Introduction

In 1960, authorization was given to construct a 188 mile-long concrete drain from the western San Joaquin Valley of California to the San Francisco Bay area. The expressed use for this drain would be for the removal of saline agricultural drainage from producers in the western half of the San Joaquin Valley. Due to budget constraints and political/planning problems, only 84 miles of the drainage system was completed. At the northern end of the San Luis Drain was the Kesterson Reservoir, which was intended as a flood control reservoir in the middle of the drainage system (Quinn et al., 1995). When the projected was halted, the Kesterson Reservoir became a drainage disposal facility, consisting of 12 separate ponds. Without drainage, water in the reservoir was left to evaporate or percolate into the soil. With its location at the southern end of the Kesterson National Wildlife Refuge, the reservoir also became a home to migratory and resident waterfowl. In the early 1980s, reports of illness and reproductive failure among the waterfowl led to investigations which reported that the water in the Kesterson reservoir and the San Luis drain had extremely high level of selenium, predominantly in the form of the selenate anion with a concentration between 200 - 300 ppm selenate. Consequently, the drainage of agricultural wastewater to the San Luis Drain and the Kesterson Reservoir was cut off by the mid 1980s. By the end of the decade the San Luis Drain and the Kesterson Reservoir were closed (Weres et al., 1995). During the period of extensive research by many groups after the initial reporting of selenium contamination, several selenium resistant bacteria were isolated from the Kesterson Reservoir before its closure. One of these isolates found was a strain of Pseudomonas fluorescens (Burton et al., 1987).

The first work showing the reduction of selenium salts, specifically selenite, was first reported by Frederick Challenger in working with fungi. In 1945, Challenger proposed a mechanism for the methylation and reduction of selenite to dimethyl selenide CH₃SeCH₃. This mechanism was suggested based upon prior work with the methylation of arsenic. Much later, reduction and methylation of organoselenium in an aquatic environment was established in the 1970s (Chau et al., 1976). A different reduction/methylation mechanism was proposed by Doran in 1982. Doran's mechanism not only accounts for the production of dimethyl selenide, but dimethyl diselenide as well. In essence, he proposed a direct path from selenite to elemental selenium, and a two step mechanism from elemental selenium to dimethyl diselenide for the methylation of selenite (Doran, 1982). The work by Burton et al., yielded a facultative anaerobe, (Pseudomonas fluorescens K27 isolated by Ray Fall at University of Colorado, Boulder) which exhibited the production of dimethyl selenenyl sulfide and dimethyl diselenide from the static headspace of test tube cultures amended with selenate or selenite salts in complex growth media. These organoselenides were detected via sulfur chemiluminescence detection equipped capillary gas chromatography (Chasteen et al. 1990; Zhang and Chasteen, 1994; Chasteen, 1998).

Our earlier work with this microbe using replicate test tube experiments has been carried to a larger volume system that we report here. A bioreactor allows for fine control of culture growth conditions. Control of culture temperature, dissolved oxygen, pH, nutrient inflow and outflow, and agitation (stirring) can all be maintained via probe feedback and computer processing. Liquid and headspace samples can also be taken while still maintaining culture isolation (see companion paper by Akpolkat and Chasteen in this conference). The bioreactor allows for the growth of large cultures; in this work, the average culture volume was 3 liters.

Monitoring liquid bacterial cultures for pH changes over the life cycle can be insightful. It has been shown that changes in growth media pH, due to cellular wastes, can have an effect on the growth rate of bacteria. Sources of pH change can be actual waste products produced by the cells or byproducts of media or other extra cellular components reacting with extra cellular wastes. Changes in culture pH that trend toward a more acidic pH are sometimes due to production of lactic acid wastes in the fermentation cycle or other acidic byproducts. Trends towards a more basic culture pH may be due to deamination of proteins or other similar processes used in the transformation of, in our case, selenite to organoselenide. These changes in the growth media may pH cause the cell to change its intracellular pH as a buffering mechanism. However, extreme changes in intracellular pH can cause metabolic problems, which can dramatically slow or stop culture growth (Cherlet and Marc, 1998). It is feasible that this may have an affect of Pseudomonas fluorescens K27 as well. Due to the interrelated relationship of culture pH with culture growth, both of these culture aspects will be discussed.

Materials and Methods

Stock solution preparation

Stock solutions for metalloidal amendment of batch cultures were made prior to bioreactor inoculation. Two separate solutions were made, each 1 M in concentration. For the 1 M selenate solution, 42 g of sodium selenate 98% (Aldrich Chemical Company, Milwaukee, WI, USA) was dissolved in 250 mL of deionized water and sterile filtered with a 0.2 micron, 250 mL Nalgene disposable filter unit (Nalge Company Rochester, NY, USA) using a vacuum pump (Barnant Company Barrington, IL, USA) to facilitate filtration. After filtration of the solution was complete, the bottom-receiving flask was removed from the 250 mL filter unit, and a sterile cap was placed on the receiving unit for storage. For the production of 1 M selenite solution, 37.5 g of sodium selenite 99% (Aldrich Chemical Company, Milwaukee, WI, USA) was used with the above procedure.

Growth media preparation

The growth media, TSN3, used in all of the experiments for this work consists of 10% tryptic soy broth (DIFCO Laboratories, Detroit, MI USA) and 3% potassium nitrate (Fisher Scientific, Houston, TX, USA) added as electron since for these facultative anaerobes. Preparation included making the media solution (10 g tryptic soy broth, and 3 g potassium nitrate per liter water) and sterilizing in a 2540E Autoclave (TuttnauerUSA Co. LTD, NY USA), at 121^oC for 20 minutes with a secure foil cover over the mouth of the container to prevent contamination during storage (non refrigerated).

Inoculum preparation

The inoculum for the fermentor was prepared in two steps. In the first step starting two days before the fermentor experiment, an isolated culture of Pseudomonas fluorescens K27 was extracted from a streaked agar plate via sterile wire loop and placed into 40 mL of sterile TSN3 in a 125 or 250 mL Erlenmeyer flask, previously sterilized. These were placed into a waterbath shaker at 30^oC with foil cover and shaken vigorously to grow aerobically. These cultures reached stationary phase overnight. The next day, this 40 mL culture was added to sterile 250 mL TSN3, contained in a 500 mL Erlenmeyer flask and this was again vigorously shaken for aerobic growth in a water bath overnight to reach stationary phase. This provided the 10% inoculum volume for the beginning of a batch culture experiment in the fermentor.

Inoculation of fermentor

Approximately 2.7 L of TSN3 was sterilized is the bioreactor vessel and this volume inoculated with the bacterial inoculum described immediately above. To establish anaerobic growth the bioreactor was purged with sterile N₂ until the dissolved oxygen probe reached a minimum reading (5 to 20 min). For the unamended (control) cultures the initial pH was 7.40.

Liquid culture sampling

Liquid samples were taken through a liquid sample port that allows for metered control. Approximately 10 mL of liquid sample was taken hourly. Liquid medium sampling optical density measurements were determined from these same samples. An absorbance reading was taken and the sample was transferred from the test tube back to the sample vial for pH monitoring via Coring pH meter 340 (Corning, NY USA) standardized with 7 pH buffer.

Determination of specific growth rate

To determine the specific growth rate, the natural logs (ln) of the optical density readings were plotted versus time for each culture. From the plot, the data points that best represented the logarithmic phase of growth were used to make a linear least squares fit. From this equation, the slope is the calculated specific growth rate (Pirt, 1975; Paran et al., 1990). Lastly, the specific growth rates for each of the triplicate set were averaged and the standard deviation was found.
 
 

Results and Observations

Triplicate runs were performed for each type of experiment. Cultures amended with selenate and/or selenite were done with a final concentration of 10 mM based on prior work (Chasteen et al., 1990; McCarty et al. 1993). Final biomass at stationary phase was determined by previous work by Stone et al. (1998) that showed a direct correlation between optical density and biomass. Error bars in all graphs represent one standard deviation for triplicate runs.

Control cultures

As a control, batch cultures of P. fluorescens were grown unamended. The pH of the cultures showed a drop from an initial pH of 7.40 to a final pH of approximately 6.60. The specific growth rate was 0.300 hr^-1, with a final biomass of 0.35 g/mL. When large changes in pH occurred, a slowing or reversal in the rate of pH change corresponded with the achievement of culture stationary growth phase. Similar work has shown an increase in pH for control cultures grown in similar system but using 1% nitrate (Stone et al. 1998). Unlike this work, these workers noted an increase in pH and a specific growth rate of 0.266 hr^-1, approx. 12% slower than reported by this work. Results from our triplicate control cultures are shown in Figure 1. Changes in cell population are designated by the log of optical density at 526 nm.

Ten millimolar selenate amended cultures

Work from the control culture led to work with 10 mM selenate amended cultures. Selenate is the most common selenium oxyanion found in aquatic environment where K27 was isolated (reference tom), and was therefore our first oxyanion of interest. Unlike with the control cultures, the 10 mM selenate amended cultures showed almost no increase in pH over the approximate 10 hours of those experiments but the hourly samples pH did exhibit large variation. The average initial pH was 7.46 after SeO₄^2- was added as the sodium salt. This reading was made within 10 minutes of preculture inoculation and selenite amendment to the bioreactor. Over the next 10 hours, the average culture pH (among three runs) increased to 7.49. The specific growth rate and final biomass at stationary phase were also effected by amendment. As compared to the control, the specific growth rate declined to 0.250 hr^-1. Also, the 10 mM selenate amended cultures also showed the lowest final biomass at stationary phase at 0.30 g/mL. Results from the triplcate 10 mM selenate amended cultures are shown in Figure 2.

Ten millimolar selenite amended cultures

Selenite is also encountered in the aquatic environments, whether environmental or industrial in nature. Ten millimolar selenite was selected as an amendment concentration for consistency. The initial pH of the 10 mM selenite amended cultures was 7.86, after Se amendment. From this initial pH there was a variable increase to 8.02 at stationary phase. Variations in pH change occurred at the end of the experimental time course, with very similar readings at the beginning. The specific growth rate was 0.190 hr^-1. Although a lower specific growth rate as compared to the 10 mM selenate amended culture, the final biomass at stationary phase was the highest at 0.37 g/mL. Results from the triplicate 10 mM selenite amended cultures are shown in Figure 3.

Ten millimolar 1:1 selenate to selenite mixed amendment cultures

In most selenium oxyanion bioremediation situations, selenate and selenite are both present in the water being treated. Therefore, it was decided to explore the affect of a mix of selenate and selenite on the culture growth and pH. A total of 10 mM Se was add to freshly inoculated cultures as 0.5 mM selenate + 0.5 mM selenite. The change in pH for the 1:1 selenate to selenite cultures was similar to the 10 mM selenite culture in the fact that a increase in pH is shown with little variation throughout the experimental time course. Initial pH begins at 7.35, after Se amendment, and climbed to an average pH of 7.6 at stationary phase. The affect of the amendment on specific growth rate was much lower than expected. The working hypothesis before experimentation was that the specific growth rate would be an average of the 10 mM selenate and 10 mM selenite cultures, which would be approximately 0.220 hr^-1. However, the reported specific growth rate was 0.12 hr^-1, slightly less than half of the expected specific growth rate and the second lowest growth rate reported. That said, the final biomass at stationary phase was 0.34 g/mL, just below the concentration of the control. The variation in pH change among the three runs for these mixed experiments was also the lowest reported. The unexpectedly low specific growth rate may suggest some type of synergistic affect of the presence of both selenate and selenite. This may also may have an effect on the pH readings as well, as seen in the initial pH and subsequent increase. Results from the triplcate 10 mM 1:1 selenate to selenite amended cultures are shown in Figure 4.

Ten millimolar 2:1 selenate to selenite mixed cultures

Using the 1:1 selenate to selenite mix as a "mixed control," the affects of an uneven mix of selenate to selenite on pH and specific growth rate were explored to see how a larger concentration of one selenium oxyanion would affect growth rates and pH. The first uneven mixed culture was 2:1 selenate to selenite. A total of 10 mM Se was add to freshly inoculated cultures as 0.67 mM selenate + 0.33 mM selenite.

The 2:1 selenate to selenite cultures had a change in pH very similar to the control. The initial pH was 7.35, after Se amendment, and declined to 6.48, very similar to the control. This was unexpected, as the 10 mM selenate culture results suggest that the pH would either stay close to the same or exhibit a slight increase. A large variation in pH change however, occurred latter in these experiments time course, and the final biomass at stationary phase was 0.34 g/mL. However, the specific growth rate was higher than the 10 mM 1:1 selenate to selenite mixed culture at 0.160 hr^-1. This suggests that if there is a synergistic affect, the higher concentration of selenate relative to selenite could be the cause for the higher specific growth rate when comparing the growth grate of the 10 mM selenate and 10 mM selenite cultures. Results from the triplcate 10 mM 2:1 selenate to selenite amended cultures are shown in Figure 5.

Ten millimolar 1:2 selenate to selenite mixed cultures

Finally, to test the hypothesis that an uneven mix of selenate to selenite would affect specific growth rate and pH, batch cultures of 10 mM 1:2 selenate to selenite were performed. A total of 10 mM Se was add to freshly inoculated cultures as 0.33 mM selenate + 0.67 mM selenite.

The starting pH was similar to the 10 mM selenite cultures, and the difference as compared to the 10 mM 2:1 selenate to selenite cultures are dramatic. The initial pH started at 7.86, virtually identical to the initial pH of the 10 mM selenite cultures. The pH for the 10 mM 1:2 selenate to selenite cultures declines to 7.5, a final pH very similar to the 10 mM selenate cultures. The specific growth rate (0.078 h^-1 was the lowest of all culture types, but most importantly lower than the mixed control. This suggests that the higher concentration of selenite causes a decrease in specific growth rate as compared to the the mixed control. The large variation in pH change towards the end of experimental time course suggests a low reproducibility. The final stationary phase biomass was 0.34 g/mL.

Discussion

Control experiments involving sterile TSN3 at 30^o for 10 hours showed only very small change in pH form the initial pH of 7.4 (data not shown); therefore the components of the sterile system do not appear to be the origin on the pH changes seen in Se-amended bacterial cultures of this microbe. Selenium-free cultures of P. fluorescens moved downward in pH over the approximate 8 hours from inoculation to stationary phase (Figure 1) probably due to the production of organic acids. These facultative anaerobes used nitrate as the electron sink in denitrification and may also release acids in anaerobiosis. When 10 mM SeO₄^2- was added soon after inoculation (10% vol/vol), the pH changed very little over the next 8 hours (Figure 2); furthermore the oxyanion amendment cause little change in the initial pH of the sterilized media: pH = 7.4. In the analagous 10 mM SeO₃^2- experiments (Figure 3), the initial selenite addition cause the unbuffered growth medium to go slightly more alkaline, to an average pH immediately after Se amendment of >7.8. This is probably due to selenite anion acting as a weak base and capturing protons in the aqueous solution, slightly increasing the pH. Through this time course, the selenite-amended cultures generally drifted higher in pH; however, the small differences in the pH among these triplicate runs in the early part of the experiment increase greatly as stationary phase was approached. The mean final average pH was above 8.

 The equal mix of selenate/selenite (5mM/5mM) also showed an increase in pH but with much less variability among the triplicate runs (Figure 4). In these the first mixed oxyanion experiments, it appears that the drift in the selenite-only experiment was dampened by the presence of an equal concentration of selenate, and these results were very reproducible.

 To see whether selenite or selenate was more important to the changes in pH, 2:1 selenate/selenite (Figure 5) or 1:2 selenate\selenite (Figure 6) experiments were carried out. The pH in both these suites decreased over the triplicate time courses; however the variability among runs increased greatly clouding any clear trend. The standard deviation in the 2:1 selenite to selenate triplicates encompasses an entire pH unit in the last sampling in stationary phase.

 Although in comparison, small differences are seen in the stationary phase biomass of these cultures, the largest difference was between the straight 10 mM selenate (0.30 g/ml biomass) and 10 mM selenite (0.37 g/mL) amendments. As a measure of relative toxicity it would thus appear that selanate is more toxic to this microbe than selenite at least as far as the production of bacterial cells. This runs generally opposite to the trend in the literature which finds SeO₃^2- more toxic to test organisms than SeO₄^2- (Wilber, 1980; Ingersoll et al., 1990; Spallholz, 1994).

Instead, using specific growth rate (see Table 1 below) as a relative measure of toxicity, the 1:2 selenate/selenite amendment experiment was the most toxic, exhibiting an average specific growth rate of 0.078 h^-1 compared to 0.30 for the unamended P. fluorescens K27 control.

This also highlights the trend in decreasing specific growth rate from control (0.30 h^-1) to straight selenate (0.25 h^-1) to straight selenite (0.19 h^-1) amendments. While the overall biomass of the 1:2 selenate/selenite runs were merely average, the synergistic toxicity of these two oxyanions in solution supressed the growth of K27 to about 25% of the control (0.078 versus 0.30 h^-1). These are some of the very few data involving mixed metalloidal cultures. Most relative toxicity reports involving selenium have been with one or the other oxyanion alone. These results tend to confirm the relative toxicity of selenite as greater than selenate.

The pH trend in these experiments do not seem to correlate with either biomass production nor specific growth rate. Using the final average pH of 7.2 from the unamended controls as a benchmark, only 2:1 selenate/selenite dropped lower (6.55 average pH); however, the standard deviation among these three runs was large. That said, those cultures exhibited average biomass and specific growth rates comparable to other amendment experiments (Table 1).

Bibliography

Burton, G.A.; Giddings, T.H.; DeBrine, P.; Fall, Ray. Appl. Environ. Microbiol. 1987, 53, 185.

Challenger, F. Chem. R. 1945, 36, 315-361.

Chasteen, T.G. In "Environmental Chemistry of Selenium;" Frankenberger, Jr. W.T. and Engberg, R.A., Eds.; Marcel Dekker: New York, 1998; Chapter 29.

Chasteen, T.G.; Silver, G.M.; Birks, J.W.; and Fall, R. Chromatographia. 1990, 30, 181-185.

Chau, Y.K.; Wong, P.T.S.; Silverberg, B.A.; Luxon, P.L.; Bengert, G.A. Science 1976, 192, 1130-1131.

Cherlet, M; Marc, A. Biotechnol. Prog. 1998, 14, 526-538.

Doran, J. In " Advances in Microbial Ecology," Vol. 6; Plenum Press: New York and London, 1982; pp.14-22.

Ingersoll, C.G.; Dwyer, F.J.; May,T.W. Environ. Toxicol. Chem. 1990, 6, 1171-1181.

McCarty, S.; Chasteen, T.G.; Marshall, M.; Fall, R.; Bachofen, R. FEMS Lett. 1993, 112, 93-98.

Paran, J.H.; Sharma, S.; Qureshi, A.A. Toxicity Assessment: An Int. J. 1990, 5, 251-265.

Pirt, S. J. "Principles of Microbe and Cell Cultivation"; Halsted Press: New York, 1975; 5-35.

Quinn, N.W.T.; McGahan, J.C.; Delamore, M.L. Cal. Agr. 1996, 52 (5), 12-18.

Spallholz, J.E. Free Radical Biol. Med. 1994, 17, 45-64.

Stone, A; Van Fleet-Stalder V.; Chasteen T.G. "Following the Headspace Production of Organometalloids Produced by a Facultative Anaerobe Amended with Metalloidal Salts", 215^th National ACS Dallas, March 26-April 2, 1998, Dallas, TX; ENVR. 46.

Weres, O.; Jaouni, A.; Tsao, L. Appl. Geochem. 1989, 4, 543-563.

Wilber, C.G. Clin. Toxicol. 1980, 17, 171-230.

Zhang, L. and Chasteen, T.G. Appl. Organomet. Chem., 1994, 8(6), 501-508.

Acknowledgements

 This research was supported by a Cottrell College Science Award of Research Corporation, the Texas Regional Institute for Environmental Studies, Sam Houston State University Research Enhancement Funds, and the Robert A. Welch Foundation.