In vitro endocrine and cardiometabolic toxicity associated with artificial turf materials

Abstract

Artificial turf, a consumer product growing in usage in the United States, contains diverse chemicals, some of which are endocrine disruptive. Endocrine effects from turf material extracts have been primarily limited to one component, crumb rubber, of these multi-material products. We present in vitro bioactivities from non-weathered and weathered turf sample extracts, including multiple turf components. All weathered samples were collected from real-world turf fields. Non-weathered versus weathered differentially affected the androgen (AR), estrogen (ER), glucocorticoid (GR), and thyroid receptors (TR) in reporter bioassays. While weathered extracts more efficaciously activated peroxisome proliferator activated receptor γ (PPARγ), this did not translate to greater in vitro adipogenic potential. All turf extracts activated the aryl hydrocarbon receptor (AhR). High AhR-efficacy extracts induced modest rat cardiomyoblast toxicity in an AhR-dependent manner. Our data demonstrate potential endocrine and cardiometabolic effects from artificial turf material extracts, warranting further investigation into potential exposures and human health effects.

Keywords: Adipogenesis; Cardiotoxicity; Endocrine disruption; Leachate; Nuclear receptor; Synthetic turf.

Fig. 1.

Schematic giving the general structure of an artificial turf field cross-section. Typical components include (i) plastic fibers and backing, (ii) infill, usually crumb rubber-derived, (iii) synthetic support, (iv) concrete, and (v) natural dirt beneath.

Fig. 2.

Turf extract activation bioactivities on steroid and thyroid hormone receptors. Heat map depicts effects of non-weathered or weathered extracts on receptor activation. Only statistically significant (p < 0.05, Kruskal-Wallis + Dunn post-hoc from vehicle) changes are visualized. Increasing redness indicates increasing activation. Extracts are organized by type of sample (i.e., fibers, infill). Nanoluciferase decreases proxying toxicity (> 15% decrease compared to vehicle) are indicated by light stripes on a panel. In-panel values are the mean ± SEM of at least 3 independent experiments.

Fig. 3.

Turf extract augmentation or inhibition bioactivities on steroid and thyroid hormone receptors. Heat map depicts effects of non-weathered or weathered extracts on receptor antagonism (augmentation or inhibition). Only statistically significant (p < 0.05, Kruskal-Wallis + Dunn post-hoc from vehicle) changes are visualized. Increasing redness indicates increasing augmentation, increasing blueness indicates greater inhibition. Extracts are organized by type of sample (i.e., fibers, infill). Nanoluciferase decreases proxying toxicity (> 15% decrease compared to EC₅₀ exposed cells) are indicated by light stripes on a panel; panels without stripes do not In-panel values are the mean ± SEM of at least 3 independent experiments.

Fig. 4.

PPARγ and adipogenic activities of artificial turf extracts. Reporter PPARγ bioactivities for non-weathered (A) and weathered (B) turf extracts. (C-F) Human mesenchymal stem cells were induced to differentiate as described in Methods. PPARγ activity (A-B) and triglyceride accumulation (C-D) are presented as percent of maximal rosiglitazone-induced effects and normalized to DNA content. (E-F) Proliferation/cytotoxicity is provided as percent increase or decrease in DNA content relative to differentiated vehicle controls. Asterisks (*) are color-coded to each chemical/extract and denote statistical difference (p < 0.05) between assay baseline and chemical exposure at the given concentration.

Fig. 5.

Spearman correlation matrices for turf extract lipogenesis and human nuclear receptor bioactivities. Values are aggregated by type of turf, non-weathered (A) or weathered (B). Increasing redness suggests positive correlations; increasing blueness suggests negative. Only statistically significant (p < 0.05) relationships are visualized.

Fig. 6.

Turf extract effects on AhR activity and H9c2 phenotypes. Luminescent reporter AhR bioactivities for non-weathered (A) and weathered (B) turf extracts. (C) H9c2 rat cardiomyoblast viability following acute (24 hour) single turf exposure or turf and IC₅₀ GNF351 (AhR antagonist). Statistical significance determined by Kruskal-Wallis test and Dunn post-hoc (p < 0.05). * denotes significant change from vehicle. # denotes significant change between single and co-exposures.

Fig. 7.

Spearman correlation matrices for (A) NJ-1 and (B) NJ-2 changes in H9c2 viability and NR bioactivities. Statistically significant correlation tiles are filled; non-significant correlation tiles are blank. Increasing redness indicates a positive correlation between two parameters, while increasing blueness indicates a negative correlation.