Stem Cells and Toxicity Testing, Part II: models and applications

Home / In the Spotlight / Stem Cells and Toxicity Testing, Part II: models and applications

In the Spotlight

Stem Cells and Toxicity Testing, Part II: models and applications

Sherry L. Ward, AltTox Contributing Editor

Published: September 15, 2013

(This is Part II of a three-part series; see also Part I and Part III.)

The isolation and growth of the first human embryonic stem cell line (hESC) in 1998 created a new world of opportunities for biomedical researchers. In reality, however, ethical and technical hurdles in procuring and using hESCs have continued to stymie progress. The creation of human induced pluripotent stem cells (hiPSC) from human skin cells in 2007 greatly expanded the opportunities for researchers. The use of hiPSCs overcomes the ethical concerns of destroying human embryos. Technical challenges remain in creating, maintaining, characterizing, and differentiating hiPSCs, but researchers are tackling these with astounding progress.

Part I of Stem Cells and Toxicity Testing reviewed a portion of the brief but explosive history of stem cell discovery and related scientific breakthroughs. A “Stem Cell Primer” that explains stem cell nomenclature and terminology was included.

Part II of Stem Cells and Toxicity Testing will briefly review some of the current and potential uses of stem cells in the field of toxicity testing.

Stem Cell Models for Reproductive Toxicity Testing

Industries that are required to conduct toxicity testing for regulatory purposes are interested in replacement and screening methods for human health endpoints like reproductive toxicity that are time consuming and expensive to conduct in animal models. A great deal of research on alternative methods for reproductive/developmental toxicity has occurred in the past decade due to the impact of two European regulations: the EU ban of animal testing for cosmetics under the Cosmetics Directive, and the large testing burden of the REACH regulation. Even so, a full replacement method or testing scheme has yet to be endorsed by a validation or regulatory authority.

Dr. Pauling would be pleased

California’s stem cell agency, the California Institute for Regenerative Medicine, reported in July that “CIRM-funded researchers…discovered a surprising role for vitamin C in how stem cells turn genes on and off” (CIRM, 2013). Tet enzymes were known to mediate DNA methylation in the early embryo and germ cell lines, and vitamin C had previously been linked to epigenetic changes in hESCs (Chung et al., 2010). The new findings, published in the journal Nature (Blaschke, 2013), showed that vitamin C increased Tet enzyme activity, which increased demethylation of certain genes, increasing their activity in the stem cells. The researchers concluded that vitamin C acts “as a direct regulator of Tet activity and DNA methylation fidelity in ES cells.” Thus, vitamin C protects germline cells from DNA methylation. The addition of methyl groups to DNA, or DNA methylation, is a heritable epigenomic modification implicated in diseases such as cancer, diabetes, and obesity.Illustration showing DNA methylated at two sitesIllustration showing DNA methylated at two sites (Source:

Alternative methods for reproductive toxicity testing face the major hurdle of having to assess potential adverse effects over the various stages of reproduction/development (Spielmann, 2005). The critical toxicity targets at each stage need to be identified and modeled in one or more alternative test system(s). Stem cell models for reproductive toxicity are perfectly suited for the assessment of toxic effects on early embryo development, or embryotoxicity. Krtolica, et al. (2009), who advocate for the use of human stem cells for reproductive toxicity studies, reported that “the early-dividing embryo is very sensitive to numerous factors present in its microenvironment.”

The mouse Embryonic Stem Cell test, developed in 1990’s by Spielmann, et al., is one of three ECVAM-validated alternative methods that can be used as a screening assay for reproductive toxicity testing (Genschow et al., 2002). The EST predicts in vitro embryotoxicity using two cell lines and the following three endpoints: inhibition of differentiation of mouse ESCs into beating cardiomyocytes, cytotoxic effects on mouse ESCs, and cytotoxic effects on mouse 3T3 fibroblasts. The original EST method had some limitations and was technically difficult to perform, so additional studies were conducted to improve the EST. Protocol modifications have been implemented to enhance the performance of the EST (Marx-Stoetling et al., 2009). It was proposed that differentiation into cardiomyocytes alone is not sufficient to assess all embryotoxic events, so additional endpoints expressed as the ESCs differentiate toward other cell lineages, such as bone or neurons, have been developed. More quantitative ways to assess differentiation, such as marker protein expression, can replace the microscopic evaluation of beating cardiomyocytes. In 2011, Seiler and Spielmann published the ECVAM-validated protocol as well as a new protocol for the molecular-based FACS-EST method, which uses “highly predictive protein markers specific for developing heart tissue.” The predictive capability of the EST method continues to be assessed and refined, such as in the recent study of Dreisig, et al. (2013) where the predictive value of the EST assay was enhanced when considered along with the data from a panel of other cell-based assays.

Standing on past achievements

The vitamin C protection of embryonic stem cell genes (see previous text box) is a long-evolving tale.

Dr. Linus Pauling began studying the effects of vitamin C on health and disease in the 1960s. He defined the new filed of orthomolecular medicine as “the preservation of good health and the treatment of disease by varying the concentrations in the human body of substances that are normally present in the body” (The Linus Pauling Institute, 1999).

Dr. Horst Spielmann, developer of the EST assay (see previous paragraph), also made early observations on the effect of ascorbic acid (vitamin C) on mouse embryos exposed to the genotoxic agent cyclophosphamide (CPA) (Vogel & Spielmann, 1989; Kola et al., 1989): ascorbic acid seemed to “protect early embryos against damage induced by genotoxic agents like CPA.”

We now know, thanks to the new field of epigenetics research, that cyclophosphamide exposure “induces aberrant epigenetic programming in early embryos” (Barton et al., 2005). And we now know (from research discussed in previous text box) the mechanism of how vitamin C mitigates the genetic damage, thus protecting the early embryo.

Even the greatest visionaries in bioscience research are limited by the technologies and methods available at that time in history. It is certainly interesting and respectful to look back to see how we arrived at our current state of knowledge, as one day we will be that history.

The European project, ReProTect (2005-2009), was set up to develop a tiered testing strategy based on a combination of in vitro/in silico methods for assessing reproductive and developmental toxicity. In the 5th year, a ring trial on 10 chemicals in a battery of 14 in vitro assays was conducted. Project researchers concluded that the results provided “a robust prediction of adverse effects on fertility and embryonic development of the 10 test chemicals” (Schenk et al., 2010). The EST assay was an important component of the ReProTect project. Outcomes for the EST included the decisions to expand the chemicals database to better define the EST applicability domain, to obtain a better mechanistic understanding of the assay, and to continue to optimize the assay. The effect of metabolic activation of test substances was not determined.

Reproductive toxicity testing is also part of the novel toxicity test platform based on human embryonic stem cells (hESCs) being developed by the EU FP7 research project, Embryonic Stem cell-based Novel Alternative Testing Strategies or ESNATS. Consortium members are developing a battery of toxicity tests using standardized hESC lines and protocols for reproductive toxicity, neurotoxicity, toxicogenomics and toxicoproteomics, and metabolism and toxicokinetics that “will finally be integrated into an ‘all-in-one’ test system.” Krug, et al. (2013) evaluated transcriptomics endpoints in five hESC-based systems developed under the ESNATS project that recapitulate different phases of early neuronal development, and concluded that this assay battery “allows classification of human DNT/RT [developmental neurotoxicity/reproductive toxicity] toxicants on the basis of their transcriptome profiles.” Additional publications resulting from the ESNATS project are listed on their website, and the final ESNATS Conference, Use of Human Embryonic Stem Cells for Novel Toxicity Testing Approaches, will be held within the European Society for Alternatives to Animal Testing (EUSAAT) 2013 Congress in Linz, Austria, on September 16, 2013. The ESNATS Conference will cover the major achievements of the five-year project, application of the ESNATS assays, and proposed further research; Dr. Horst Spielmann will give the opening lecture.

The US EPA’s ToxCast computational toxicology program was established in 2007 as “part of the federal Tox21 consortium to develop a cost-effective approach for efficiently prioritizing the toxicity testing of thousands of chemicals” (Knudsen et al., 2013). The ToxCast approach is to test a chemical library in a large number of in vitro assays adapted for high-throughput screening (HTS). In the first phase of ToxCast, 309 chemicals with substantial in vivo data were evaluated. This testing included using the EST assay modified for HTS with mouse ESCs. Perturbed transcription of certain developmental genes was associated with impaired stem cell differentiation (Chandler et al., 2011). Further research to identify the metabolic and regulatory pathways of human ESCs disrupted by chemicals is underway (Kleinstreuer et al., 2011). The Phase I in vitro data (ToxCast) and in vivo database (ToxRefDB) are available on the EPA website.

Industry has shown great interest in the EST, and many companies are using it as an in-house screen for reproductive toxicity. Companies and other researchers are also adapting the EST to HTS, evaluating the use of human ESCs, differentiating the ESCs to other tissues such as bone and neural cells, and using the EST as part of a tiered or integrated testing scheme for better decision making. Some researchers, including Adler, et al. (2008), Krtolica and Giritharan (2010), Krtolica, et al. (2009), and West, et al. (2010), have concluded that it is important to use human ESCs (rather than mouse ESCs) for human health testing applications.

Stem Cell Pre-clinical Models

Pharmaceutical firms are developing human induced pluripotent stem cells as pre-clinical models for human toxicity, most commonly as screens for cardiotoxicity, hepatotoxicity, nephrotoxicity, and neurotoxicity. “Stem cells are expected to dramatically improve the ability of drug companies to screen for side effects of new drugs much earlier in the development process…” (CIRM, 2013). Undifferentiated stem cells can be maintained over time and used to develop the large quantities of different cell types needed for testing. Pre-clinical test results using hiPSCs will also correspond to human biology, thus avoiding the potential species-differences encountered with the use of animal models. ESCs and iPSCs have been adapted to high-throughput and/or high-content screening platforms (Sherman et al., 2011), which provides the capability to screen many compounds for a variety of mechanistic and toxicity parameters. Scott, et al. (2013), however, concluded that “significant improvements in hiPSC production and differentiation processes are required before cell-based toxicity assays that accurately reflect mature tissue phenotypes can be delivered and implemented in a cost-effective manner.”

There have been some interesting successes in the development of stem cell models for cardiotoxicity testing (Sinnecker et al., 2012; Kraushaar et al., 2012; Mordwinkin et al., 2013; Mercola et al., 2013). Human iPSC-derived cells have been differentiated into functional cardiac myocytes as a model for drug-induced QT interval prolongation (long-QT syndrome) (Moretti et al., 2010; Sinnecker et al., 2013).  Liang, et al., (2013) generated a library of human iPSC-derived cardiomyocytes from healthy patients and those with various genetic-related cardiac diseases, and verified the disease phenotypes in the cultured iPSCs as well as their different susceptibilities to cardiotoxic drugs. One of the authors in this study commented that “the use of patient-specific stem cells to detect cardiotoxic properties of pharmaceutical compounds may be more accurate than the current drug-safety assays mandated by the FDA” (Conger, 2013). The stem cell assay for drug-induced QT interval prolongation would replace another in vitro method, and so may be less problematic for regulatory acceptance than replacing an animal test.

Clinical Trial in a Dish

In vitro clinical trials” have also been proposed “in which iPS cells derived from a wide variety of individuals could be used to predict patients’ response to a drug” (Dimond, 2012). Promising cardiotoxicity test results achieved with an iPSC model were described as a “springboard to Phase 1 clinical trials” (Iacone & Anson, 2013). In addition to this being a more species-specific model than current animal-based pre-clinical testing, it would have the additional benefit of providing wide human population diversity without the risk of current early clinical trials. Both normal and disease-specific stem-cell derived models would be included to provide early efficacy as well as toxicology assessments. Since stem cells can be directed to differentiate into any tissue, once the many technical hurdles are resolved, this long-term goal is a promising approach that could significantly reduce animal testing in drug development, while improving human safety and reducing post-marketed drug attrition.

Bioengineered Human Tissue Models

Bioengineered human tissue constructs have the potential to serve as species- and tissue-specific in vitro models to replace animal models in some types of toxicity testing. In vitro three-dimensional (3D) models that replicate the physiology of in vivo human tissues/organs would provide more relevant information on the human safety/toxicity of drugs and chemicals than do the simpler cell-based models.

Tissue-engineered constructs are being used as pre-clinical models for testing drugs, gene therapies, and medical devices, as described in the recent review article of Gibbons, et al. (2013). They identified pre-clinical testing in the following types of bioengineered human tissues: “blood vessels, skeletal muscle, bone, cartilage, skin, cardiac muscle, liver, cornea, reproductive tissues, adipose, small intestine, neural tissue, and kidney.” The models reviewed by Gibbons were not limited to those that incorporated stem cells. However, substantial progress has been made in the development of bioengineered human tissues that incorporate stem cells, including those for bone, liver, cornea, musculoskeletal, bronchial, and hypertrophic scar. Many of these bioengineered human tissues are being developed for transplantation purposes, but could be repurposed for toxicology testing applications.

Except for an avascular tissue such as the cornea, or a very small tissue, a functioning vasculature (blood vessels) is needed to supply in vitro tissues/organs with nutrients and oxygen. In toxicological studies, additional issues such as metabolism and toxicokinetics often need to be assessed. These issues as well as a brief review of some of the bioengineered human tissues under development that are using human stem cell technologies will be the focus of the upcoming Part III of Stem Cells and Toxicity Testing.

Parting Thoughts

Human stem cell-derived in vitro models hold great potential for the development of biologically relevant models for evaluating the toxicity of substances to humans. Due to the concurrent and more lucrative demand for stem cell technologies in the fields of regenerative medicine and drug development, scientific discoveries are occurring at a rapid pace and technological obstacles are being resolved. Now is an appropriate time to more fully apply this emerging technology to the development of human cell and tissue-based models for assessing human health endpoints in toxicological studies.