Whole-animal models are still the gold standard for toxicity predictions and are required by regulatory agencies, yet for many biological pathways and mechanisms they do not provide a good extrapolation to humans.
In vitro surrogate assays provide more representative methods to rule out adverse effects early in the screening process. Presentations at the recent “ETS-2012: 10th International Conference on Early Toxicity Screening” focused on emerging trends.
Dedifferentiated HepG2 cells, originally derived from the liver, do not have liver-like transporter function or metabolism. However, this cell line offers a cheap methodology to answer qualitative questions, and to quantitatively compare the potency and efficacy of a broad spectrum of chemicals early in the process.
Life Technologies’ cryopreserved HepaRG™ cells are a new alternative. Terminally differentiated hepatic cells derived from a single human donor, liver progenitor cell line, HepaRG cells model an analogous biology that appears to be comparable to primary human hepatocytes.
These cells retain many primary human hepatocyte characteristics, such as morphology and expression of key metabolic enzymes, drug transporters, nuclear receptors including support of the CAR pathway, in addition to primary hepatocyte-like P450 activities.
“We believe this is the first hepatic cell line that supports ‘mature’ hepatic phenotypes analogous to cultured primary human hepatocytes, a clear step above HepG2 cells and other reported hepatic cell lines,” explained Stephen Ferguson, Ph.D., associate director R&D, Life Technologies.
“It may be possible to further improve predictivity with various in vitro hepatic models by incorporation of human primary Kupffer cells to model diseased/inflamed liver phenotypes in vitro. This model may be more reflective of a patient sub-population that has compromised liver function with diminished ability to metabolize and clear drugs.”
Early-stage research studies performed on primary hepatocytes and primary hepatocyte/Kupffer cell co-cultures suggest that Kupffer cells modulate metabolic activity, leading to altered toxicity profiling.
Also phagocytic, Kupffer cells potentially could be used as therapeutic targets, as well as to study cytokine signaling, chemical effects, and the interactions between these cells and hepatocytes.
“A comprehensive predictive in vitro hepatic model system does not exist right now. However, we can validate models that answer specific questions. As the use of in vitro hepatotoxicity screening increases, consensus will grow around the model systems,” concluded Dr. Ferguson.
Modeling Transport Function
The most important determinant of a drug’s effect in hepatocytes is the intracellular concentration, which is a function of the balance between drug uptake and efflux. Since most drugs are substrates or inhibitors of multiple transporters, a cell-based model with in vivo relevant transporters is critical.
“Culturing hepatocytes on collagen with an overlay of collagen, a sandwich configuration, results in hepatocytes with more in vivo-like characteristics. Under appropriate culture conditions, hepatic transport proteins are expressed, localized, and function similar to in vivo. The overlay of collagen allows the hepatocytes to repolarize and form functioning bile pockets where the hepatocytes touch,” said Kenneth Brouwer, Ph.D., CSO, Qualyst Transport Solutions.
Bile pockets are isolated from the media by “tight junctions”, which are controlled by calcium levels. A drug can only get into the bile pockets if it is taken up into the hepatocyte and excreted into the bile pocket.
“To evaluate a drug’s potential to cause hepatotoxicity, the in vitro model must generate an in vivo relevant intracellular concentration. All of the transporters identified in vivo are expressed in Qualyst’s B-CLEAR system and localized on the correct domain. Transporters on the cannicular, basolateral, or blood domain in vivo are on the same domain in vitro and functioning like in vivo.
“A direct in vivo to in vitro correlation has been demonstrated using compounds that demonstrated transporter problems in the clinic,” continued Dr. Brouwer. “Internal certification protocols provide functional information on both uptake and efflux transporters for B-CLEAR transporter-certified hepatocytes.
“Understanding interaction mechanisms is difficult to do in in vivo models. You need to take it down one level of complexity and sandwich-cultured hepatocytes seem to be the sweet spot for answering transporter-type questions. Most other models look at one transporter at a time; we look at the integrated effects of all transporters.”
Expanding Assay Applications
Developed in the 1960s, hematopoietic colony-forming unit (CFU) assays are used to measure the number of blood-forming progenitor cells in bone marrow or blood cell grafts prior to transplantation. This cell assay allows the identification and quantitation of the rare clonogenic cells that give rise to all of the different types of mature blood cells. Now, the assay is also used for hematotoxicity testing.
“For hematotoxicity assays, the trend is to move toward primary cells,” commented Jackie Damen, Ph.D., director of contract assay services, Stemcell Technologies.
“With human cells, when proliferation starts to slow and differentiation begins, a whole host of proteins become more predominant in the cell; that is a prime time for toxicity. Today, new drugs are more targeted; the advantage of using primary cells is that the protein milieu is similar.
“Animal testing is still used, but can be minimized by using representative, alternative in vitro assays. We typically use human cells and compare results to animal models to predict a maximum tolerable human dose and to help determine the animal model’s relevance.”
Stemcell Technologies is developing an automated counter for hematopoietic colony assays. Introduced first to cord blood banks, the company plans to expand the application to toxicity screening. Certain drugs will alter the size and morphology of colonies before reducing colony numbers in the CFU assay. The automated system allows such changes to be detected with high sensitivity.
“Historically, the CFU assay has been used to detect primitive hematopoietic cells in different human tissues. The number of CFUs in a cell sample is the measurement that correlates most strongly with engraftment after transplantation. This is also true when looking at drug toxicity, because the assay measures effects on the most important cell type and their ability to do what those primitive cells are supposed to do: proliferate and differentiate into the right lineage,” concluded Dr. Damen.
Analyzing More with Less
Specifically developed for blood stem and progenitor cells, HemoGenix’ HALO assay analyzes the effect of large numbers of compounds on a small number of cells. Instrument-based, standardized and validated, HALO is one of the few stem-cell high-throughput assays—up to 384-well.
A metabolic assay, HALO is based on changes that occur within the cells that are measured using intracellular ATP (iATP), a biochemical marker for chemical energy and mitochondrial integrity. The cellular effects of different molecules are proportional to the amount of iATP production. Similar assays for mesenchymal or multipotent stromal stem cells use the same iATP read-out.
Different cell types from different species can be directly compared to look at the effect of different compounds; although culture environments vary, the read-out is exactly the same for each cell type.
“The important thing is to use an in vitro assay that provides a direct extrapolation to the human system; to use primary human cells as the target,” discussed Ivan Rich, Ph.D., founder and CEO, HemoGenix.
The drawback is that primary human cells are supply-limited and expensive. To produce larger quantities of cells at lower cost, embryonic stem (ES) and induced pluripotent stem (iPS) cells are being used to generate mature cells thought to be similar to functional mature cells found in the body.
“It is just too early to make any definitive statement or conclusion. The primary cells need to be characterized first and compared to cells generated from ES and iPS cells. Without this characterization, it is unknown if ES and iPS cells can replace primary cells from a specific organ and give the same response. Using them routinely for drug screening could provide inaccurate information,” continued Dr. Rich.
“For example, a fibroblast could be reprogramed into an iPS cell, and then induced into a different cell type, such as a hepatocyte. Is the genetic make-up that of the hepatocyte or does the cell retain some of the original fibroblast genome?”
A Promising Way Forward
“Although neuron development is a tightly orchestrated, complex process, the plasticity of the nervous system is one of the biggest developmental neurotoxicity problems we grapple with,” explained Pam Lein, Ph.D., professor, department of molecular biosciences, University of California at Davis.
“We understand the major processes involved in the development of the nervous system but we do not have good testing paradigms in place. This is a prime time to develop techniques before legislation and regulatory actions take place using the old way of doing things.”
In vitro and alternative systems-based models, such as the zebrafish and C. elegans, have been used in the developmental neurobiology field for some time. The significant change is not the models but the methods being applied to measure the endpoints of interest, such as high-content analysis and high-throughput systems (HTS).
HTS are being developed for detecting fluorescent signals and cellular morphology in cultured neurons and alternative systems-based models. Even behavioral analyses in C. elegans and zebrafish are being automated with the goal of increasing throughput. More recent is the development of microelectrode arrays, specialized culture systems in which microelectrodes are embedded in the substrate. As the neurons begin to form functional networks and communicate with each other, electronic signals are detected, decoded, and read as endpoints.
By looking at multiple endpoints in the same model system, scientists hope to discover novel insights regarding potential mechanisms of action that will facilitate extrapolation across species, and will identify the most sensitive endpoint with respect to the chemical.
The goal is to use alternative models to rule out compounds early for adverse effects, but many variables cannot be taken into account with these approaches. Confirmation still needs to be accomplished in a whole-animal model.
“The molecular- and mechanistic-level understanding of how compounds cause developmental neurotoxicity is still a bit of a black box. Alternative cell- and systems-based models are a promising way forward,” concluded Dr. Lein.