The objective of this study is to calibrate bioaccumulation of PCBs by zebra mussels with sorption by solid-phase microextraction fibers (SPMEs) in order to develop a baseline model that can be used to predict the uptake of PCBs by zebra mussels without conducting tests with live organisms. Since SPMEs mimic the bioconcentration process by simple partitioning, dietary uptake and chemical metabolism would not be accounted for.  However, since zebra mussels readily accumulate hydrophobic compounds and are relatively poor at metabolizing them, they are a good animal model for examining PCB bioconcentration.

Persistent, bioaccumulative, and toxic (PBT) chemicals, such as polychlorinated biphenyls (PCBs), are of particular concern because of their resistance to degradation and high affinity for organic phases. Such extremely persistent compounds are hydrophobic, or lipophilic in nature and can reach considerable concentrations in biota.  The hydrophobicity of PCBs causes them to partition to and bioconcentrate in lipid phases of organisms, such as cell membranes, which often results in baseline toxicity.  They can also accumulate in storage lipid and present a source of chemicals for food chain accumulation. 

            Bioaccumulation tests with live organisms are often expensive, time-consuming and highly variable.  Passive sampling devices (PSDs) such as Empore disks, semi-permeable membrane devices (SPMD), and solid-phase microextraction (SPME) fibers have been introduced as a method to detect hydrophobic compounds in the environment.  The biomimetic application of these sampling devices has recently been suggested. For chemicals that are not metabolized to any great degree, PSDs are capable of mimicking chemical accumulation in biota based upon simple physiochemical partitioning of compounds between aqueous and hydrophobic phases.  PSDs have many advantages including ease of deployment, lower variability, low production costs, and increased sample size due to ease of replication since the amount of time required to complete tests is considerably less than with live organisms.  All of these passive sampling devices have been demonstrated to be effective in environmental sampling; however, due to the small volume of SPME fibers and rapid accumulation kinetics, steady state is achieved much faster than in SPMDs or Empore disks.        

This research will investigate the capacity of SPME fibers to mimic the bioaccumulation of PCBs by zebra mussels by conducting standard bioaccumulation tests.  Uptake of PCBs by zebra mussels will be determined over a geometric time series (e.g., 0, 1, 2, 4, 8, 16 days) to establish uptake kinetics and steady-state PCB concentrations in the organisms.  Uptake kinetics by the SPME fibers will be determined prior to the bioaccumulation tests in order to assess the amount of time required to achieve steady state in the fiber so that the required exposure time for SPMEs can be determined.  SPME measurements of PCBs in the zebra mussel exposure water will be completed over the course of the bioaccumulation study to ensure that zebra mussel PCB exposure remains constant.  The  correlation between the accumulation of PCBs by zebra mussels and the partitioning of PCBs to the SPME will provide the basis for a model to predict PCB bioaccumulation from SPME uptake of PCBs. 

The Great Lakes collectively represent the largest body of fresh water on the planet.  They are an integral part of the North American heritage for a number of different reasons.  The Great Lakes provide a source of potable water, food, power, transportation, and recreational habitat for the approximately 30 million people that reside in the Great Lakes Region (U.S. EPA, 2006).  In addition, the Great Lakes are a source of habitat for numerous wildlife species. 

Toxic chemicals continue to threaten the Great Lakes ecosystem, and are harmful to both humans and wildlife.  Polychlorinated biphenyls (PCBs) represent a group of chemicals that are of great concern because they are persistent, bioaccumulative, and toxic (PBTs).  PCBs are distributed widely in the environment and in the Great Lakes Region even though production of them ceased in the United States in 1977.  PCBs were used as coolants and lubricants in transformers, capacitors, and other electrical equipment since they are good insulators and do not burn easily.  There are 209 different PCB congeners, which differ in the number and placement of chlorine atoms on the biphenyl rings.  PCBs are relatively insoluble in water and the solubility decreases with increased chlorination (ATSDR, 2006). 

The study of PBTs, such as PCBs, is prevalent in the field of environmental toxicology because of their resistance to degradation and high affinity for organic phases.  These characteristics allow PBTs to remain in the environment long after the emission of such chemicals has ceased.  Some PBTs can also be subject to long-range transport, which means they can migrate to areas that are not related to their production and usage (Mayer et al., 2000).  Such extremely persistent compounds are hydrophobic, or lipophilic, in nature and can reach considerable concentrations in biota.  The hydrophobicity of PBTs causes them to partition to and bioaccumulate in lipid phases of organisms, such as cell membranes, which often results in baseline toxicity.  They can also accumulate in storage lipids and present a source of chemicals for food chain accumulation. Thus, a major focus in environmental toxicology is to study the accumulation of such contaminants in organisms. 

The Great Lakes have also fairly recently been impacted by the introduction of zebra mussels (Dreissena polymorpha).  Zebra mussels are extremely invasive and have a very high reproductive rate. On average, a single female produces 40,000 eggs per reproductive period (USGS, 2006).  In addition, zebra mussels rapidly disperse to new areas in aquatic systems in two ways: (1) passive drifting of the free-floating larval stage (or “pelagic” veliger), and (2) proteinaceous byssal threads are extended allow them to attach to a variety of surfaces (e.g., rocks, other mollusks) and be transported.  Zebra mussels are extremely efficient filter feeders and are capable of removing particles ranging from 0.7-750 um in size from the water column (Morrison et al., 1998).  Due to the high density and efficient filtering capabilities of zebra mussels, they have an immense potential to alter water quality through clarification and removal of nutrients as well as accumulation and redeposition of pollutants.  Zebra mussels have a relatively high lipid content, and do not readily metabolize chemical compounds.  Therefore, they represent a good study organism for assessing the uptake of hydrophobic organic contaminants, such as PCBs, in the Great Lakes.

Estimating the bioaccumulation potential of chemicals typically involves exposing live organisms to known concentrations of the test chemical in water or sediment and measuring the uptake of the chemical over time.  Typically, a bioaccumulation factor (BAF) or bioconcentration factor (BCF) is measured at steady state or is estimated using a one-compartment first order kinetics (1CFOK) model (Mayer et al., 2003). There are various disadvantages to determining BAFs from bioaccumulation tests with live organisms.  These tests are often expensive, time-consuming, and highly variable due to differences in species and chemical metabolism.

Passive sampling devices (PSDs) such as Empore disks, semi-permeable membrane devices (SPMD), and solid-phase microextraction (SPME) fibers have been introduced as an alternative method to detect hydrophobic compounds in the environment (Górecki & Namiesnik, 2002; Leslie et al., 2002a).  PSDs typically consist of a hydrophobic phase that can be deployed in a medium and act as a sink to which hydrophobic molecules partition.  For chemicals that are not metabolized to any great degree, PSDs are capable of mimicking chemical accumulation in biota based upon physiochemical partitioning of compounds between aqueous and hydrophobic phases.  This “biomimetic” application of these sampling devices has recently been suggested (Verbruggen et al. 2000).  PSDs have many advantages including ease of deployment, lower variability, low production costs, and increased sample size due to ease of replication since the amount of time required to complete tests is considerably less than with live organisms. 

The uptake of hydrophobic chemicals by PSDs is based on an equilibrium-partitioning model.  According to this model, hydrophobic compounds are distributed between lipids of organisms, pore-water, and organic carbon; and partitioning between these three compartments is at equilibrium (Kraaij et al. 2003). The key to equilibrium partitioning sampling is the ability to control or reduce sampling time to hours or days, which would allow for ease of replication and data collection.  Typically, the higher the surface-to-volume ratio, the faster the sampling device reaches equilibrium with respect to chemical partitioning. 

As mentioned previously, there are various types of PSDs that are based upon equilibrium partitioning.  SPMDs are composed of low-density polyethylene layflat tubing containing a known weight of neutral lipid.  SPMDs have been shown to effectively mimic the uptake of environmental contaminants in biological systems and successfully extract hydrophobic chemicals, such as PCBs, from water, sediment, and soil (Huckins et al. 1990; Booij et al. 1998; Lanno & Wells, 2001).   SPMEs were developed by Pawliszyn et al. (1990) as a simple extraction method with several advantages over other PSDs.  The SPME device consists of a fiber coated with a thin polymer phase, such as poly(dimethylsiloxane) (PDMS).  SPMEs can be used for sampling hydrophobic contaminants and can be subsequently introduced into a gas chromatograph system via thermal desorption.  The polarity and thickness of the polymer coating can be varied to optimize the uptake of contaminants with varying polarities. 

The introduction of PSDs provides a mechanism for directly comparing the uptake of chemicals by organisms, but with the advantage of a more rapid response time.  This type of environmental monitoring is not necessarily a new idea; however, the application of recent SPME technology has enormous beneficial implications that need to be further investigated.  SPMDs provided a brilliant tool for detecting the presence of hydrophobic contaminants in environmental media.  Nonetheless, the sheer mass of the membrane provides an essentially infinite sink for the chemical partitioning of PBTs.  According to Mayer et al. (2003), a comparison of literature uptake rate constants shows that SPMDs can require weeks, months, or years to reach equilibrium, whereas SPMEs can reach equilibrium in a time span of hours or days in water samples with hydrophobic substances such as PCBs.  Due to their size and duration of exposure, SPMDs are also subject to biofouling by colonizing organisms (i.e., diatoms), which can alter the membrane characteristics and hinder chemical uptake. 

One advantage of SPMEs over other PSDs (i.e., SPMDs) is that because only a very small amount of chemical is extracted from solution, the deployment of the fiber does not influence the existing equilibrium between the bound and free form of a chemical, and the fiber only measures the dissolved, or potentially bioavailable, concentration (Van der Wal et al., 2004).  When SPME fibers are placed in test water, the hydrophobic compounds diffuse from the aqueous phase and are absorbed into the polymer coating.  This passive process is driven by the differences in the fugacity of the chemical in the aqueous (freely dissolved) and hydrophobic phases (Leslie et al., 2004).  As mentioned, a second advantage is that the time required for SPMEs to reach equilibrium varies depending upon the coating thickness, exposure conditions, and hydrophobicity of the target compound; however, the equilibration time is much faster than other PSDs.  The rapid equilibration of the fiber offers advantages over traditional chemical measures to assess bioaccumulation potential, such as K_ow.  Since accumulation of hydrophobic chemicals by aquatic organisms (i.e., zebra mussels) and SPMEs both follow first order kinetics, both the fiber and the organism will achieve a steady state concentration that facilitates a correlation between the two uptake values (Wells & Lanno, 2001).  In this manner, the bioaccumulation potential of contaminants can be predicted from the equilibrium concentration of the chemical in the fiber, with the fiber acting as a biomimetic sampling device.

Thus, PSDs have been demonstrated to be effective in environmental sampling; however, due to the small volume of SPME fibers and rapid accumulation kinetics, steady state is achieved much faster than in SPMDs or Empore disks (Leslie et al., 2002a), which makes SPMEs the most appropriate sampling technique.  The effectiveness of SPMEs as biomimetic sampling devices will be investigated further in this proposed study. 

This research will investigate the capacity of SPME fibers to mimic the bioaccumulation of PCBs by zebra mussels by conducting standard bioaccumulation tests. Uptake of PCBs by zebra mussels will be determined over a geometric time series (e.g., 1, 2, 4, 8, 16 hours) to establish uptake kinetics and steady-state PCB concentrations in the organisms. The correlation between PCB accumulation by zebra mussels and PCB partitioning to the fiber will provide the basis for a model to predict PCB bioaccumulation from SPME uptake of PCBs.

Zebra Mussel Collection and Culture Maintenance

Adult zebra mussels (20-25 mm shell length) will be collected from relatively uncontamintated nearshore sites in Lake Erie, Ohio. The exact locations of field collection sites are yet to be determined; however, they will likely be related to Stone Laboratory or at locations previously sampled and located via GPS from studies conducted in Dr. Susan Fisher’s laboratory.Mussels will be placed in sealable plastic bags with a wet paper towel on ice during transport to the laboratory.

Zebra mussels will be maintained in the laboratory according to standard operating procedures in observance of federal invasive species handling protocols. Mussels will be maintained in aquaria with aged tap water following chlorine dissipation. Similar to the protocol used by Fisher & Bernard (1991), mussel cultures will be fed a mixture of algal species every other day. The water in the aquaria will be maintained between 18-22°C, and water quality parameters such as pH, ammonia, dissolved oxygen and temperature will be measured daily.

Methods of Laboratory Test

Uptake kinetics will be determined in the organisms to determine steady state chemical values or a sufficient number of data points will be generated to estimate organism steady-state concentrations using a 1CFOK model. Uptake kinetics by the SPME fibers will also be determined prior to the bioaccumulation tests in order to assess the amount of time required to reach equilibrium with respect to the partitioning of each of the selected PCB congeners to the SPME fiber and to validate that sampling in the test system is non-depletive. A detailed treatment of these methods is presented in Mayer et al. (2003).PCBs 99 and 169 (IUPAC) will be used for this experiment, which are pentachlorinated and hexachlorinated, respectively.

Bioassays will comprise a set of exposure chambers containing two to three organisms in each chamber.Approximately 24 hours before the bioassay, zebra mussels will be removed from the culture aquaria by cutting their byssal threads with a razor blade.Mussels will be placed in glass Petri dishes and allowed to reattach themselves over a 24-hour period.Organisms will not be fed during this 24-hour period or for the duration of the test (i.e., 16 hours).Mussels that reattach to the Petri dish will be considered healthy and will be used in the bioaccumulation test; organisms that do not reattach will be discarded. All organisms attached to glass Petri dishes will be placed in individual chambers and sampled at specific sampling times, with duplicate chambers sampled at each time. Simultaneous measurements of chemical concentrations in the water will be taken with SPMEs. The goal is to establish correlations between steady-state organism chemical concentrations and steady state chemical concentrations absorbed by the SPME. SPME chemical analysis and analysis of organism body residues will be conducted at each sampling time. The lipid content of organisms will be determined colorimetrically by the sulphophosphovanilline method of Van Handel (1985).

Methods of Analysis

Chemical Analysis

Zebra mussels will be shucked and soft tissues homogenized in a minimum volume of water. Ten g of homogenate will be ground with 40 g anhydrous sodium sulfate and the mixture will be poured into a 2.5 cm x 60 cm glass column, plugged with glass wool, and filled with 70 ml of 1:1 dichloromethane:hexane.Samples will be allowed to sit overnight, eluted with 210 ml of 1:1 dichloromethane:hexane, and evaporated to ~2 ml by a rotary evaporator.The extract will then be further eluted through a Florisil column (6 g: 60-100 mesh), activated and cleaned by drying overnight at 130ºC and running 50 ml hexane through the column.Target compounds will be eluted with 50 ml hexane, the eluate will be evaporated, and then re-suspended with 50 ml hexane on an activated silica-gel column (70-230 mesh) for a second clean-up step. The sample will be eluted with 50 ml hexane and evaporated to dryness again.At the end of clean-up, samples will be resuspended in 2 ml of iso-octane for injection into the gas chromatograph with an electron capture detector (GC-ECD).

PCB congeners will be identified and quantified using a Hewlett-Packard 5890 Series II GC-ECD with a splitless injection port.The analytical column is a DB-5 (J&W Scientific, Folson, CA; 60 m x 0.25 mm, 0.25 um film).The injection port will be set to 250ºC, and the detector to 325^oC.The beginning oven temperature will be set at 100^oC and increased (1^o C/min) to 265^oC, followed by 20^oC/min to 300^oC.Constant head pressure in the column will be set at 65 psi.The carrier gas and make-up gas will be hydrogen and nitrogen, respectively, and each chromatographic peak will be quantified by the internal standard method.

Quality assurance will be achieved by sample replication and reagent blanks in each sample batch.Reagent blanks will be prepared and analyzed identically to biotic samples.To evaluate the detection limit of the methods, method detection limit (MDL) values will be calculated using the two appropriate congener reference standards, and replicated seven times in clean tissue.Values calculated by the MDL-test will be used to separate background and the response of target compounds.

Data Analysis

Uptake kinetics of the PCB congeners in zebra mussels and SPMEs will be examined using a 1CFOK model to estimate steady-state PCB concentrations.The correlation between the steady-state PCB concentration in mussels and SPMEs will then be examined using appropriate statistical models.

As stated, the major objective of this study is to establish a quantitative model linking steady state levels of PCBs in SPME fibers with steady state levels of PCBs in zebra mussels. Calibration and validation of this model would provide a greater understanding of chemical and/or biological interactions and enhance the ability to predict the bioaccumulation of hydrophobic contaminants.  This model could further be applied to assist in the efforts to assess and remediate the Great Lakes. 

This predictive biomimetic model will provide a simple means for estimating the bioaccumulation potential of chemicals without the time and expense of conducting bioaccumulation tests with live organisms.  Once proof of concept for this type of model is established for PCBs and zebra mussels, the potential exists for the extension of this approach to other PBT chemicals and other organisms.  Furthermore, once calibrated with other organisms, the potential exists for SPME fiber measurements to be used as a rapid screening tool to assess the potential for trophic transfer of PBTs in Great Lakes food webs associated with zebra mussels.