Use of High Throughput Assays and Computational Tools in the Endocrine Disruptor Screening Program
Learn more about the future of the Endocrine Disruptor Screening Program (EDSP):
The ability to screen chemicals rapidly for bioactivity in several endocrine pathways, and reducing the use of animals in testing, have been EDSP goals since 1998, when the program was first adopted.
As previously noted, when the first Tier 1 orders (for List 1 chemicals) were issued in 2009, EPA had not confirmed the reliability and relevance of the ToxCastTM results so that they could be cited as “other scientifically relevant information” to satisfy the Tier 1 order. However, since that time, EPA has reached a critical juncture, determining that the science has progressed such that reevaluation of EPA's earlier position is warranted.
Based on scientific advances, EPA intends to implement the use of high throughput screening assays and computational models to evaluate, and to a significant extent, screen chemicals. The in vitro high throughput and computational model alternatives provide an accurate quantitative measure of specific endocrine receptor binding bioactivity and mechanisms that can serve as alternatives to the current Tier 1 estrogen receptor (ER) binding, ER transactivation (ERTA), uterotrophic assays and androgen receptor (AR) binding in vitro assay.
Learn more about TOX21 and the future of evaluating chemical safety with this short video.
High throughput assays are automated methods that allow for a large number of chemicals to be rapidly evaluated for a specific type of bioactivity at the molecular or cellular level. This approach, which can help identify compounds that may modulate specific biological pathways, was initially developed by pharmaceutical companies for drug discovery. The results of these methods provide an initial understanding of a biochemical interaction or possible role of a chemical in a given biological process.
In vitro high throughput assays are usually conducted using a microtiter plate: a plate containing a grid with a large number of small divots called “wells.” The wells contain chemical and/or biological substrate (e.g., living cells or proteins). Depending on the nature of the experiment, changes can be detected (e.g., color, fluorescence, etc.) when the chemical is added to indicate whether there is bioactivity. High throughput microtiter plates typically come in multiples of 96 wells (96, 384, or 1536), so that through the use of robotics, data processing and control software, liquid handling devices, and sensitive detection methods, an extremely large number of chemicals can be evaluated very efficiently.
High throughput assays can be run for a range of test chemical concentrations and produce concentration-response information representing the relationship between chemical concentration and bioactivity. The concentration-response data from multiple assays can be mathematically integrated in a computational model of a biological pathway, providing values representative of a chemical's bioactivity in that pathway (e.g., estrogen receptor pathway).
To reduce non-specific results, the computational model can use results from multiple assays and technologies to predict whether a chemical is truly bioactive in the pathway being evaluated. The most prominent cause of non-specific results (activity in an assay that is likely not due to bioactivity of the chemical in the pathways) is cytotoxicity in cell-based assays. In other cases, chemicals influence the assays through a manner dependent on the physics and chemistry of the technology platform (i.e.,“assay interference”).
EPA is adopting in vitro high throughput assays and computational models for detecting and ER and AR agonist and antagonist bioactivity as alternatives for three current Tier 1 assays:
- ER binding in vitro assay;
- ER transcriptional activation in vitro assay;
- In vivo Uterotrophic assay; and
- AR binding in vitro assay.
EPA will be accepting existing results for chemicals that have been evaluated using the ToxCastTM “ER Model” for bioactivity. The accompanying database contains the ER and AR agonist bioactivity and ER and AR antagonist bioactivity for over 1800 chemicals and identifies those chemicals that are pesticide active ingredients, pesticide inert ingredients, and on EDSP Lists 1 or 2. This is a “living” database that will continue to incorporate bioactivity results for chemicals as they become available.
It is important, however, not to equate a determination of a chemical's bioactivity from the “ER Model” with a determination that a chemical causes endocrine disruption. The World Health Organization (WHO)/International Programme on Chemical Safety (IPCS) defines endocrine disruption as being caused by "an exogenous substance or mixture that alters function(s) of the endocrine system and consequently causes adverse health effects in an intact organism or its progeny, or (sub)populations." Bioactivity is an indicator that a chemical has the potential to alter endocrine function, but (1) whether the chemical actually alters endocrine function and (2) whether that altered function produces an adverse outcome in an intact animal cannot be determined without further testing (i.e., Tier 2 testing).
The EDSP has been developed over the past 19 years, and has demonstrated that the current screening process may take upwards of 5 years before a Tier 1 decision is available or Tier 2 test orders are issued. In light of recent advances in high throughput assays and computational models, and advances likely to come in the next two years, it is prudent for the Agency to consider new, rapid screening methods.
The availability of additional alternative high throughput assays and computational models in the near term will allow EPA to screen more chemicals in less time, involve fewer animals, and cost less for everyone. Furthermore, reconsideration of the EDSP List 2 chemicals may be appropriate since “ER Model” data are available for many List 2 and other chemicals. Ongoing use of high throughput screening assays and computational models will address thousands of chemicals in the future.
These advancements in the EDSP screening program will not affect the overall framework—i.e., the Tier 1 screening battery and Tier 2 testing approach focused on estrogen, androgen and thyroid pathways in humans and wildlife remains unaffected. Instead, as discussed above, EPA is planning to adopt sensitive, specific, quantitative, and efficient screening methods that will rapidly screen many chemicals and substantially decrease costs and animal use and may be used as an alternative to some EDSP Tier 1 screening assays.
Accordingly, EPA intends a future recipient of an EDSP test order to be able to satisfy the screening requirement for ER, ERTA, and uterotrophic in one of three ways: (1) cite existing ToxCastTM “ER Model” for bioactivity data as “other scientifically relevant information” (where available); (2) generate new data relying on the 18 ER high throughput assays and the ToxCastTM “ER Model” for bioactivity; or (3) generate their own data using the current Tier 1 ER binding, ERTA, and uterotrophic assays.
EPA believes that ongoing adoption of alternative methods and technologies will continue to advance EDSP screening of chemicals for bioactivity in the estrogen, androgen, and thyroid pathways. EPA is continuing research on the “ER Model” to determine if ToxCastTM assays can provide comparable information as that of the Female Rat Pubertal and the Fish Short Term Reproduction assays.
In addition, research continues on the ToxCastTM “AR Model” for bioactivity which, if fully validated, may be considered as an alternative (alone or with the “ER Model”) for the following current Tier 1 assays: AR binding, Male Rat Pubertal, Hershberger, and Fish Short Term Reproduction. Research is also underway to develop steroidogenesis ToxCastTM (STR) and thyroid (THY) bioactivity models.
Over time, the Agency's goal is to develop a set of “non-vertebrate” high throughput assays and computational bioactivity models as an alternative to all of the assays in the current Tier 1 screening battery. The following table is intended to illustrate the evolution of screening in the EDSP:
|Current EDSP Tier 1 battery of assays||Alternative high throughput assays and computational model for EDSP Tier 1 battery|
|Estrogen Receptor (ER) Binding||ER Model (Alternative).|
|Estrogen Receptor Transactivation (ERTA)||ER Model (Alternative).|
|Uterotrophic||ER Model (Alternative).|
|Female Rat Pubertal||ER, STR, and thyroid (THY) Models (Future).|
|Male Rat Pubertal||AR, STR, and THY Models (Future).|
|Androgen Receptor (AR) Binding||AR Model (Alternative).|
|Hershberger||AR Model (Future).|
|Aromatase||STR Model (Future).|
|Steroidogenesis (STR)||STR Model (Future).|
|Fish Short Term Reproduction||ER, AR, and STR Models (Future).|
|Amphibian Metamorphosis||THY Model (Future).|
The table indicates combinations of various alternative assays and models that might overlap for evaluating potential endocrine bioactivity of chemicals. The in vitro high throughput and computational model alternatives provide a focused evaluation of the mechanistic aspects of endocrine pathways, thereby providing specific and quantitative measures of bioactivity.
Several assays in the Tier 1 battery rely on intact animals and identify bioactivity in the multiple biological pathways present. For this reason, the specificity of the in vitro high throughput and computational model alternatives may be more informative of specific endocrine pathway bioactivity.