2
726
Environ. Toxicol. Chem. 20, 2001
G.M. Kle cˇ ka et al.
MATERIALS AND METHODS
Subsamples of the composite sediment and water samples
were submitted for characterization to Midwest Laboratories
Test substance
(
Omaha, NE, USA). Sediment samples were analyzed for tex-
The BPA was obtained from Research Triangle Institute
Research Triangle Park, NC, USA). The sample was identified
ture, organic carbon content, inorganic content, cation ex-
change capacity, and total heterotrophic bacteria. River water
samples were analyzed for conductivity, alkalinity, anions, cat-
ions, total dissolved solids, and total heterotrophic bacteria.
(
by lot number B0070138. The purity was reported to be 99.1%,
and the supplier confirmed the identity by infrared spectros-
1
copy and nuclear magnetic resonance spectroscopy ( H-NMR).
1
4
Experimental approach
Radiolabeled [ C]BPA (uniformly ring labeled) was ob-
tained from Wizard Laboratories (West Sacramento, CA,
USA). The sample was identified by lot number 980622. Ra-
diochemical purity was reported to be 99.5%, with a specific
activity of 41.17 milliCuries/mmole. Upon receipt, the com-
pound identity was confirmed by gas chromatography with
electron impact mass spectroscopy. The radiochemical purity
and specific activity specified by Wizard Laboratories were
confirmed by high-performance liquid chromatography
The experimental design for the river die-away studies was
based on standard methods issued by the U.S. Environmental
Protection Agency (U.S. EPA), the American Society for Test-
ing and Materials, and the Society of Environmental Toxicol-
ogy and Chemistry [11–13]. The effects of BPA concentration,
river sediment concentration, and microbial adaptation on BPA
degradation kinetics were addressed in the experimental de-
14
sign. The degradation of [ C]BPA was routinely followed
(
HPLC) with radiochemical and ultraviolet (UV) detection.
Chemicals used as analytical reference compounds, re-
agents, and solvents were obtained from commercial sources
with the appropriate documentation of purity. Deionized water
was purified in a Milli-Q Water Purification System (Milli-
through at least three half-lives or greater than 90% degra-
dation. The duration of the studies was sufficient to define the
degradation pattern of BPA and possible metabolites. All phas-
es of the study were conducted in compliance with Good Lab-
oratory Practice Standards as described by the U.S. EPA [14]
and OECD [15].
pore, New Bedford, MA, USA).
The biodegradation of BPA was examined in river die-away
experiments using both surface water alone and sediment/wa-
ter mixtures. Initial biodegradation experiments were run using
two different methods conducted in parallel. With one method,
River water and sediment samples
River water and sediment samples were collected from eight
locations on seven rivers or estuaries across the United States
and Europe from July 1998 to January 2000. Samples were
collected both upstream and downstream of outfalls of waste-
water treatment systems known to receive BPA. Sampling lo-
cations included the Ohio River in Mt. Vernon, Indiana, USA;
the Ware River in Ware, Massachusetts, USA; the Monte Sano
Bayou in Baton Rouge, Louisiana, USA; the Mississippi River
in Baton Rouge, Louisiana, USA; the Rhine River in Krefeld,
Germany; the Elbe River in Stade, Germany; and the Wester-
schelde River in Bergen-op-Zoom, The Netherlands.
The procedure for collecting water and sediment samples
depended on the depth and size of the river. For the Rhine,
Ohio, and Westerschelde rivers, water sampling required use
of a boat. For the other rivers, water samples were collected
from the bank or by wading into the river. Water samples were
collected by immersing 4-L amber glass bottles just below the
surface to minimize collection of floating debris. During sam-
pling, a number of water quality parameters were measured,
including water temperature, pH, dissolved oxygen concen-
tration, and conductivity.
Surface sediment samples (i.e., the top 1–2 cm) were col-
lected from depositional zones in 1-L polyethylene bottles. A
diver was used to locate and collect sediments from the Ohio
River. Sediments from the Rhine and Westerschelde rivers
were collected using grab sampling devices operated from
boats. For the remaining rivers, sediment samples were col-
lected from the bank or by wading into the river and scooping
the sediment into the collection bottle.
Water and sediment samples were chilled with ice packs
and shipped via overnight express to the laboratory. Upon
receipt, separate composite samples of upstream and down-
stream sediment samples were prepared and passed through a
14
[
C]BPA was added to river water and water/sediment mix-
tures in sealed shake flasks fitted with CO traps and sampling
2
ports. The advantage of this procedure was the ability to ex-
14
amine [ C]BPA biodegradation over a range of initial test
chemical concentrations and to monitor the fate the chemical,
14
formation of [ C]metabolites, and the mineralization of
14
14
[
C]BPA to CO . Studies were also performed using a Co-
2
lumbus MicroOxymax respirometer (Columbus Instruments,
Columbus, OH, USA), where oxygen consumption and pro-
duction of CO due to the biodegradation of BPA were mea-
2
sured. The advantages of the respirometer were efficiency,
continuous monitoring of the degradation process, and low
cost, although higher concentrations of BPA (ϳ5,000 g/L)
were required to achieve the necessary sensitivity.
Biologically inhibited (killed) controls were included in all
studies to verify degradation was due to biological activity.
Killed controls were prepared by autoclaving the water and
sediment (121ЊC, 15 psi, 30 min) prior to the addition of BPA.
Autoclaving was used because preliminary experiments in-
dicated that chemical sterilants such as mercuric chloride (500
mg/L) or formalin (2%) could react with BPA under the test
conditions.
[14
C]BPA die-away studies
Initial die-away experiments were conducted with [ C]BPA
14
to define biodegradation kinetics over a range of BPA con-
centrations. The range was bounded at the low end by the
detection limit of the analytical methods and at the high end
by the lowest concentration that could be tested in the respi-
rometer (5,000
placing 200-ml portions of river water in 500-ml test vessels.
Reaction mixtures were amended with 50 and 500 g/L
l portions of stock solutions
of [ C]BPA in 1,4-dioxane. To prepare microcosms containing
g/L). Microcosms mixtures were prepared by
2
-mm sieve to remove stones and improve homogeneity. Water
14
samples containing high levels of particulates (by visual in-
spection) were filtered through glass wool prior to use. Die-
away studies were set as soon as possible, generally within 24
to 48 h of receipt of the samples.
[ C]BPA by the addition of 15
14
1
4
5,500
g/L [ C]BPA, a large volume of water (5 L) was
g of nonlabeled BPA, distrib-
initially amended with 25,000