumption and H2S formation. The stoichiometric coefcient, 0.95 T 0.04 (error reported as standard error), was obtained from the slope of a regression line tted to the data (r 2 = 0.913, P G 0.0001). These results are consistent, within the experimental error, of expected stoichiometry. However, there is a distinct change in the slope before and after the termination of the rst exponential growth phase at 15 h (Fig. 4, marked with a vertical dashed line). The stoichiometric coefcient obtained during the initial exponential growth phase is 0.68 T 0.14 (r 2 = 0.956, P G 0.0001), and for the secondary exponential growth phase is 1.25 T 0.14 (r 2 = 0.945, P = 0.0011). A similar result was observed when comparing SO2 4 utilization to the accumulated volume of NaOH titrated into the H2S trap (Fig. 4). A 2:1 stoichiometry is expected between SO2 4 consumption in the reactor and HCl production in the H2S trap [see Eq. (1)]. The stoichiometric coefcient estimated from the slope of a regression line tted to the data was 2.2 T 0.06 (r 2 = 0.955, P G 0.0001). The stoichiometric coefcient obtained for the overall data set is larger than expected; this may result from precipitation of a small amount of ZnCO3 in the H2S trap or, alternatively, as a result of overtitration of ZnCL2 with NaOH due to the precipitation of ZnOH2. The stoichiometric coefcients obtained from the slope of regression lines tted to the initial exponential growth phase and the secondary exponential phase were 1.7 T 0.09 (r 2 = 0.958, P G 0.0001) and 2.9 T 0.06 (r 2 = 0.951, P = 0.0012), respectively. The results from these experiments suggest, again, that not all the SO2 4 consumed during the

Table 1. Specic growth rates (hj1) calculated for each growth experiment of T. indicus using the log of slope (LOS) data (see text for explanation)

Experiment No. High H2 ux 1 2 3 Low H2 ux 4 5

H2 ow rate (mL minj1) 10 10 10 2 1.5

Specic growth rate (hj1) 0.36 0.36 0.33 0.009 0.008

r2 0.89 0.98 0.96 0.91 0.12

P G0.0001 G0.0001 G0.0001 G0.0001 0.309

Specic growth rates were calculated from linear regression analysis to linear sections of the LOS data.

J. HOEK

ET AL.:

A BIOREACTOR

FOR

GROWTH

OF

SULFATE-REDUCING BACTERIA: ONLINE ESTIMATION

OF

SPECIFIC GROWTH RATE

AND

BIOMASS

FOR THE

DEEP-SEA

475

indicus under high H2 ux conditions. (---) Linear regression line for SO4 2 utilized, slope = 14.69, r2 = 0.963, P G 0.0001. () Linear regression line for titrant added, slope = 4288, r2 = 0.986, P G 0.0001.

Figure 3. Concordance plot of biomass vs titrant addition (q) and SO4 2 consumption () for the initial exponential growth phase of T. _

initial exponential growth phase was completely reduced to H2S, and that an intermediate product accumulated, which was later metabolized to H2S during the second period of exponential growth. The presence of metabolic intermediates was not recognized until after analyses had been completed and therefore was not examined. Subsequent experiments did not detect the presence of sulfur intermediates.
Low H2 Flux Growth Experiments. A second set of experiments was carried out to measure growth rates and biomass concentration of T. indicus at low H2 ux rates. Two fed-batch culture experiments were run under the same conditions described above except that H2 ow was adjusted to 2 and 1 mL minj1, respectively (Table 1). The reactor was allowed to equilibrate for a period of 24 h prior to inoculation. In these experiments computer control was not activated until just before inoculation. Cumulative titrant addition (Fig. 5A) for both experiments was exponential with the same kinetics as the 2 biomass accumulation rate, SOj 4 uptake rate, and H2Sproduction rate (Fig. 5B, off-line measurements shown for one of the experiments). The base titration rate was used to calculate the specic growth rate with the LOS

method as described above. The LOS data (Fig. 5A, plot is shown for one of the experiments) indicate that exponential growth initiated soon after inoculation of the culture and continued until the experiment was terminated after approximately 75 h of growth. The specic growth rate calculated from the best t to the rectilinear segments of the data was (0.0095 and 0.008) T 0.0005 h1 (r 2 = 0.91 and 0.12, respectively, P G 0.0001 and P = 0.309, respectively) (Table 1). The poor t of the LOS data in the second low H2 ux experiment is likely due to the possibility that the H2 concentration in the bioreactor was approaching threshold levels and that growth under these conditions was becoming H2-limited, resulting in high levels of scatter in the LOS plot. A linear relationship between titrant addition and the change in biomass concentration, SO2 4 consumption, and H2S production for all experiments at low H2 ux conrm the conclusion that the kinetics of titrant addition follows growth and metabolism even during extremely slow growth rates. A linear relationship also exists between SO2 4 consumption and H2S formation and the volume of added base. The stoichiometric coefcient obtained from the slope of a regression line tted to the H2S-production data was estimated to be 0.96 T 0.02

476

J. HOEK

ET AL.:

A BIOREACTOR

FOR

GROWTH

OF

SULFATE-REDUCING BACTERIA: ONLINE ESTIMATION

OF

SPECIFIC GROWTH RATE

AND

BIOMASS

FOR THE

DEEP-SEA

2 Figure 4. Estimation of stoichiometry of H2S formation ()) and NaOH titration () with SOj 4 consumption for T. indicus under high H2

ux conditions. () Linear regression of NaOH titration, (---) linear regression of H2S formation. Vertical line, point of termination of initial exponential growth phase.

(r 2 = 0.996, P G 0.0001). These results are consistent with complete and immediate reduction of SO2 4 to H2S. The stoichiometric coefcient estimated from the base titration data was 2.37 T 0.06 (r2 = 0.991, P G 0.0001).

Discussion

The bioreactor described in this study allows for the online estimation of biomass accumulation and energetics of sulfate-reducing bacteria. The amount of acid or base titrant required to maintain a constant pH over a given time interval is an indirect measure of changes in biomass concentration. By monitoring titrant addition to maintain a constant pH, subtle changes in growth parameters, energetics, and metabolic fate can be monitored online. Another advantage of this approach is that the online monitoring of growth provides reliable highresolution data for the calculation of specic growth rates. Furthermore, no specialized apparatus is required, as most bioreactors are equipped with pH control. Because the measurements depend on the amount of active

cells in the whole bioreactor, the method is equally suited for homogeneous bacterial suspensions and biolms. However, because the measurements are indirect, titrant data must be correlated to a direct measure of biomass, such as cell dry weight or cell density. With this experimental setup we were easily able to observe a subtle shift in the metabolic state of the culture in one of the high H2 ux experiments (Fig. 3). This was clearly expressed as a change in specic growth rates in a LOS plot of the base titration data. Furthermore, we were able to correlate this shift in specic growth rates to changes in the stoichiometry of SO2 4 consumption and H2S production. The results from this experiment suggest that accumulation of a metabolic intermediate during the initial exponential growth phase, such as sulte or thiosulfate, provided an alternative source of H2S formation in the H2-limited exponential growth phase [15]. This interpretation is supported by earlier reports on the accumulation of intermediate oxidation states during SO2 4 reduction [6, 19]. In the low H2 ux experiments exponential growth was observed, although at extremely low specic growth

J. HOEK

ET AL.:

A BIOREACTOR

FOR

GROWTH

OF

SULFATE-REDUCING BACTERIA: ONLINE ESTIMATION

OF

SPECIFIC GROWTH RATE

AND

BIOMASS

FOR THE

DEEP-SEA

477

Figure 5. Growth of T. indicus under low H2 ux conditions. (A) NaOH titration response curves: 1, H2 ow = 2 mL min 1; 2, H2 ow =

1 mL min . () LOS data for titration response curve 1. (B) Off-line growth measurements: () SO4 biomass.

_1

_2

utilized, ()) H2S formed, and (q)

rates (0.0095 h1). Thus, H2 limitation suppressed but did not inhibit growth. The stoichiometric relationship between SO2 4 consumption and H2S production remained 1:1 during the entire growth period, suggesting that there was no accumulation of metabolic intermedi2 ates during slow SOj 4 reduction rates limited by electron transfer from H2. The bioreactor described here is particularly suited for laboratory experiments where the H2S that is

generated will be utilized for further analysis. For instance, this setup is appropriate for experiments to measure the fractionation of sulfur isotopes by sulfatereducing bacteria. The primary objective of the work reported in this paper was the development of a bioreactor system that can be used for the growth analysis of bacterial cultures at variable and carefully controlled low H2 gas pressure that allows for continuous sampling of the reaction products.

478

J. HOEK

ET AL.:

A BIOREACTOR

FOR

GROWTH

OF

SULFATE-REDUCING BACTERIA: ONLINE ESTIMATION

OF

SPECIFIC GROWTH RATE

AND

BIOMASS

FOR THE

DEEP-SEA

Acknowledgments

We thank the members of the J. J. L. Iversen and D. E. Caneld laboratories for technical and analytical assistance, and two anonymous reviewers for their helpful comments. This research was supported by a NASAExobiology grant (NAG5-13503) and the Danish National Research Council.

References
1. Benthin, S, Villadsen, J (1996) Amino acid utilization by Lactococcus lactis subsp. cremoris FD1 during growth on yeast extract or casein peptone. J Appl Bacteriol 80: 6572 2. Cline, JD (1969) Spectrophotometric determination of hydrogen sulde in natural waters. Limnol Oceanogr 14: 454458 3. Eriksen, NT, Iversen, JJL (1997) On-line determination of respiration rates of aquatic organisms in a mono-phase oxystat at steady-state dissolved oxygen tensions. Mar Biol 128: 181189 4. Eriksen, NT, Kratchmarova, I, Neve, S, Kristiansen, K, Iversen, JJL (2001) Automatic inducer addition and harvesting of recombinant Escherichia coli cultures based on indirect on-line estimation of biomass concentration and specic growth rate. Biotechnol Bioeng 75: 355361 5. Eriksen, NT, Poulsen, BR, Iversen, JJL (1998) Dual sparging laboratory-scale photobioreactor for continuous production of microalgae. J Appl Phycol 10: 377382 6. Fitz, RM, Cypionka, H (1990) Formation of thiosulfate and trithionate during sulte reduction by washed cells of Desulfovibrio desulfuricans. Arch Microbiol 154: 400406 7. Huth, J, Blasig, R, Werner, S, Muller, HG (1990) The proton extrusion of growing yeast cultures as an online parameter in fermentation processesdetermination of biomass production and substrate consumption in batch experiments with Candida maltosa Eh15 D. J Basic Microbiol 30: 481488 8. Iversen, JJL, Thomsen, JK, Cox, RP (1994) On line growth measurements in bioreactors by titrating metabolic protonexchange. Appl Microbiol Biotechnol 42: 256262 9. Kempe, LL, Gillies, RA, West, RE (1956) Acid production by homofermentative Lactobacilli at controlled pH as a tool for studying the unit process of fermentation. Appl Microbiol 4: 175 178

10. Lejeune, R, Callewaert, R, Crabbe, K, De Vuyst, L (1998) Modelling the growth and bacteriocin production by Lactobacillus amylovorus DCE 471 in batch cultivation. J Appl Microbiol 84: 159168 11. Moussard, H, LHaridon, S, Tindall, BJ, Banta, A, Schumann, P, Stackebrandt, E, Reysenbach, AL, Jeanthon, C (2004) Thermodesulfatator indicus gen. nov., sp nov., a novel thermophilic chemolithoautotrophic sulfate-reducing bacterium isolated from the Central Indian Ridge. Int J Syst Evol Microbiol 54: 227233 12. Nagpal, S, Chuichulcherm, S, Livingston, A, Peeva, L (2000) Ethanol utilization by sulfate-reducing bacteria: an experimental and modeling study. Biotechnol Bioeng 70: 533543 13. Okabe, S, Nielsen, PH, Characklis, WG (1992) Factors affecting microbial sulfate reduction by Desulfovibrio desulfuricans in continuous culturelimiting nutrients and sulde concentration. Biotechnol Bioeng 40: 725734 14. Poulsen, BR, Ruiter, G, Visser, J, Iversen, JJL (2003) Determination of rst order rate constants by natural logarithm of the slope plot exemplied by analysis of Aspergillus niger in batch culture. Biotechnol Lett 25: 565571 15. Rabus, R, Hansen,T, Widdel, F (2000) The dissimilatory sulfate and sulfur reducing bacteria. In: Dworkin, M (Ed.) The Prokaryotes: An Evolving Electronic Resource for the Microbiological Community, 3rd ed. Springer Verlag, New York. http://141.150.157.117:8080/ prokPUB/index.htm 16. Reis, MAM, Almeida, JS, Lemos, PC, Carrondo, MJT (1992) Effect of hydrogen-sulde on growth of sulfate-reducing bacteria. Biotechnol Bioeng 40: 593600 17. San, KY, Stephanopoulos, G (1984) Studies on online bioreactor identication 4. Utilization of pH measurements for product estimation. Biotechnol Bioeng 26: 12091218 18. Siano, SA (1995) On the use of the pH control reagent addition rate for fermentation monitoring. Biotechnol Bioeng 47: 651665 19. Vainshtein, MB, Matrosov, AG, Baskunov, VP, Zyakun, AM, Ivanov, MV (1980) Thiosulfate as an intermediate product of bacterial sulfate reduction. Microbiology 49: 672675 20. Van Houten, RT, Pol, LWH, Lettinga, G (1994) Biological sulfate reduction using gas-lift reactors fed with hydrogen and carbon dioxide as energy and carbon source. Biotechnol Bioeng 44: 586594 21. Vicente, A, Castrillo, JI, Teixeira, JA, Ugalde, U (1998) On-line estimation of biomass through pH control analysis in aerobic yeast fermentation systems. Biotechnol Bioeng 58: 445450 22. Walach, MR, Balyuzi, HHM, Bazin, MJ, Lee, YK, Pirt, SJ (1983) Computer control of an algal bioreactor with simulated diurnal illumination. J Chem Technol Biotechnol B 33: 5975