Approaches to mixture risk assessment of PFASs in the European population based on human hazard and biomonitoring data.

Per- and polyfluoroalkyl substances (PFASs) are a highly persistent, mobile, and bioaccumulative class of chemicals, of which emissions into the environment result in long-lasting contamination with high probability for causing adverse effects to human health and the environment. Within the European Biomonitoring Initiative HBM4EU, samples and data were collected in a harmonized way from human biomonitoring (HBM) studies in Europe to derive current exposure data across a geographic spread. We performed mixture risk assessments based on recent internal exposure data of PFASs in European teenagers generated in the HBM4EU Aligned Studies (dataset with N = 1957, sampling years 2014-2021). Mixture risk assessments were performed based on three hazard-based approaches: the Hazard Index (HI) approach, the sum value approach as used by the European Food Safety Authority (EFSA) and the Relative Potency Factor (RPF) approach. The HI approach resulted in the highest risk estimates, followed by the RPF approach and the sum value approach. The assessments indicate that PFAS exposure may result in a health risk in a considerable fraction of individuals in the HBM4EU teenager study sample, thereby confirming the conclusion drawn in the recent EFSA scientific opinion. This study underlines that HBM data are of added value in assessing the health risks of aggregate and cumulative exposure to PFASs, as such data are able to reflect exposure from different sources and via different routes.

Modelling of the feed-to-fillet transfer of ethoxyquin and one of its main metabolites, ethoxyquin dimer, to the fillet of farmed Atlantic salmon (Salmon salar L.).

Ethoxyquin (EQ) is an antioxidant supplemented to feed ingredients, mainly fish meal, which is currently under re-evaluation for use in the food production chain. EQ is partly metabolized into several metabolites of which the ethoxyquin dimer (EQDM) accumulates most in the farmed fish fillet. In this study, the feed-to-fillet transfer of dietary EQ and EQDM in Atlantic salmon fillet was investigated, and a physiologically based pharmacokinetic (PBPK-) two-compartmental model was developed, based on experimental determined EQ and EQDM uptake, metabolism, and elimination kinetics. The model was verified with an external data-set and used to simulate the long term (>1.5 years) EQ and EQDM feed-to fillet transfer in Atlantic salmon under realistic farming conditions such as the seasonal fluctuations in feed intake, growth, and fillet fat deposition. The model predictions showed that initial EQDM levels in juvenile fish are the driving factor in final levels found in food-producing animals, while for EQ the levels in feed, and seasonal variations were the driving factor for food EQ levels.

Does EU legislation allow the use of the Benchmark Dose (BMD) approach for risk assessment?

Hazard characterisation is largely based on an approach of (statistically) comparing dose groups with the controls in order to derive points of departure such as no-observed-adverse-effect levels (NOAELs) or lowest-observed-adverse-effect levels (LOAELs). This approach suggests the absence of any relevant effect at the NOAEL. The NOAEL approach has been debated for decades. A recent Scientific Opinion by the European Food Safety Authority (EFSA) concluded that the Benchmark Dose (BMD) approach should be preferred over the NOAEL approach for deriving human (health-based) limit or guidance values. Nonetheless, the BMD approach is used infrequently within European regulatory frameworks. The reason for this may lie in legislation or guidelines requiring the use of the NOAEL approach. In this context, various EU regulatory frameworks were examined on such demands. Interestingly, no single legislation was identified containing statutory requirements in conflict with the use of the BMD approach.

Toxicokinetic model assessment on the dechlorination of dietary toxaphene CHB-62 into CHB-44 in Atlantic salmon (Salmo salar L.).

European Union legislation on the upper limits of toxaphene in feed and food include the congeners CHB-26, CHB-62 and CHB-50 and is set at 50 µg kg⁻¹ feed for the sum of these three congeners. However, due to their elevated presence in fish, the congeners CHB-40 and CHB-41, CHB-44, and CHB-42 should also be included according to the European Food Safety Authority (EFSA) in 2005. Earlier trials with model zebra fish have shown in vivo dechlorination of dietary CHB-62 to CHB-44 and, to a lesser degree, of CHB-50 to CHB-40. Biomagnification patterns of Atlantic salmon, fed with technical toxaphene-enriched feeds, indicated that Atlantic salmon have a similar dechlorination. In the present study, a serial one-compartment physiological kinetic model, which includes differentiated growth of body components, is used to quantify the contribution of dechlorination to the congener-specific fillet accumulation of a mixture of dietary toxaphene congeners in Atlantic salmon. The model is assessed from experimental uptake and elimination kinetics of Atlantic salmon smolt fed with technical toxaphene for 122 days followed by a depuration period of 75 days in which the fish were fed toxaphene-free control feed. The serial one-compartment model shows that about 31% of CHB-44 that accumulated in the fillet originated from dietary CHB-62. In contrast, dechlorination of CHB-50 into CHB-40 is not significant. The results show that previously demonstrated in vivo dechlorination of CHB-62 into CHB-44 in zebra fish also occurs in the farmed fish species Atlantic salmon. This dechlorination can at least partly explain the relatively elevated CHB-44 observed in toxaphene fish surveys.

Transfer of perfluorooctane sulfonic acid (PFOS) from contaminated feed to dairy milk.

Dietary intake is the predominant route for human exposure to perfluorooctane sulfonic acid (PFOS). Single pollution events may thus affect human exposure if polluted ground and water is used to produce animal feed or food. In this study, a physiologically based pharmacokinetic (PBPK-) model is derived that describes the uptake of PFOS from contaminated feed by cows and its subsequent elimination through the cows' milk. Parameter values of the model were estimated by fitting to experimental data of a cow feeding trial. Model calculations showed that almost all PFOS ingested is excreted through the cows' milk. The elimination rate, however, was low as the estimated half-life in the cow was 56days and it may, thus, take a long time after an initial pollution event to produce PFOS-free milk. The derived model can be used to estimate the transfer of PFOS through the dairy food chain and can be used for comparison of various contamination routes.

Benzo[a]pyrene-induced transcriptomic responses in primary hepatocytes and in vivo liver: toxicokinetics is essential for in vivo-in vitro comparisons.

The traditional 2-year cancer bioassay needs replacement by more cost-effective and predictive tests. The use of toxicogenomics in an in vitro system may provide a more high-throughput method to investigate early alterations induced by carcinogens. Recently, the differential gene expression response in wild-type and cancer-prone Xpa (-/-) p53 (+/-) primary mouse hepatocytes after exposure to benzo[a]pyrene (B[a]P) revealed downregulation of cancer-related pathways in Xpa (-/-) p53 (+/-) hepatocytes only. Here, we investigated pathway regulation upon in vivo B[a]P exposure of wild-type and Xpa (-/-) p53 (+/-) mice. In vivo transcriptomics analysis revealed a limited gene expression response in mouse livers, but with a significant induction of DNA replication and apoptotic/anti-apoptotic cellular responses in Xpa (-/-) p53 (+/-) livers only. In order to be able to make a meaningful in vivo-in vitro comparison we estimated internal in vivo B[a]P concentrations using DNA adduct levels and physiologically based kinetic modeling. Based on these results, the in vitro concentration that corresponded best with the internal in vivo dose was chosen. Comparison of in vivo and in vitro data demonstrated similarities in transcriptomics response: xenobiotic metabolism, lipid metabolism and oxidative stress. However, we were unable to detect cancer-related pathways in either wild-type or Xpa (-/-) p53 (+/-) exposed livers, which were previously found to be induced by B[a]P in Xpa (-/-) p53 (+/-) primary hepatocytes. In conclusion, we showed parallels in gene expression responses between livers and primary hepatocytes upon exposure to equivalent concentrations of B[a]P. Furthermore, we recommend considering toxicokinetics when modeling a complex in vivo endpoint with in vitro models.

Toxicokinetics and carry-over model of α-hexabromocyclododecane (HBCD) from feed to consumption-sized Atlantic salmon (Salmo salar).

A two-compartmental model for the kinetics of carry-over of the brominated flame retardant α-hexabromocyclododecane (HBCD) from feed to the fillet of farmed harvest-sized Atlantic salmon (Salmo salar L.) was developed. The model is based on a fat compartment for storage of the lipophilic α-HBCD and a central compartment comprising all other tissues. Specific for this model is that the salmon has a continuous growth and that fillet contaminant levels are explained by both the fat and the central compartments. The uptake and elimination kinetics are obtained from experimental data where consumer sized (start weight approximately 1 kg) Atlantic salmon was fed α-HBCD spiked feed (280 ± 11 µg kg(-1)) for 2 months followed by a depuration period of 3 months. The model was used to simulate the HBCD feed-to-fillet transfer in Atlantic salmon under realistic farming conditions such as the seasonal fluctuations in feed intake, growth and fillet fat deposition. The model predictions gave fillet concentrations of 0.2-1.8 µg kg(-1) depending on the level of fish oil inclusion in the salmon diets when using fish oil with high POP background levels. Model simulations show that currently farmed Atlantic salmon can contribute to a maximum of 6% of the estimated provisional food reference dose for HBCD.

Carry-over of pyrrolizidine alkaloids from feed to milk in dairy cows.

Pyrrolizidine alkaloids are toxins present in many plants belonging to the families of Asteraceae, Boraginaceae and Fabaceae. Particularly notorious are pyrrolizidine alkaloids present in ragwort species (Senecio), which are held responsible for hepatic disease in horses and cows and may lead to the death of the affected animals. In addition, these compounds may be transferred to edible products of animal origin and as such be a threat for the health of consumers. To investigate the possible transfer of pyrrolizidine alkaloids from contaminated feed to milk, cows were put on a ration for 3 weeks with increasing amounts (50-200 g day(-1)) of dried ragwort. Milk was collected and sampled twice a day; faeces and urine twice a week. For milk, a dose-related appearance of pyrrolizidine alkaloids was found. Jacoline was the major component in milk despite being a minor component in the ragwort material. Practically no N-oxides were observed in milk, notwithstanding the fact that they constituted over 80% of the pyrrolizidine alkaloids in ragwort. The overall carry-over of the pyrrolizidine alkaloids was estimated to be only around 0.1%, but for jacoline 4%. Notwithstanding the low overall carry-over, this may be relevant for consumer health considering the genotoxic and carcinogenic properties demonstrated for some of these compounds. Analysis of the faeces and urine samples indicated that substantial metabolism of pyrrolizidine alkaloids is taking place. The toxicity and potential transfer of metabolites to milk is unknown and remains to be investigated.

A novel source for dioxins present in recycled fat from gelatin production.

Within a survey on dioxins in animal fat used as feed ingredient, a sample originating from pigs offal was shown to contain 50 ng Toxic Equivalents (TEQ) PCDD/PCDFs kg(-1) fat. Further investigation revealed fat samples with levels as high as 440 ng TEQ kg(-1) fat and contaminated feed with a highest level of 8.4 ng TEQ kg(-1) feed. The congener pattern was dominated by 1,2,3,7,8-PeCDD and 2,3,7,8-TCDD, and was not recognized from any previous incident or known dioxin source. Remarkably, 2,3,7,8-substituted congeners were much more abundant than their non-2,3,7,8-substituted counterparts. The sampled fat was derived from a gelatin production plant. Broken filters, used to clean the hydrochloric acid (HCl) used in the process, caused the dioxin contamination. The fat was primarily used for pig feed. A new physiologically-based pharmacokinetic (PBPK) model for lipophilic contaminants in growing slaughter pigs predicted levels at slaughter varying between 40 pg TEQ g(-1) fat (worst-case) and 2.5-7pgTEQ g(-1) fat under more realistic scenarios. Almost 300 farms were temporarily blocked. Many fat samples of pigs were analyzed using a combined approach of DR CALUX and GC/HRMS. Levels in contaminated pig fat were around the EU-limit of 1 pg TEQ g(-1) fat, with some samples up to 2-3 pg TEQ g(-1) fat. Of 80 negative samples analyzed by DR CALUX and GC/HRMS no false-negatives were obtained, whereas 36 and 62 of the 80 samples classified suspected with the bioassay had GC/HRMS levels above respectively the tolerance and action limits. It is concluded that novel and unexpected dioxin sources remain a threat to the food chain and require the proper evaluation and monitoring of production processes, including chemicals used therein.

A toxicokinetic model for the carry-over of dioxins and PCBs from feed and soil to eggs.

A mathematical model for the kinetics of carry-over of dioxins and dioxin-like PCBs from feed mixed with contaminated oil to eggs has been developed. This model incorporates uptake of the compounds over the gut wall and their subsequent transport by blood, distribution over the body, hepatic metabolism and excretion through egg yolk fat. The model is analysed with respect to the possibility of identifying as yet unknown model parameters by fitting these to the experimental data. The model was fitted to the experimental data on the carry-over from feed to eggs. The calibrated model was applied to calculate the steady-state concentrations in eggs which were compared to European Maximum Residue Levels for dioxins in feed and eggs, showing that these limits do not match. The feed limit of 0.75 ng TEQ/kg should be reduced to about 0.17 ng TEQ/kg in order to guarantee egg levels below the residue limit of 3 pg TEQ/g fat. Experimental results of carry-over from contaminated soil were used to estimate the absorption of dioxin-like compounds from soils as compared to the absorption from feed, resulting in a value around 40 to 60% absorption from soil as compared to around 90% absorption from feed.

Carry-over of dioxins and PCBs from feed and soil to eggs at low contamination levels-- influence of mycotoxin binders on the carry-over from feed to eggs.

Laying hens were fed with compound feed containing six different levels of dioxins, dioxin-like PCBs and indicator PCBs for a period of 56 days. This was followed by a period of 56 days on clean feed. Dioxin levels in feed varied from background levels to three times the current EU tolerance limit of 0.75 ng TEQ/kg. At all dose levels a rapid increase was observed in the dioxin levels in eggs. There was a clear linear dose-response relationship between the dioxin levels in eggs and feed. The feed containing 0.4 ng TEQ dioxins per kg resulted in egg levels just above the EU limit of 3 pg TEQ/g fat. Dioxin-like and indicator PCB residues followed a pattern very similar to that of dioxins. Exposure to the highest indicator PCB level of 32 microg/kg resulted in egg levels around 300 ng/g fat. Exposure to dioxins through contaminated soil, mixed at 10% into the feed, resulted in a similar carry-over as from feed. Mycotoxin binders, mixed at 0.5% into the feed, had little effect on the carry-over of dioxins from the feed to the egg. It can be concluded that consumption of feed or soil with even moderate levels of dioxins and dioxin-like PCBs rapidly results in increased levels in eggs. The current EU dioxin limit for feed cannot guarantee egg dioxin levels below the EU-limit.

Evaluation of the DNA-repair host-mediated assay. III. Relationship between metabolic activation of dimethylnitrosamine and organ-specific differential lethality induced in E. coli indicator strains.

In the present study the sensitivity of differential lethality as an endpoint for monitoring the presence of organ-specific genotoxic factors within the DNA-repair host-mediated assay (HMA) was determined. The induction of differential lethality in chemically exposed animals was assessed by measuring the recovery ratio Q, i.e., the relative survival of a repair-deficient E. coli K-12 derivative in comparison with its repair-proficient counterpart. Using untreated animals the interindividual fluctuation of the recovery ratio Q was first quantified and then used to determine the level below which it could be considered indicative of chemically induced differential lethality. This Q value was found to be 0.65 or lower. Using this criterion, a significant decrease of the Q value was observed in mice exposed to DMNA at a dose level as low as 15-30 mumole/kg, i.p. Inter-organ transport (liver----extrahepatic organs) of indicator bacteria was studied in reconstruction experiments using the direct-acting methylating agent MNU. These studies showed that inter-organ transport of indicator bacteria did not interfere with MNU-induced differential lethality. Time-related experiments were used to study the effects of inter-organ transport of genotoxic DMNA metabolites. In these studies significant, time-related differences were found in the induction of differential lethality in various organs of mice treated with DMNA. At a dose level of 200 mumole/kg (i.p.) genotoxic factors appeared within 25 min after administration in the liver. In the lungs and kidneys such factors appeared at a substantially slower rate, e.g., 20-120 min after DMNA administration. In persistence experiments differential lethality reached a maximum 30 min after DMNA treatment. No residual effects were detected 60 min after the injection of the carcinogen. These experiments showed that DMNA-derived genotoxic factors diffused from the liver into the bloodstream. The diffusion of these reactive species followed by their transport via the bloodstream to the lungs accounted for maximally 50% of differential lethality observed in bacteria recovered from the latter organ. In contrast, no indications were found for the transport of genotoxic DMNA metabolites from the liver via the bloodstream to the spleen and the kidneys. These results show that organ-specific effects observed in the DNA-repair HMA procedure after DMNA exposure can be primarily attributed to in situ metabolism, rather than diffusion of genotoxic metabolites from the liver to extrahepatic organs.

Mutational specificities of environmental carcinogens in the lacl gene of Escherichia coli H. V: DNA sequence analysis of mutations in bacteria recovered from the liver of Swiss mice exposed to 1,2-dimethylhydrazine, azoxymethane, and methylazoxymethanolacetate.

The host-mediated assay (HMA) was used to determine the spectra of mutations induced in the lacl gene of Escherichia coli cells recovered from the livers of Swiss mice exposed to the carcinogens 1,2-dimethylhydrazine (SDMH), azoxymethane (AOM), and methylazoxymethanolacetate (MAMA). These spectra were further compared with changes induced by dimethylnitrosamine (DMNA) in the HMA methodology. A total of 177 independent lacl mutations arising in the HMA following exposure to SDMH, AOM, and MAMA were analyzed. Single-base substitutions accounted for 97% of all mutations analyzed. The vast majority of the single-base substitutions consisted of G:C----A:T transitions (94% of all mutations). The remaining mutations consisted of A:T----G:C transitions (3% of all mutations) while non-base substitutions accounted for only 3% of the total mutagenesis. The latter mutations consisted of one frameshift mutation and four lacO deletions. The distribution of G:C----A:T transitions induced by the three chemicals in the first 200 bp of the lacl gene was not random, but rather clustered at sites where a target guanine was flanked at the 5' site by a purine residue.

Preventive action of thioethers towards in vitro DNA binding and mutagenesis in E. coli K12 by alkylating agents.

Thioethers are effective scavengers of electrophilic metabolites derived from the hepatocarcinogen N-hydroxy-2-acetylaminofluorene (van den Goorbergh et al., 1987). In this study 2 of these thioethers, 4-(methylthio)benzoic acid (MTB) and its methylester, methyl 4-(methylthio)benzoate (MMTB), have been tested for their ability to prevent in vitro DNA binding and mutation induction in E. coli K12 by the direct alkylating agents ethylnitrosourea (ENU), methylnitrosourea (MNU), ethyl methanesulfonate (EMS) and methyl methanesulfonate (MMS). In addition to MTB and MMTB, the thioether L-methionine (Met), and the thiols glutathione (GSH) and L-cysteine (Cys) were included for reasons of comparison. MTB was able to (partially) prevent DNA binding and mutation induction by ENU. However, this thioether was ineffective with EMS. DNA binding and mutagenesis by EMS were (partially) prevented by GSH and Cys, while these thiols could not prevent DNA binding and mutation induction by ENU. MMTB was unable to prevent mutation induction by these ethylating agents. With the methylating agents, similar effects of MTB were observed: MTB effectively prevented mutation induction by MNU while it was much less effective towards MMS. GSH and Cys were comparably effective as antimutagenic agents towards both methylating agents. Met was unable to prevent either DNA binding or mutation induction by these agents. Taken together, the results show that aromatic thioethers are able to trap genotoxic electrophiles derived from the nitrosoureas ENU and MNU, and may therefore act as potential anticarcinogens towards these agents, which are only poorly detoxified by GSH.

Mutational specificities of environmental carcinogens in the lacl gene of Escherichia coli. II: A host-mediated approach to N-nitroso-N,N-dimethylamine and endogenous mutagenesis in vivo.

An intrasanguineous host-mediated assay was used to determine the mutational specificity of the hepatocarinogen N-nitroso-N,N-dimethylamine metabolized in vivo. A total of 114 forward mutations in the lacl gene of Escherichia coli reisolated from the livers of treated Swiss albino mice were characterized at the DNA sequence level. Consistent with the methylating ability of this compound and the demonstrated mutagenic specificity of O6-methylguanine, the predominant mutation was the G:C----A:T transition. These were recovered, on average, seven times more frequently at guanines flanked (5') by a purine residue than at those preceded by a pyrimidine residue--a specificity similar to that reported for many direct-acting SN1 alkylating agents. This nitrosamine appears to be distinguished from related N-nitroso methylating compounds by the induction of additional mutational events. Here, the exceptions consisted of four A:T----G:C transitions, four A:T site transversions, and a single G:C----T:A transversion. In addition, the DNA sequence alterations of 34 I- mutants of E. coli reisolated from otherwise untreated mice were identified. The predominant mutation was the G:C----A:T transition, which accounted for almost half of all background mutations. The sites at which these mutations were recovered appear to indicate that some of these mutations may have arisen as a result of an accelerated rate of cytosine deamination. These data suggest that many of the additional "spontaneous" mutations observed under in vivo conditions resulted from genotoxic events occurring during the host-defense (immune) reaction.

Inhibition of junctional intercellular communication as a possible short-term test to detect tumor-promoting agents: results with nine chemicals tested by dye transfer assay in Chinese hamster V79 cells.

In an attempt to establish an in vitro short-term test to detect tumor-promoting agents, we studied the effects of these agents on junctional intercellular communication in cultured Chinese hamster V79 cells using a microinjection-dye transfer technique. When Lucifer Yellow CH solution is injected into a cell, the average number of cells that become fluorescent after 10 min is 11.6 +/- 7.8 (SD). When the phorbol ester 12-O-tetradecanoylphorbol-13-acetate was used as a positive control, the extent of dye transfer was reduced to 2.9 +/- 2.1 cells within 2 h after incubation with 12-O-tetradecanoylphorbol-13-acetate (100 ng/ml). Nine chemicals that have been reported to have or suspected of having tumor-promoting activity in experimental animals were tested at different doses and after different incubation times. 1,1,1-Trichloro-2,2-bis(p-chlorophenyl)ethane, lindane (1,2,3,4,5,6-hexachlorocyclohexane), phenobarbital, and butylated hydroxyanisole showed inhibitory properties in V79 cells, but with kinetics different from that of 12-O-tetradecanoylphorbol-13-acetate. With 1,1,1-trichloro-2,2-bis(p-chlorophenyl)ethane and lindane, exposure for 24 h resulted in full blockage of dye transfer; with phenobarbital, a treatment time of 96 h was necessary to achieve this effect, and butylated hydroxyanisole was more active after 48 h than after 24 or 72 h incubation. Five of the reported or suspected tumor-promoting agents, benzoyl peroxide, anthralin, deoxycholic acid, lithocholic acid, and butylated hydroxytoluene, had no effect on communication between V79 cells at noncytotoxic doses; deoxycholic acid, lithocholic acid, and butylated hydroxytoluene but not anthralin inhibited communication only at cytotoxic doses. Our results indicate that we can detect several, but not all, types of tumor-promoting agents, using microinjection-dye transfer assay of junctional communication between Chinese hamster V79 cells.