Paul A. White

The effect of temperature and algal biomass on bacterial production and specific growth rate in freshwater and marine habitats.

We analyzed heterotrophic, pelagic bacterial production and specific growth rate data from 57 studies conducted in fresh, marine and estuarine/coastal waters. Strong positive relationships were identified between 1) bacterial production and bacterial abundance and 2) bacterial production and algal biomass. The relationship between bacterial production and bacterial abundance was improved by also considering water temperature. The analysis of covariance model revealed consistent differences between fresh, marine and estuarine/coastal waters, with production consistently high in estuarine/coastal environments. The log-linear regression coefficient of abundance was not significantly different from 1.00, and this linear relationship permitted the use of specific growth rate (SGR in day−1) as a dependent variable. A strong relationship was identified between specific growth rate and temperature. This relationship differed slightly across the three habitats. A substantial portion of the residual variation from this relationship was accounted for by algal biomass, including the difference between marine and estuarine/coastal habitats. A small but significant difference between the fresh- and saltwater habitats remained. No significant difference between the chlorophyll effect in different habitats was identified. The model of SGR against temperature and chlorophyll was much weaker for freshwater than for marine environments. For a small subset of the data set, mean cell volume accounted for some of the residual variation in SGR. Pronounced seasonality, fluctuations in nutrient quality, and variation of the grazing environment may contribute to the unexplained variation in specific growth.

The genotoxicity of priority polycyclic aromatic hydrocarbons in complex mixtures

Risk assessment of complex environmental samples suffers from difficulty in identifying toxic components, inadequacy of available toxicity data, and a paucity of knowledge about the behavior of geno(toxic) substances in complex mixtures. Lack of information about the behavior of toxic substances in complex mixtures is often avoided by assuming that the toxicity of a mixture is simply the sum of the expected effects from each mixture component, i.e. no synergistic or antagonistic interactions. Although this assumption is supported by research investigating non-genotoxic end-points, the literature describing the behavior of genotoxic substances in complex mixtures is sparse and, occasionally, contradictory. In this study, the results of polycyclic aromatic hydrocarbon (PAH) analyses on freshwater bivalves were used to prepare realistic mixtures containing up to 16 PAHs. The SOS genotoxicity of the mixtures and each component were then assessed in an effort to evaluate the additivity of PAH genotoxicity. At nominal PAH concentrations above 1 microg/ml, observed genotoxic responses were far lower than those predicted under the assumption of additivity. At nominal concentrations below 0.75 microg/ml, differences are smaller and occasionally negligible, indicating that the genotoxicity of unsubstituted homocyclic PAHs is additive or slightly less than additive. Other researchers who have investigated the mutagenicity, carcinogenicity, and DNA binding activity of mixtures containing unsubstituted homocyclic PAHs have also reported additive effects. Therefore, the mutagenic risk posed by simple, well-characterized mixtures of priority PAHs can reasonably be estimated as the sum of the risks posed by the mixture components. Current data indicate that less-than-additive effects likely result from saturation of metabolic pathways needed to activate mutagenic PAHs.

Comparing the presence, potency, and potential hazard of genotoxins extracted from a broad range of industrial effluents

We examined the genotoxicity of dichloromethane extracts from 50 final effluent samples collected from 42 industries, including pulp and paper, chemical manufacturing, metal refining, metal surface treatment, and municipal waste water treatment. Effluents were initially fractionated into dissolved substances, and substances adsorbed to suspended particulate matter. Acid/base partitioning was used to further fractionate aqueous extracts. Genotoxicity was measured using the SOS Chromotest. Genotoxicity of extracts was found to be related to sample type, industry type, metabolic activation status, and extract fluorescence (380 nm excitation, 430 nm emission). S9 metabolic activation reduced genotoxic potency in over 90% of the extracts examined. Expression of potency values per equivalent unit of original sample revealed that effluent particulate matter is, on average, almost four orders of magnitude more potent than aqueous filtrates. Suspended particulate matter from organic and inorganic chemical production, petroleum and metal refining, and from metal surface treatment facilities, provided extracts that were significantly more genotoxic than those from sewage treatment and pulp and paper facilities. Aqueous filtrates from inorganic and organic chemical production, metal refining, and surface treatment facilities were significantly more genotoxic than those emitted by aluminum and petroleum refineries. Overall, the results suggest that pulp and paper mills emit mostly soluble genotoxins, while petroleum and aluminum refineries emit predominantly particle‐associated genotoxins. Although some extracts elicited a strong SOS response, the potency of the extractable residues was low when compared to highly potent pure substances such as benzo(a)pyrene. On average, a mg of dichloromethane‐extractable residue has an SOS genotoxicity equivalent to 0.1–1.0 μg of benzo(a)‐pyrene. Predicted Ames mutagenic potency values corresponded reasonably well with industrial waste mutagenic potency values published by other researchers. Genotoxic loading values were calculated to quantify the total daily genotoxic emission and potential hazard of each industry. Highest loadings were from sewage treatment, pulp and paper, and metal refining facilities. Highest loading values were the SOS genotoxic equivalent of over 30 kg of benzo(a)pyrene per day. The ultimate hazard of genotoxic emissions is not known. Actual hazard assessment is complicated by a poor understanding of the postemission behavior of genotoxins. Exposure of downstream biota is likely substantial.

Mutagenic characteristics of river waters flowing through large metropolitan areas in North America

The hanging technique using blue rayon, which specifically adsorbs mutagens with multicyclic planar structures, has the advantages over most conventional methods of not having to bring large volumes of water back to the laboratory for extraction of organic materials. Therefore, for the same effort the hanging blue rayon technique allows for the analysis of more samples from remote sites, although it has a disadvantage of not allowing quantitative analysis. In this study, the blue rayon hanging technique was used to collect organic mutagens in river waters that flow through metropolitan areas in northeastern North America. Monitoring was performed at a total of 21 sites: the Providence River system (4 sites), the Charles River (2 sites), the Potomac River (6 sites), the St. Lawrence River (5 sites), the Hudson River (3 sites), and the East River (1 site). Mutagenicity was evaluated using the Salmonella assay with strains TA98, TA100, YG1024, YG1041, and YG1042 with and without metabolic activation. The results demonstrated that strains YG1041 and YG1024 were much more sensitive than TA98 with S9 mix. Fifteen samples out of 21 were positive in YG1041 with S9 mix. Six samples gave 5000-18,400 revertants/g blue rayon equivalent. YG1042 was also much more sensitive than TA100. Eight samples were positive in YG1042 with S9 mix. The highest activity was 10,200 revertants/g blue rayon equivalent. The overall results showed that rivers flowing through major cities in North America contain frameshift-type, aromatic amine-like mutagenic activity. However, the levels of mutagenic activity in these rivers were much lower than expected based on prior analyses and calculated population-to-discharge ratios. Further research, such as detailed chemical analyses and/or simultaneous comparisons of several different adsorbents (e.g. XAD and blue rayon), will be needed to clarify the observed differences between North American blue rayon values and published values for European and Asian river systems.

A semi-automated, microplate version of the SOS Chromotest for the analysis of complex environmental extracts☆

Environmental monitoring for genotoxicity requires that a large number of measurements be made across space and time. This requirement demands a rapid and efficient bioassay system. The SOS Chromotest is a rapid, efficient bacterial system for the detection of DNA damaging agents. Over 100 publications have described its use on a variety of samples. Relatively few studies have used the test to examine complex mixtures. Effective testing of complex samples poses a variety of problems. Although solutions have been proposed, few have validated the resulting protocol. In this work we present a semi-automated microplate version of the SOS Chromotest for the examination of complex mixtures. Experiments were conducted to determine the optimal cell concentration, exposure time, substrate conversion time and S9 enzyme concentration. The performance of the method was evaluated using 6 reference genotoxins and 3 complex mixtures. The complex mixtures examined are extracts of diesel particulate matter, urban dust and coal tar. The results obtained indicate that optimal responses often require fewer cells (approximately equal to 5-10 x 10(6) CFU/ml) and a longer exposure (3 h) than that recommended in the original protocol. Interfering effects of colored and turbid samples are removed using centrifugation and initial optical density readings taken 60 min after cell resuspension and lysis. The performance of the established protocol was evaluated using mitomycin C and benzo[a]pyrene results for 207 microplates and solvent control results for 293 microplates. The results indicate that the established method is accurate, sensitive and precise. Coefficient of variation on mean SOSIP values for mitomycin C and benzo[a]pyrene are < 5%. Solvent control data indicate that the standard threshold for determination of a positive response (induction factor > 1.5) is excessively conservative. All liquid transfers were automated using the Biomek automated laboratory workstation. Automation permits a throughput of up to 72 samples per day and maintains excellent precision and accuracy.

Steady-State Solutions to PBPK Models and Their Applications to Risk Assessment I: Route-to-Route Extrapolation of Volatile Chemicals

Although analysis of in vivo pharmacokinetic data necessitates use of time-dependent physiologically-based pharmacokinetic (PBPK) models, risk assessment applications are often driven primarily by steady-state and/or integrated (e.g., AUC) dosimetry. To that end, we present an analysis of steady-state solutions to a PBPK model for a generic volatile chemical metabolized in the liver. We derive an equivalent model that is much simpler and contains many fewer parameters than the full PBPK model. The state of the system can be specified by two state variables-the rate of metabolism and the rate of clearance by exhalation. For a given oral dose rate or inhalation exposure concentration, the system state only depends on the blood-air partition coefficient, metabolic constants, and the rates of blood flow to the liver and of alveolar ventilation. At exposures where metabolism is close to linear, only the effective first-order metabolic rate is needed. Furthermore, in this case, the relationship between cumulative exposure and average internal dose (e.g., AUCs) remains the same for time-varying exposures. We apply our analysis to oral-inhalation route extrapolation, showing that for any dose metric, route equivalence only depends on the parameters that determine the system state. Even if the appropriate dose metric is unknown, bounds can be placed on the route-to-route equivalence with very limited data. We illustrate this analysis by showing that it reproduces exactly the PBPK-model-based route-to-route extrapolation in EPAs 2000 risk assessment for vinyl chloride. Overall, we find that in many cases, steady-state solutions exactly reproduce or closely approximate the solutions using the full PBPK model, while being substantially more transparent. Subsequent work will examine the utility of steady-state solutions for analyzing cross-species extrapolation and intraspecies variability.

SOS chromotest results in a broader context: Empirical relationships between genotoxic potency, mutagenic potency, and carcinogenic potency

Environmental monitoring requires that large numbers of samples be processed in a relatively short period of time. While microbioassays facilitate rapid testing, the results are often difficult to interpret in the broader context of human or animal health. Determining the consequences of exposure to genotoxic substances will ultimately require in situ monitoring of exposed organisms. However, it is immediately possible to construct a broad empirical framework within which available microbioassay results can be interpreted. To do this for SOS Chromotest results, we investigated the empirical relationships between SOS genotoxic potency and mutagenic potency (as measured with the Salmonella/microsome assay), as well as between genotoxic potency and carcinogenic potency (as measured using standard, chronic animal bioassays). Strong relationships were identified between; 1) genotoxic potency and mutagenic potency for 268 direct‐acting substances (r2 = 0.76) and 2) genotoxic potency and mutagenic potency for 126 S9‐activated substances (r2 = 0.65). Ordinary least squares regression analyses of the SOS genotoxicity‐Salmonella mutagenicity relationship revealed a significant effect of SOS genotoxicity as well as differences in mutagenic potency that can be attributed to the Salmonella strain used to measure mutagenic potency. Analyses of S9‐activated substances revealed a significant interaction between the SOS genotoxic potency (SOSIP) effect and the Salmonella strain effect. Two regression models relating SOS genotoxicity and Salmonella mutagenicity were used to predict the mutagenic potency of several industrial effluent extracts previously analyzed for SOS genotoxicity by White et al. [(1996): Environ Mol Mutagen 27:116–139]. Predictions are consistent with published mutagenic potency values for similar industrial waste materials. A consistent relationship was also identified between genotoxic potency and carcinogenic potency for 51 substances. Linear regression analyses revealed an effect of SOS genotoxic potency as well as differences in carcinogenic potency that may be attributable to experimental animal and route of exposure. The correlation between genotoxicity and carcinogenicity was fairly weak (maximum r value = 0.51). Previous studies revealed similar strength of association between Ames mutagenicity and carcinogenicity. Predicted carcinogenic potencies of previously examined genotoxic, industrial effluent extracts are generally low compared to the pure substances included in the data set.

Detection of genotoxic substances in bivalve molluscs from the Saguenay Fjord (Canada), using the SOS chromotest.

Few studies have employed bioassays to investigate the accumulation of genotoxins in aquatic biota that inhabit areas contaminated with industrial and municipal wastes. This study employed the SOS Chromotest, a short-term bacterial genotoxicity assay, to investigate the presence of genotoxins in bivalve molluscs from the Saguenay Fjord (Canada). Genotoxicity analyses were performed on dichloromethane extracts of Mya arenaria and Mytilus edulis collected downstream from several aluminum refineries and forestry products industries known to produce and release genotoxic substances. The results confirmed that bivalve molluscs inhabiting downstream regions are contaminated with both direct-acting and pro-genotoxic substances. In several cases, SOS response induction factors exceeded 3.0. The results failed to reveal a clear downstream trend of decreasing genotoxicity with increasing distance from the presumed industrial sources(s). A significant relationship (r2 = 0.61, p < 0.007) between a demographic variable (population near shoreline) and lipid-corrected genotoxic potency suggest that the accumulated direct-acting genotoxins may be of municipal origin. Significant relationships between tissue extract genotoxicity (r2 = 0.75, p < 0.003) and tissue PAH contamination (r2 = 0.77, p < 0.0001) and drainage basin area suggests that the bivalves are accumulating airborne contaminants deposited on the surface of the relevant drainage basins. In spite of contamination with genotoxic PAHs, the addition of rat liver microsomal enzymes reduced the genotoxic potency of all samples investigated (31-94% decrease). The results also revealed a significant relationship between tissue extract genotoxicity and PAH concentration (r2 = 0.72, p < 0.0005). Further analyses confirmed that a variable portion (7-97%) of the S9-activated tissue extract genotoxicity can be attributed to the detected PAHs. Although the sources, identity and effects of genotoxins accumulated by bivalves of the Saguenay Fjord remain to be determined, the study has confirmed the utility of the SOS Chromotest in environmental monitoring of aquatic biota.

Sorption of organic genotoxins to particulate matter in industrial effluents

In an earlier work [White PA et al. (1996): Environ Mol Mutagen 27:116–139] we examined the genotoxicity of dichloromethane extracts from a variety of industrial effluent samples. In this companion work, we used the SOS Chromotest to investigate the sorption of the extracted genotoxins to effluent suspended particulate matter. The affinity of the genotoxins for particulate matter is expressed as a genotoxicity sorption partition coefficient (Kd‐genotox). The results indicate that industries known for their emission of combustion by‐products, such as polycyclic aromatic hydrocarbons, often have high Kd‐genotox values (≥106). These include metal refining and founding industries as well metal surface treatment facilities. In contrast, Kd‐genotox values for pulp and paper mills and sewage treatment facilities are several orders of magnitude lower (≤104). In several cases the calculated Kd‐genotox values are in agreement with the Kow values of genotoxic substances isolated from genotoxic industrial waste samples studied by other researchers. The sorption partition coefficient, in conjunction with the concentration of available particulate matter, was used to determine the percent of organic genotoxins adsorbed to effluent suspended particulate matter. Values range from 2.3% to 99.8%. High values (>70%) were obtained for metal surface treatment and inorganic and organic chemical production facilities. Low values (<30%) were obtained for sewage treatment facilities and pulp and paper mills. The results also demonstrate the effect of variations in the concentration of available particulate matter on the genotoxicity of both aqueous and particulate extracts. The results suggest that the sorptive properties of the particulate matter itself are reduced when the concentration of particulate matter is very high (>1,000 mg per l). The use of sorption partition information in inferring the physical‐chemical nature of the putative genotoxins and the implications of the results for assessing the hazard posed to aquatic biota by industrial genotoxins are discussed.

The use of PBPK models to inform human health risk assessment: case study on perchlorate and radioiodide human lifestage models.

Physiologically-based pharmacokinetic (PBPK) models are often submitted to or selected by agencies, such as the U.S. Environmental Protection Agency (U.S. EPA) and Agency for Toxic Substances and Disease Registry, for consideration for application in human health risk assessment (HHRA). Recently, U.S. EPA evaluated the human PBPK models for perchlorate and radioiodide for their ability to estimate the relative sensitivity of perchlorate inhibition on thyroidal radioiodide uptake for various population groups and lifestages. The most well-defined mode of action of the environmental contaminant, perchlorate, is competitive inhibition of thyroidal iodide uptake by the sodium-iodide symporter (NIS). In this analysis, a six-step framework for PBPK model evaluation was followed, and with a few modifications, the models were determined to be suitable for use in HHRA to evaluate relative sensitivity among human lifestages. Relative sensitivity to perchlorate was determined by comparing the PBPK model predicted percent inhibition of thyroidal radioactive iodide uptake (RAIU) by perchlorate for different lifestages. A limited sensitivity analysis indicated that model parameters describing urinary excretion of perchlorate and iodide were particularly important in prediction of RAIU inhibition; therefore, a range of biologically plausible values available in the peer-reviewed literature was evaluated. Using the updated PBPK models, the greatest sensitivity to RAIU inhibition was predicted to be the near-term fetus (gestation week 40) compared to the average adult and other lifestages; however, when exposure factors were taken into account, newborns were found to be populations that need further evaluation and consideration in a risk assessment for perchlorate.