Phytoremediation of Contaminated Soil and Water - Chapter 16 pot

A close examination of the TLDD basin waters revealed that they are abundant in micro-phytes with chlorophyll a content up to 1 μg/ml water including coccoid and filamentous cyanobacteri

16 Microphyte-Mediated

Selenium Biogeochemistry

and its Role in In Situ

Selenium Bioremediation

Teresa W.-M Fan and Richard M Higashi

CONTENTS

Introduction

Site Description and Rationale

Approach

Isolation and Culturing of Microphytes

Measurement of Se Volatilization Kinetics

Se Analysis

Findings and Implications

Se Volatilization by Isolated Microphytes

Se Allocation in Microphyte Biomass

Nonvolatile Forms of Se in Microphyte Biomass

Conclusion

Acknowledgment

References

INTRODUCTION

Only during the last 2 decades has the complexity of the selenium (Se) biogeochem-ical cycle begun to be realized Based on the limited information available, Se biogeochemistry appears to be largely analogous to that of sulfur For example, similar to sulfur, the oxyanions of Se (selenite and selenate) are often the dominant forms in oxic aquatic environments and groundwater (Sugimura et al., 1977; Cutter and Bruland, 1984; White and Dubrovsky, 1994) Volatilization of both Se and S via biomethylation represents a major process by which these two elements enter the atmosphere (Duce et al., 1975; Craig, 1986; Cooke and Bruland, 1987) Further-more, like sulfur, Se is extensively metabolized by organisms into amino acids and antioxidant compounds (Lewis, 1976) However, many of the processes involved in the Se cycle are yet to be elucidated

This complex biogeochemical cycling of Se, together with an unusually narrow tolerance between nutritional requirement and toxicity, plus a highly heterogeneous distribution have all contributed to the difficult problem of Se pollution This is exemplified by the environmental problems associated with California’s agricultural drainage disposal systems since the Kesterson Reservoir (California) incident in the early 1980s The wildlife deformities observed there and in the present agricultural drainage evaporation basins have been attributed to Se bioaccumulation and biotrans-formation through the food chain (Skorupa and Ohlendorf, 1991; Maier and Knight, 1994) However, the actual Se biotransformations through the aquatic and terrestrial foodweb are still largely unknown The Se-related ecotoxic effect is rapidly becom-ing a serious concern for agricultural operations in seleniferous soils of the western

17 states and for industrial discharges throughout the U.S (Skorupa and Ohlendorf, 1991; Reash et al., 1996; Fairbrother et al., 1996) Long-term economic solutions

to the problem need to be explored to make these agricultural and industrial activities sustainable

Many Se removal schemes that are physically and/or chemically based have been tested but they were shown to be cost-prohibitive or ineffective for waterborne

Se concentrations of 10 μg/l or lower (Mudder, 1997) Biologically based remedi-ation schemes have also been proposed that involve precipitremedi-ation and/or volatilizremedi-ation

of selenium by soil bacteria and vascular plants (Gerhardt et al., 1991; Thompson-Eagle and Frankenberger, 1991; Terry et al., 1992) The efficacy of these bioremoval schemes has yet to be evaluated

In this chapter, we describe an approach that utilizes the biotransformation activities of aquatic microphytes intrinsic to the Se-laden agricultural evaporation basins for “natural” remediation Similar to the vascular plant-based scheme, this process is solar-driven and does not require extensive maintenance; thus the eco-nomic advantage is self-evident In addition, evaporation basins are currently employed for disposing large volumes of agricultural drainage waters, some of which have been in operation for over 2 decades Moreover, we illustrate the importance

of acquiring a molecular-level understanding of the Se fate in the environment to achieve not just Se removal but, more importantly, reduction in ecotoxic risk, which

is the real goal of remediation

SITE DESCRIPTION AND RATIONALE

The evaporation basin system of the Tulare Lake Drainage District (TLDD) located near Tulare, CA has been in continuous use for disposing large volumes of agricul-tural drainage waters (e.g., 12,000 acre-ft/yr) for more than 2 decades These basins are arranged as a sequence of shallow (e.g., 0.7 to 1.5 m) cells, where drainage waters are channeled sequentially from the first to the terminal cell to optimize water evaporation Since the drainage water itself is moderately saline (e.g., 7 ‰), salinity (primarily Na2SO4 and NaCl) in successive cells increases progressively due to evaporation (Figure 16.1A), reaching levels as high as 300 ‰ in the summer months Contaminant salts including Se oxyanions can be predicted to increase based on evaporite chemistry (Tanji, 1989), as was observed in systems such as the Peck Pond basin, located in Fresno County, CA (Figure 16.1B) However, waterborne Se

con-centrations at the TLDD basins often exhibit a decreasing trend, despite the large

buildup of salinity across the sequence of cells as illustrated in Figure 16.1A Although selenate is often the dominant form in the subsurface source waters, a significant fraction (up to 60%) of the waterborne Se has been reported to be in the selenite form during residence at these basins (Tanji and Gao, personal communi-cation; Fan and Higashi, unpublished data; Maier, 1997)

Since the disappearance of waterborne Se in the presence of 106-fold higher sulfate concentration cannot be readily accounted for by abiotic physical and chem-ical mechanisms, we focused our attention on the biologchem-ical mechanism(s) A close examination of the TLDD basin waters revealed that they are abundant in

micro-phytes (with chlorophyll a content up to 1 μg/ml water) including coccoid and

filamentous cyanobacteria, diatoms, and green phytoplankton, while vascular plants

FIGURES 16.1A and B Relationship between waterborne Se concentration and salinity at

two multicell evaporation basins currently used for agricultural drainage disposal Waterborne

Se concentration ([Se]) is often observed to increase progressively with increasing salinity through a squence of evaporation cells as in the Peck basin system, CA (Figure 16.1B)

(Modified from Tanji, K.K Toxic Substances in Agricultural Water Supply and Drainage,

1989, 109-121.) This behavior is consistent with that predicted from evaporite chemistry as shown However, an opposite trend was observed for the TLDD basins, CA (Figure 16.1A), where waterborne [Se] decreased with increasing salinity from inlet to the terminal cell (HEB A4) This trend is persistent year-round (Fan et al., 1998b) Cell waters were collected monthly

in acid-washed polyethylene bottles by TLDD staff at the designated location of each cell After removing particulate matters by centrifugation, waters were acidified to <pH 2 with HCl and analyzed for total Se via microdigestion and fluorescence measurement

and aquatic macrophytes are absent Except for the most saline cell (i.e., terminal cell), microphytes persist in these waters year-round However, there has been little information regarding the Se biogeochemistry associated with these microphytes,

or even microphytes in general Our recent investigation of a green coccoid

phy-toplankton (Chlorella sp.) isolated from a similar system (Pryse Pond, Tulare County,

CA) demonstrated that aquatic microphytes are active in transforming Se oxyanions into organic Se metabolites including volatile alkylselenides (Fan et al., 1997a) Further studies on microphyte-mediated biogeochemistry should help reveal the Se dissipation mechanism(s) against the strong salinity gradient at TLDD basins More importantly, by tracing the environmental fate of the biotransformed prod-ucts, particularly through the foodweb, the ecotoxic consequence of Se contamina-tion may be better understood This knowledge should also lead to a general under-standing of the role(s) of aquatic microphytes in the biogeochemical cycling of Se

or other elements in surface waters These organisms have been postulated to be a major driver of Se biogeochemistry (Cutter and Bruland, 1984; Cooke and Bruland, 1987), as they are in the case of the sulfur cycling (e.g., Andreae, 1986; Bates et al., 1987) Recent findings in the marine environment corroborate this hypothesis (Amouroux and Donard, 1996)

Because it is unlikely that the natural phenomenon occurring at TLDD basins

is fortuitously optimal in Se removal, understanding the removal mechanism(s) may help develop a more efficient scheme for reducing the load of Se or other contam-inants in these and other basins At the same time, by integrating the ecotoxic

FIGURE 16.1B (Continued)

considerations into the removal scheme, we hope to achieve the true goal of in situ

bioremediation — wildlife protection

APPROACH

We have taken a parallel laboratory and field approach to investigate the Se biotrans-formation activities in microphytes occurring at the TLDD basins Here, we will describe mainly the laboratory studies which involved fractionating the commonly occurring microphyte species from the basin waters, establishing monocultures of these species, measuring their Se volatilization kinetics, and characterizing the biotransformation products of Se

ISOLATION AND CULTURING OF MICROPHYTES

Microphyte species were isolated from the basin waters according to the procedure described by Fan et al (1997a) Briefly, 10 to 100 μl of water was streaked onto a 1% agarose plate prepared in f/2 seawater medium and incubated at 20 to 22°C under a light/dark cycle of 16/8 h Distinct and isolated colonies were then inoculated aseptically into the f/2 seawater medium and incubated similarly as above to establish stock cultures The plating procedure was repeated using the stock culture to min-imize bacterial contamination The stock culture was maintained by periodic inoc-ulation into fresh medium Under careful microscopic examination, only a single species of microphyte was visible in each isolated culture with no apparent bacterial

contamination One phytoplankton species (Chlorella sp.) was isolated from the

Pryse evaporation basin water and two cyanophyte species were isolated from the TLDD basin water One TLDD basin species was a filamentous cyanophyte possibly

of the LPP (Lyngbya, Phormidium, and Plectonema) group (Rippka et al., 1979; T.

Hanson, personal communication) while the other was tentatively assigned to be a

Synechocystis sp We are conducting further species identification of the microphytes

based on the 16S ribosomal RNA method

MEASUREMENT OF SE VOLATILIZATION KINETICS

The rate of Se volatilization from growing microphyte cultures supplemented with selenite or selenate was measured as described previously (Fan et al., 1997a) Briefly, 0.8 liter of f/2 seawater medium was supplemented with 10 μg/l to 100 mg/l Se and inoculated with microphyte stocks The culture was then constantly bubbled with 0.22 μm filtered air at 30°C under continuous fluorescent light with a light intensity

of approximately 360 cd The air aeration provided the CO2 needed for microphyte growth, and purged the volatile Se compounds (e.g., alkylselenides) out of the

medium into an alkaline peroxide trap containing 50 mM NaOH plus 30% H2O2, 4/1 (v/v) for 18 to 22 h The volatile Se compounds were also trapped in their original forms into a 1/4 in Teflon tube kept at liquid nitrogen temperature Total

Se in the alkaline peroxide trap was measured by GC-ECD or fluorescence while the liquid nitrogen-trapped Se forms were analyzed by GC-MS (see below) The rate of Se volatilization was calculated from the total Se and the duration of the

trapping Suspended microphyte cell density was monitored by taking visible spectra from 350 to 800 nm, from which the optical density at 680 nm was obtained

SE ANALYSIS

Medium water, the trap, and biomass samples were digested using a microdigestion method described previously (Fan et al., 1997a), which substantially simplified the procedure and reduced sample requirement compared with reported macrodigestion methods (e.g., McCarthy et al., 1981) Total Se in the digest was determined by using the GC-ECD method (Fan et al., 1997a) or by fluorescence The fluorescence method was modified from the Analytical Methods Committee (1979; Fan et al., 1998b) Briefly, microdigestion with either nitric acid or alkaline peroxide converted various Se forms into selenate which were then reduced to selenite by 6 N HCl at

105°C, followed by derivatization with 4-nitrophenylene-o-diamine or

2,3-diami-nonaphthalene to form the corresponding piazselenol derivatives, which were quan-tified by GC-ECD or fluorescence, respectively The detection limit for the GC-ECD method was 1 μg/l while that for the fluorescence method was in the sub-μg/l range, both based on a 250 μl of water sample The fluorescence method is more convenient

to perform with a broader linear range of concentrations, while the GC-ECD method should be less subject to matrix interference A typical correlation coefficient obtained for the fluorescence-based standard curve of 12 Se concentrations ranging from 0-250 μg/l was better than 0.99 Known addition method was used to check interference from sample matrix, which was negligible in all cases

The biotransformed products including volatile alkylselenides, selenonium com-pounds, and selenoamino acids were analyzed using a combination of GC-MS and NMR techniques (Fan et al., 1997a and 1998b) Alkylselenides were trapped at liquid nitrogen temperature and analyzed directly on an open-tubular DB-1 column

in a Varian 3400 gas chromatograph interfaced to a Finnegan ITD 806 mass spec-trometer Selenonium compounds and selenoamino acids were extracted from micro-phyte biomass using 5% perchloric acid (PCA) or 10% trichloroacetic acid (TCA)

Selenoamino acids in the extract were then silylated with MTBSTFA (N-methyl-N-[tert-butyldimethylsilyl]trifluoroacetamide) before GC-MS analysis, or removed of

paramagnetic ions by passing through Chelex-100 resin (BioRad) column before NMR analysis The combined MTBSTFA derivatization and GC-MS analysis enabled a simultaneous determination of selenomethionine, selenocysteine, and methylselenocysteine at trace levels plus a wide range of other metabolites with structure confirmation (Fan et al., 1998a) Selenonium compounds in the biomass

or extracts were analyzed indirectly by GC-MS for dimethylselenide (DMSe) which was liberated by 5 M NaOH treatment at room temperature or at 105 to 110°C (Fan

et al., 1997a) The precursor to DMDSe (e.g., methylselenocysteine) was similarly determined by alkaline treatement at 105 to 110°C, followed by GC-MS analysis for DMDSe (Fan et al., 1998a) Selenonium compounds such as dimethylselenonium propionate (DMSeP) and methylselenomethionine (CH3-Se-Met) may be the bio-logical precursors of DMSe, while selenoamino acids such as Se-Met are considered

to be the toxicity surrogate to wildlife (e.g., Heinz et al., 1996)

Protein-bound Se-Met was analyzed using the following procedure The protein fraction was first obtained by extracting the biomass with a Tris-SDS buffer, followed

by dialyzing the extract against a 3.5-kDa molecular weight cutoff membrane (MWCO) to remove low molecular mass components (Fan et al., 1997b) The resulting protein-rich fraction was digested in 6 N HCl, followed by MTBSTFA derivatization and GC-MS analysis (Fan et al., 1998b) The HCl digestion allowed

a full recovery of Se-Met but was unsuited for recovering Se-Cys and selenocystine (Fan et al., 1998a) The GC-MS method provided trace-level analysis for selenoam-ino acids with rigorous structure confirmation, while circumventing the use of radiotracer 75Se (e.g., Wrench, 1978) The latter coupled with liquid or thin-layer chromatography is also suited for trace analysis but impractical for the analysis of samples collected from the field Without 75Se radiotracers, conventional chromato-graphic methods do not allow sensitive detection and fall short of structure confir-mation (Bottino et al., 1984) The ability to analyze various food chain components for protein-bound Se-Met is needed since this Se form may be the vehicle through which Se is biotransferred and bioconcentrated via the food chain, thereby causing ecotoxic effects at the higher trophic levels

FINDINGS AND IMPLICATIONS

SE VOLATILIZATION BY ISOLATED MICROPHYTES

All three microphyte species could volatilize Se from selenite-supplemented f/2 seawater media, as shown in Figure 16.2 The Chlorella culture also volatilized Se

from selenate-supplemented medium, albeit at a lower rate (Fan et al., 1997a) The

Se volatilization activity of the two cyanophytes in selenate media has not been tested The peak rate of volatilization from selenite for the filamentous cyanophyte

was at least fourfold higher than that for the Chlorella sp at 1 mg/l Se supplement.

Assuming that the volatilization rate is proportional to the Se concentration of the

medium, this rate may be comparable to that for the Synechocystis sp.

In addition, the rate of Se volatilization by the three microphytes was dependent

on the suspended cell density (optical density at 680 nm), although the growth-dependent time course of the volatilization rate differed among the three species

The rate of Se volatilization by the Chlorella sp tracked the suspended cell density

and chlorophyll content, while that by the filamentous cyanophyte exhibited a sharp peak coincident with the suspended cell and chlorophyll density, followed by another more sustained peak during cell senescence (Figure 16.2 and data not shown) The

Se volatilization kinetics by the Synechocystis sp also showed two peaks, with one

during the exponential growth and the other during early senescing period In all cases, the cells aggregated and precipitated to the bottom while the optical density

of the culture dropped to near zero during senescence (Figure 16.2) There was no increase at any point during senescence in the visible spectrum from 350 to 800 nm, indicating the absence of heterotrophic bacterial growth in suspension

The close association of Se volatilization kinetics with the exponential growth

of microphytes in all three cultures indicates that this activity was intrinsic to the

FIGURE 16.2 Growth-dependent time courses of Se volatilization by

three selenite-treated microphyte species isolated from agricultural drainage

waters The Chlorella species was isolated from the Pryse basin system while the two cyanophytes (filamentous and Synechocystis) were isolated

from the TLDD basins (Fan and Higashi, 1998, and adapted from Fan et al.,

Environ Sci Technol 31, 569-576, 1997a, and Fan and Higashi, in press).

The three microphytes were grown in f/2 seawater medium supplemented

with 1 mg/l (Chlorella and filamentous) or 0.4 mg/l (Synechocystis) Se (as

Na2SeO3) Volatile Se was air-purged from the media and trapped in an alkaline peroxide solution while the optical density (OD) at 680 nm of the media was monitored for suspended cell density Total Se in the trap was analyzed by GC-ECD or fluorescence and the rate of Se volatilization was calculated from total Se trapped and the duration of the trap.

Copyright © 2000 by Taylor & Francis

microphytes Because the microphyte cultures investigated have not been proven to

be axenic, the possibility of some contribution of heterotrophic bacterial activity to

Se volatilization during the senescing period cannot be eliminated However, this contribution is unlikely to be major based on the lack of suspended growth of heterotrophic bacteria and the predominance of senescent microphyte cells in the aggregate biomass The more sustained peak of Se volatilization by the two cyano-phyte cultures during the senescing period could result from a switch of Se metab-olism in the cyanophytes and/or interaction with closely associated heterotrophic bacteria which could not be differentiated from microscopic examination

The relationship between Se depletion from the medium and rates of Se vola-tilization as a function of selenite concentrations by the filamentous cyanophyte culture was also investigated (Fan et al., 1998b) From treatments of 20 μg/l to 1 mg/l Se, the rate of Se volatilization followed a qualitatively similar time course as that in Figure 16.2 and was roughly proportional to the treatment concentrations (Fan et al., 1998b) Meanwhile, Se depletion from the medium exhibited an opposite time course (e.g., Figure 16.3) Even at treatment of 10 μg/l Se, Se was volatilized from the medium by the filamentous cyanophyte culture as the medium Se concen-tration decreased with time to 2 μg/l (Figure 16.3) The volatilization process accounted for a major fraction of the Se depletion from the medium (up to 77%), although Se incorporation into the biomass was also significant, as described below These results indicate that both processes are capable of contributing to the decrease

in waterborne Se concentrations of the TLDD basins (cf Figure 16.1)

Moreover, the chemical form(s) of the volatilized Se trapped at liquid nitrogen temperature were examined using GC-MS Figure 16.4 illustrates the total ion

FIGURE 16.3 Growth-dependent time courses of Se volatilization and depletion from

medium by selenite-treated filamentous cyanophytes The filamentous cyanophyte was grown

in f/2 seawater medium supplemented with 10 μg/l Se (as Na2SeO3) (Reprinted with permis-sion from Fan et al., 1998a.) Total Se volatilized and suspended cell density were measured

as in Figure 16.2 Medium [Se] was analyzed by the fluorescence method (Fan et al., 1998b)

chromatograms of volatile compounds released by a Chlorella culture grown at 100

mg/l Se (Figure 16.4A) and a filamentous cyanophyte culture grown at 10 mg/l Se (Figure 16.4B) For both microphytes, three volatile Se compounds were identified: DMSe, DMDSe (dimethyldiselenide), and DMSeS (dimethylselenenyl sulfide; Fan

et al., 1998b) Dimethylsulfide (DMS) and dimethyldisulfide (DMDS) were also

liberated from the Chlorella (Fan et al., 1997a) and filamentous cyanophyte cultures

(Fan et al., 1998b), respectively However, at lower Se treatment concentrations (e.g., ð1 mg/l), the only detectable product from both species was DMSe (data not shown) It is likely that the production of DMSe and DMDSe shares a similar mechanism as that of DMS and DMDS, respectively As for the release of DMSeS, the mechanism may involve a cross-reaction between the production of DMDS and DMDSe (see below)

SE ALLOCATION INMICROPHYTE BIOMASS

To examine Se allocation into the microphyte biomass, we developed a fractionation scheme that involved extraction of the biomass with 10% TCA (Fan et al., 1997a) and with Tris-SDS buffer plus dialysis The TCA-soluble fraction generally includes water-soluble, small molecular mass metabolites including small peptides, while the residue fraction is expected to be comprised of proteins, lipids, other macromolecules such as cell wall constituents, and Se0 The combination of Tris-SDS buffer extraction and dialysis against a 3.5-kDa MWCO membrane should yield a protein-enriched fraction The Se distribution in these four fractions and the whole biomass of

Chlorella and filamentous cyanophyte is summarized in Figure 16.5 Clearly, a major

fraction (60%) of the Se in Chlorella biomass resided in the TCA residue when cells

were grown in the 100 mg/l Se (as selenite) medium, while the TCA-soluble fraction accounted for 15% (Figure 16.5A) It should be noted that the TCA residue may contain Se0, since a red amorphous material codeposited with the biomass and was not extracted by TCA (Fan et al., 1997a) The Tris-SDS extract constituted only 6%, while the protein-rich fraction (3.5-kDa retentate) contributed even less (1%) to total

Se in biomass (Figure 16.5A)

In contrast, the Tris-SDS extract contained the majority of the Se (89%) in the filamentous cyanophyte biomass when grown at 10 mg/l Se (Figure 16.5B; Fan et al., 1998b) Selenium present in the protein-rich fraction (3.5-kDa retentate) also contributed to a significant fraction (17%) of the biomass Se Moreover, very little

Se was extracted with TCA (0.17%), while 52% of the biomass Se remained in the TCA residue The TCA extract and residue combination did not account fully for the biomass Se It is possible that the missing Se was a result of degradation of labile Se compounds during the TCA extraction

NONVOLATILE FORMS OF SE IN MICROPHYTE BIOMASS

The fractionation of Se into TCA and protein extracts allowed further analysis of the chemical form(s) of Se present in these extracts Our present focus is on sele-nonium metabolites and selenoamino acids, since the former may be indicative of