Are stem cells ready as a next generation drug discovery tool?
Posted on: December 17, 2010
In recent years, there is a considerable push in adopting stem cell based assays as a drug discovery tool. The major argument behind this initiative is stem cells can differentiate into specific cell types, which can be used for targeted drug screening. Major pharmaceutical companies and public funded academic institutes have been started investing significant amount of effort and money on stem cell based drug discovery with the hope that this approach will ultimately provide real breakthroughs in drug discovery research as well as significant cost savings. There is no doubt that stem cells may offer unique opportunities in drug discovery, but the scientific data to support this notion may not yet well established to the extent that is applicable to drug discovery screening.
A most popular application of stem cells is the use of stem cell derived cardiomyocytes as a screening tool for testing cardiac toxicity of drugs. One can argue that there is enough scientific evidence to prove that stem cell differentiated cells are genetically similar to corresponding normal cells and can be used as a drug screening tool. This may or may not be true based on answers to following questions. a) can stem cells be a better tool in drug screening assays than currently used cell/animal based assays? b) are stem cell derived cell/tissue types are genetically and physiologically identical to the desired natural cell types, for example cardiomyocytes? c) are these proposed stem cell based assays will provide better tools to reduce or eliminate fatal side effects of drugs, e.g. cardiotoxicity? It is important to note that current use of cell based assays (non-stem cell based) failed to reduce or prevent fatal side effects of drugs. Based on this, it is hard to argue that currently proposed stem cell based assays provide any better solution to current problems. We need to understand genetic and physiological make up of stem cell derived cell types before we adopt this technology for making life saving decisions. In any means, our opinion does not imply that stem cell based products are not suitable for developing drug discovery assays. Our goal is to critically analyze the scientific rational behind current approaches.
Currently used confirmatory tests are not sufficient enough to establish the use of stem cells in drug discovery screening.
We will analyze stem cell based cardiomyocytes, which is one of the most “publicized” applications of stem cell based drug assay, as an example to understand whether stem cell derived products can be used in drug discovery screening assays. We would like to raise one question. Are we in a position to predict confidently that these cardiomyocytes, generated from stem cells have genomic stability? This means, stem cell derived cardiomyocytes do not posses genetic and somatic mutations, DNA polymorphism, chromosomal translocations, genomic instability due to polyploidy or anueploidy, miRNA polymorphism, metabolite variations and proteome polymorphism/modifications. These cellular or genetic changes can result from long-term cell culture system, molecules or genetic modifications that are used for the induction of differentiation and induced genetic rearrangements that are needed for the generation of stem cells, especially iPS cells, which may have some tumor cell properties. Mere confirmation of stem cell derived cardiomyocytes using DNA microarrays is not enough to establish the fact that stem cell derived cell types are genetically similar to normal cardiac cells/tissues. DNA microarray based approach does not address or identify unknown genes derived from alternative splicing/transposons, gene modifications and polymorphism, polyploidy or anueploidy, RNA modifications and epigenetic/mitochondrial DNA modifications. Therefore, the use of current DNA microarray based screening may be insufficient to establish the validity of stem cell derived products.
In addition to genes, miRNA, metabolites and proteins play significant role in genetic and physiological regulation in a particular cell type. A single change in post-translational modification in a protein can drastically change the cellular signaling process, which cannot be detected using DNA microarrays. This is also true with miRNAs and metabolites. The currently used stem cell derived products such as cardiomyocytes need to be established as true cardiac cells, both genetically and physiologically. This requires extensive scientific research. Until then, stem cell based assays for screening life threatening side effects such as cardiotoxicity should be carefully considered for drug development applications. There is a need for the development of reliable assays or technologies for the detection of genomic, proteomic, metabolomic, epigenetic and physiological instabilities in stem cell derived cell or tissue types.
Establishing a regulatory system for drug discovery assays will help in developing new drugs with fewer or no fatal side effects.
The development of new drugs based on stem cell derived assays, with very limited in-depth understanding of the system, needs to be reviewed diligently. How can we confirm that genetic or physiological instability of stem cell derived cardiomyocytes does not affect assay outcome, which is very critical? Can companies who develop such products, for example cardiomyocytes, guarantee that drugs developed using their stem cell technology do not have fatal cardiac side effects?. Under current system, anyone can aggressively market their products/technologies as a drug discovery tool based upon incomplete data or information because of the lack of regulation in this area. There are numerous examples where several products are marketed for drug discovery research without disclosing possible product limitations (see our earlier blog on cell based reporter assays). This is a major issue especially with proprietary/patented technologies owned by assay development companies or CROs. A very good example for this is reporter enzyme, homogeneous and cell based assays. Marketing proprietary technologies for all possible applications, without disclosing limitations, needs to be critically evaluated. So many public funded laboratories and facilities may also be using technologies, which were marketed to these customers with or without disclosing or foreseeing product limitations. It is important to note that consumers are protected from undesirable claims on pharmaceutical drugs, diagnostics and consumer products. Establishing an oversight regulatory system will help in checking product performance claims. Generating scientific and technical accountability monitoring system that starts from the bottom (e.g. reagent companies, CROs etc.) to the top (pharmaceutical companies) will help in generating robust pipeline of safer drugs. Any failure in a given drug should be scrutinized from the drug discovery research to development phase. Creating an oversight system is warranted especially for developing and marketing drug screening assays, similar to diagnostics products. This approach may slow-down development and application of drug discovery assays. However, in the long-run this will have significant positive impact on drug discovery research and development. This will help in selecting the right product for specific applications with complete understanding of limitations, which will help in developing alternative strategies. Finally, it is the choice of drug discovery researchers to address the need for a regulatory system for assays that are used in critical areas of drug discovery and development. These critical areas can be drug metabolism (ADME), cardiotoxicity, other organ/tissue specific toxicity, mutagenicity, and immunogenicity.
8 Responses to "Are stem cells ready as a next generation drug discovery tool?"
[…] promote such lines to the market hoping about BigPharma interest. The recent critical essay – Are stem cells ready as a next generation drug discovery tool?, published on SciClips blog challenged the reliability of current stem cell lines in development […]
December 22, 2010 at 2:47 pm
This article raises a very relevant question, namely, “Are stem cells ready for ‘prime-time’ in drug discovery or toxicity testing?”. I believe the answer is conditionally “No!”. The fundamental reasons are that , until, one understands (a) the generic limitations of all in vitro assays to generate results that can be extrapolated to the in vivo situation; (b) the basic biology of stem cells ( embryonic/ organ-specific adult); (c) the basic mechanisms of how radiation, chemicals affect the choices a stem cell makes on the receipt of their triggered intra-cellular signals [cell proliferation; differentiation; necrosis/apoptosis; senescence, etc.); (d) the mechanisms of toxicity; (e) pathogenesis of various disease outcomes; (f) how the mechanisms of toxicity ( mutagenesis; cell death; and epigenetic alteration of gene expression) might contribute to the pathogenesis of a disease endpoint; (g) timing of exposures and (h) how these factors might contribute to extrapolation to unique genetic, developmental, gender, dietary, pharmaceutical, life style and circadian state of the individual exposed to the drug.
Together with the above, understanding that alteration of both the quantity & quality of the stem cells can influence any stem cell-based disease state ( i.e., cancer), use of in vitro stem cell data, grown under in vivo niche conditions( low oxygen, correct extra-cellular matrices, etc.), could generate useful data. That this fundamental question has been raised, it should drawn more attention to this potential useful new application of stem cells. This something, I have have been addressing at multiple conferences and publications [ see. J.E. Trosko and C.C. Chang, Factors to consider in the use of stem cells for pharmaceutical drug development and for chemical safety assessment. Toxicology 270: 18-34, 2010; J. E. Trosko, ” Role of diet and nutrition on the alteration of the quality and quantity of stem cells in human aging and diseases of aging. Curr. Pharmaceut. Discovery 14: 2707-2718, 2008).
While I understand the rush to use these human stem cells for both drug discovery and toxicity assessment, I believe following the dictum that “one ought not apply that which one does not understand ” should support strong support for more basic scientific studies of stem cell biology.
James E. Trosko, Ph.D.
Center for Integrative Toxicology, Dept. Pediatrics/Human Development, College of Human Medicine, Michigan State University, East Lansing, Michigan
December 26, 2010 at 12:47 pm
Let us not forget that drug discovery and toxicology are things that we need today. I absolutely support more basic scientific research into stem cell biology, but while this happens, the best is the enemy of the good… We could already have large benefits from assays that qualifiy as the Sherman tanks of toxicology: Not as good as it should be, perhaps; maybe not even as good as some of the competing technologies; but informative enough to be useful and affordable enough to be available in numbers. It could already reduce our consumption of laboratory animals, help us filter lead compounds more accurately, and perhaps ease the task of environmental agencies that now have to re-assess thousands of substances.
If we should not apply that which we do not understand, we will, at the current state of biological knowledge, do nothing at all. Even our understanding of the simplest enzymes is flawed.
December 27, 2010 at 11:08 am
We are not convinced that stem cell based assays will neither reduce the cost of drug discovery nor the consumption of laboratory animals. For stem cell based assays, one has to use assays that are currently used for normal human cell based assays. The only difference will be the use of stem cells rather than normal human cell lines. Stem cells may not be cheaper than normal human cell lines. The other argument is the use of stem cell derived cell types such as cardiomyocytes will reduce the consumption of laboratory animals. This can be partly true if stem cell derived cardiomyocytes are identical to normal cardiac cells/tissues. This is not well established yet; extensive studies are required to confirm this. Genetic and “somalconal” variations can occur in stem cell derived cell or tissue types and it is hard to justify any sort of compound profiling using these “abnormal” cells/tissues. Some of these stem cell derived products are sold for drug discovery research along with a certificate of analysis using DNA microarray or karyotype analysis. Interestingly, these studies do not prove that stem cell derived cell types such as cardiomyocytes are similar to normal cardiac cells/tissue types. The “scientific responsibility” should be considered as important as or may be more important than “commercialization responsibility”. An honest and efficient combination of scientific and commercial responsibilities will lead to the discovery of better and safer drugs. One should not rush new technologies without complete understanding, especially in important areas such as drug discovery. This often happens when commercialization goals conquer novel scientific discoveries and when investors need “short-term transient” scientific achievements over long-term real benefits.
December 27, 2010 at 10:21 pm
I have no commercial interest whatsoever in the use of stem cells in assays. But I don’t buy the “complete understanding” argument. It is at odds, in my opinion, with the cornerstones of the philosophy of science. There is no such thing as “complete understanding”, because that would boil down to proof that a scientific model or theory is correct, and that isn’t possible. Science consists of a series of progressively more accurate models (we hope), each of them probably doomed to be proven flawed in the end, that is if we don’t already know that they are flawed (we usually do).
Striving for complete understanding before we use something is a recipe for inertia, and it ignores the basic realities of science and technology, if not of life. If complete understanding is the criterion, we should pull all drugs from the market, without exception, and probably ban eating while we are at it.
The question we need to answer before we use stem cells in ‘in vitro’ assays is not whether our assays are a perfect model for the ‘in vivo’ situation. Of course they aren’t. Or whether we perfectly understand the behavior of the cells and the implications of our measurement. Of course we don’t. The question is whether our new assays are significant progress over the old ones. We need to make progress one step at a time. I hope that will increase our understanding.
If there is good evidence, I am willing to accept the argument that stem cell derived cardiac cells are not, at this time, a good substitute for primary human cells. I have not problem with that. But in other areas, profiling in various “abnormal” cells is still routine or even cutting-edge. We have just laid over half a million on the table for tests that will be done in HEK293, and I am reasonably confident that the data will be worth the money. If tomorrow a new CRO appears and offers me, say, the profiling of drugs on a panel of stem-cell derived models of various tissues, my first question will not be whether they have a complete understanding of all these models, because I have no illusions about that. It will be whether they can demonstrate advantages over the stare in the art in identifying off-target effects and toxicity risks. I care about the science and I would hope that part of our payment is invested in good research. But meanwhile, we just need to use the best tools we can get.
December 27, 2010 at 10:38 pm
To which I want to add that I support your point about fully disclosing product limitations, and sensibly limiting claims of effectiveness. That is normal scientific procedure — so Newton’s law of gravity, though known to be incorrect, is still very useful on conditions that the boundaries of its validity are respected.
I am not sure, though, that CROs are that much to blame in this regard, at least not intentionally, or that putting the full burden of proof on them will help that much. Yes, I have seen some CRO people who don’t understand where the useful limits of their own models, but more often the users fail to understand although the vendor provided extensive documentation. The real problem is that modern biological practice requires a complex assembly of skills, often well beyond the knowledge of an individual scientist, and individuals who can’t fall back on expert advice are often forced to treat assays and tools as “black boxes” even if an explanation is available. We still like to pretend that the individual scientist can be a Leonardo in his own lab, and it is no longer true.
December 28, 2010 at 6:26 pm
One answer may be the new the New National Center Advancing Therapeutics (NCAT) at the NIH. The central role of the proposed National Center for Advancing Translational Sciences (NCATS) would be to establish a focused, integrated, and systematic approach for building new bridges to link basic discovery research with therapeutics development and clinical care. The Center would be formed initially by integrating selected translational research programs now located within the National Human Genome Research Institute (NHGRI), the National Center for Research Resources (NCRR), and the NIH Director’s Common Fund.
Sciclips's Blog 
December 21, 2010 at 6:12 pm
It is important to make the distinction between screening assays used for hit discovery, and profiling assays used for lead selection, optimization and preclinical studies. Assays for (primary) screening need to be cheap, perhaps 50 cents per tested compound. Most of us would like to use the most relevant screening assays possible, in the hope of getting better hits and ultimately better leads. But it would be pointless to validate them, as none of the tested compounds will reach the market without long chemical optimization. The extra cost for a regulated validation would also seriously harm innovation, in effect forcing people to abandon their own assay development and purchase all assays from CROs.
Profiling assays are something different. Here we are often asked to spend 5000 dollars to test a compound, instead of 50 cents, and it would be nice to know that we are getting value for money. There already are cross-pharma collaborations to evaluate the assays offered by CROs and share the results. But we should not crack down to hard because this is area of innovation with a lot of small players, and we should allow the technology time to grow. And I don’t see much need for the FDA to get involved in regulating profiling assays used for lead optimization, because all this is still distant from the patient. Perhaps they could develop a “validation kit” that allows CROs to demonstrate benchmark performance and level of correlation with toxicity in humans.
Where it gets more interesting is in the preclinical phase. Here there have been a number of studies, most famously by Peter O’Brien, that suggest that advanced in-vitro profiling assays on human cells would not only be a cheaper but also a better alternative for animal trials. If we can indeed use them to more reliably identify a hepatoxic compound before it gets into human volunteers, it is self-evident that we should. And if assays are going to be used to gain permission for first-in-human trials, obviously they should be validated. The health authorities should take a lead in this, actively studying how our rather ancient pre-clinical practice can be updated using new assay technologies.