TOXICOLOGICAL SCIENCES 112(1), 17–22 (2009)
doi:10.1093/toxsci/kfp202
Advance Access publication August 24, 2009
FORUM SERIES - PART VII
Endless Possibilities: Stem Cells and the Vision for Toxicology Testing
in the 21st Century
Robert E. Chapin* and Donald B. Stedman
Developmental and Reproductive Toxicology, Drug Safety R&D, Pfizer, Inc., Groton, Connecticut 06365
Received July 31, 2009; accepted August 18, 2009
The National Research Council’s (NRC) toxicity testing vision
lays out a bold future for our field. It depends heavily on
computational algorithms based on the latest knowledge of
cellular biochemistry and protein interaction pathways, exposing
human cells to novel compounds in vitro, and being able to
understand the changes seen. At the same time, significant strides
are being made in our understanding of the control, production,
and ‘‘behavior’’ of stem cells. While stem cells offer seemingly
limitless possibilities for regenerative medicine, they have already
delivered new assays to predict embryo-fetal developmental
toxicity in vitro. In addition to providing a model of cells
undergoing differentiation and proliferation, stem cells will play
a major role by giving rise to many of the differentiated cell types
on which this new vision depends. These will not be pure
populations of single cell types but mixtures of cells much more
representative of tissues in vitro. Moving from cells alone in
a culture dish toward the more physiological condition of multiple
cell types being able to interact to maintain homeostasis in the
face of a disequilibrating force (like a toxic exposure) will lead us
toward more useful and correct predictions of in vivo toxicities.
Despite the seemingly insurmountable hurdles, persistence and
creativity are on our side. We expect that a long series of
successive iterations of predictive models will eventually yield
a working process that approximates the NRC’s vision and
delivers on the promise of faster evaluation of chemicals with
reduced animal use.
Key Words: embryonic stem cells; induced pluripotent stem
cells; predictive toxicology; toxicity testing.
The vision articulated in ‘‘Toxicity Testing in the 21st
Century’’ leans heavily on cell-based assays, knowledge of
systems biology and the pathways of cell response to toxicants,
and to a large extent on computational agility. We agree that this
is an excellent vision, toxicology’s own version of the Grand
*
To whom correspondence should be addressed at Pfizer Global R&D,
Developmental and Reproductive Toxicology Center of Expertise, MS 82741336, Eastern Point Road, Groton, CT 06340. Fax: (860) 686-0090. E-mail:
[email protected].
Unification hypothesis in physics, a seeing of things that are not
and saying ‘‘Why the heck not?!’’ At this point in time, the main
value of this vision is as a framework providing the outlines of
the way toxicity testing would ideally be done X years from now.
Like all visionary frameworks, it has serious and major gaps, but
we understand and applaud the benefits of such a vision. These
gaping holes serve to identify where we collectively should be
directing our efforts. Gap filling can unite and guide the efforts
of everyone in a field because as soon as concrete advances are
made in each area, they can be put to immediate use and deliver
immediate improvement to the model.
This installment in the series of Forum articles deals with
one of the testing tools articulated in the vision and in the initial
discussion piece by Andersen and Krewski (2009): stem cells.
Embryonic stem cells are those cells which come from the
inner cell mass of the blastocyst and from which the entire
body arises. They can divide nearly endlessly and differentiate
into any cell type. Stem cells are the new darlings of both basic
science and therapeutics, and interest in them has been
exploding (Fig. 1). They are the yet-to-be-elected but popular
politician, brimming with promise and hope: they are touted as
providing new cures for previously untreatable diseases, they
might replace worn-out cells and provide new therapies for
aging, they can theoretically differentiate into any final cell
type (once we learn the cues required to guide them) and thus
provide new substrate for developing new medicines . . . the
possibilities seem limitless. As with the politician, there can be
a significant gulf between promise and delivery, but the
promises are so beguiling, so ‘‘alluring,’’ that we elect them in
the hopes that miracles still occur.
The recent production of induced pluripotent stem (iPS) cells
adds another layer of luster to the appeal (Maherali and
Hochedlinger 2008; Takahashi and Yamanaka, 2006; Zhao and
Daley, 2008). These cells can come from a (or theoretically
any) terminally differentiated cell in the body, and by
transduction with a few transcription factors (TFs) (Oct4,
Nanog, Klf4, and Sox2), they can be returned to their
undifferentiated pluripotent state. The big appeal of these cells
Ó The Author 2009. Published by Oxford University Press on behalf of the Society of Toxicology. All rights reserved.
For permissions, please email: [email protected]
18
CHAPIN AND STEDMAN
TABLE 1
Examples of cells derived from embryonic stem cells
Cell types
Neural
Fig. 1. Number of publications focused on stem cells, 1990–2008, as
generated by SciFinder.
is (1) they are derived from an adult cell and do not require the
destruction of an embryo, (2) because they could in theory be
made from anyone’s cells, they could avoid immune rejection
if the starting cells were taken from the intended ultimate
recipient, and (3) they could allow the creation of cell lines
from individuals with diverse genetic backgrounds for study of
specific diseases. If iPS cells prove to be as robust as human
embryonic stem cell in generating differentiated cells, they
could open up new doors to the understanding of the genetic
versus environmental factors that contribute to an individual’s
response to a drug or chemical. This method has implications
so profound that the journal Science proclaimed ‘‘Reprogramming’’ (the derivation of iPS cells from differentiated cells) the
Breakthrough of the Year last year (Vogel, 2008).
As we think about how stem cells might contribute to the tox
testing vision, there are two additional items to remember: (1)
Nature abhors 2D monocultures, both macroscopically and
microscopically; there is no such thing in vertebrates as a pure
population of a single kind of cell connected to other cells only
in two dimensions. Thus, it is not surprising that any given
population of stem cells, even highly purified embryonic stem
cells, is heterogeneous, with varying amounts of the pluripotency-inducing TFs in different cells, which in turn leads to
different behaviors of those cells in the cultures and in vivo
(Graf and Stadtfeld, 2008). And (2) most adult tissues contain
stem cells specific for that tissue type. Some of these turn over
quickly and are responsible for the high rates of cell
replacement (i.e., testis, skin, and gut), while others turn over
much more slowly or undetectably (brain and heart) (e.g.,
Hoffman, 2008; Hoogendoorn et al., 2008; reviewed in Slack,
2008.). These tissue-specific stem cells are also heterogeneous,
existing in multiple states characterized by different degrees of
expression of various TFs (Graf and Stadtfeld, 2008). In this
review, we are mostly referring to pluripotent embryonic stem
cells or iPS cells.
As with any promising new technique, there is a combination
of irrational exuberance and some truth to all the rumors.
Phenotype
Markers
Dopaminergic Tyrosine
hydroxylase
Serotonergic Serotonin
Motor
HB9, Isl1
GABA
FOXG1B
neurons
Astrocytes
GFAP
Glial cells
GFAP
Cardiomyocyte Cardiac cells MHC and Nkx2.5
Hepatic
Hepatocytes
ALB, AAT,
and ATP
Vascular
Endothelial
PECAM
(CD31), Flt-1,
VE-cadherin,
and CD34
Islet
b-cell
Insulin I, insulin II,
islet amyloid,
and GLUT-2
Skeleton
Chondrocyte Type II collagen
Osteoblast
Mineralized matrix
Reference
Lee et al. (2000)
Kim et al. (2002)
Wichterle et al. (2002)
Barbieri et al. (2003)
Barbieri et al. (2003)
Barbieri et al. (2003)
*Wobus et al. (2002)
Hamazaki et al. (2001)
Vittet et al. (1996)
Lumelsky et al. (2001)
Kramer et al. (2000)
Bielby et al. (2004)
Note. This table presents some of the cells that have been produced from
stem cells in vitro and provides the markers used to confirm their presence in
the cultures. GABA, gamma-aminobutyric-acid; GFAP, glial fibrillary acidic
protein.
Guiding stem cell differentiation to a desired final differentiated
cell type is a combination of skill and luck. The goal of the field
is still to generate final differentiated cells from pluripotent stem
cells in culture because only cultures can be scaled to meet the
volume needs of the testing which is envisioned, and this is an
opportunity to free ourselves from the costs and logistical and
moral constraints around using animals. Thus far, success in
producing an enriched population of a specific cell type requires
a specific cocktail of extrinsic growth factors or transfection with
specific TFs and a specific external environment, either feeder
layers or a particular 3D culture condition or implantation into
an injured in vivo model, to stimulate the final differentiation
(reviewed in Godier et al., 2008) (Table 1). We are still some
ways from having a straightforward protocol to generate pure
adult hepatocytes, for example (Banas et al., 2006; SotoGutierrez et al., 2008). It is possible to make cells that have
many characteristics of cells along the lineage of hepatocytes,
but a protocol that produces adult terminally differentiated
hepatocytes is still future tense. Similar challenges currently
exist in making insulin-producing cells (reviewed in Raikwar
and Zavazava, 2008). Stem cell–derived cardiomyocytes are
another example of a differentiated cell type that has a number of
cardiac-specific characteristics (protein markers, ion channels,
pacemaker, atrial, and ventricular action potentials). However,
these cardiomyocytes are still more fetal like than adult. There
has been more success with other cell types: pure populations of
STEM CELLS AND THE TOXICITY TESTING VISION
functional dopaminergic neurons have been produced in vitro,
for example (Kim et al., 2002).
There are a number of efforts worldwide which are currently
focused on the application of stem cells to disease treatment
and safety testing, either as stem cells or as their differentiated
daughters. The Stem Cells for Safer Medicines consortium in
the United Kingdom is a public-private partnership aimed at
converting stem cells into hepatocytes and generating predictive safety screens. Geron (Menol Park, CA), Cellartis
(Göteborg, Sweden), and VistaGen Therapeutics (South San
Francisco, CA) are companies which each currently provide or
are developing safety screens for in vivo toxicities. In addition,
Geron is pursuing the production of osteoblasts for treating
osteoporosis, chondrocytes for treating osteoarthritis, and
hepatocytes for multiple therapeutic indications. Cellular
Dynamics (Madison, WI) is another company developing
a cardiomyocyte-based assay for cardiac safety prediction. In
the regrettable current vernacular: ‘‘This is a thriving private
sector space.’’
One issue of concern is which stem cells will be most useful
for fitting into this vision. The answer is it is too soon to tell.
Again, theoretically, any truly pluripotent stem cell should be
able to fill in and provide any differentiated cell type. However,
the practical matter of not yet knowing all the necessary cues to
produce the desired cell types will likely lead us to employ
tissue-specific stem cells or mixes thereof as a way to shortcut
the occult early differentiation process. This will be an area of
much trial and error and will be a secure home for visionary
pragmatism for the foreseeable future.
The challenges of producing ‘‘pure’’ populations of a final
cell type are not surprising. A pure population may not even be
what is needed. Recall that cells in vivo do not develop or live
in pure monoculture but they live and function admixed with
other cell types as a tissue, and we predict that this
heterogeneity will eventually be recognized as crucial to the
sustained differentiated function of any cell type. Thus, flasks
of pure hepatocytes, for example, is the wrong vision. A better
goal might be mixed 3D cultures of hepatocytes, Kupfer cells,
bile ducts, and fibroblasts because only a mixed population in
three dimensions will best mimic the in vivo responses we seek.
We might call this a tissue ‘‘doppel.’’ Liver spheroids are one
step toward this goal and have been shown to better replicate
the in vivo response of the liver to many exposures,
nanoparticles being only one recent example (Lee et al., 2009).
One other useful characteristic of stem cells is that iPS cells
generated from humans with specific diseases maintain some
of the programming characteristic of that disease (Beqqali
et al., 2009; Ebert et al., 2009). This implies that we can (and
should) eventually obtain or generate iPS cells from a wide
variety of people, encompassing the broad spectrum of
metabolic abilities, drug susceptibilities, resistance or susceptibility to disease, etc. Conceptually, this could yield
enormous benefits. As with all conceptual gifts, however, the
details of harnessing this diversity and squeezing something
19
useful from it, even a fraction of the promise, seems
impressively complex. George Church’s Personal Genome
Project at Harvard is doing exactly this and will help highlight
the challenges and their solutions.
So how do stem cells fit into the vision? In two ways: (1) by
being differentiated into cultures of different ‘‘human’’ cell
types whose response to putative toxicants can then be assessed
and (2) by being evaluated for their response to toxicants in
their undifferentiated state (or during differentiation) during
toxicant exposure. The specific allure is that we can use human
cells and thus avoid the extrapolation of responses from
animals, and we can do this with a theoretically limitless
supply of cells. The goal might be a production process which
generates cultures of differentiated cells representative of
various adult tissues which would be exposed to the chemical
of concern and the responses measured. This goal assumes
continued progress in understanding the basic biology of
differentiation for each tissue (which seems reasonable) in
a time frame short enough to be usable (perhaps more of
a stretch), and it assumes that we know the limitations of each
system. These limitations will become more and more apparent
as we push these cultures to meet our expectations, which are
probably well beyond what the cells can reasonably do. So
while stem cells themselves are part of the vision, arguably
their greater contribution is to produce the differentiated cells
that we want and need.
What will it take to realize the promise inherent in these
cells?
1) Robust cell lines which are prolific, easy to maintain,
differentiate appropriately and readily, which recapitulate
in vitro the known in vivo responses to well-known stimuli
(both trophic and toxic), and which yield results which are
reproducible across replicates and laboratories.
2) A known method for the guided differentiation into, say,
15 desired ‘‘tissue doppels,’’ which will include the known
induction cocktail and environmental requirements for producing adult-state tissue recapitulations in vitro.
3) To account for differences in susceptibility, there needs to
be a well-thought-through process for obtaining cells from
a spectrum of genotypes and then applying these in a thoughtful
and systematic way to the predictive process. Again, this
should be achievable.
4) An appropriately scaled production and distribution system
for all these different types of cells. This is ‘‘merely’’ a logistical
challenge, not necessarily requiring scientific breakthroughs to
achieve, although the logistics of this are daunting.
It is worth noting that all of the above use stem cells as
a precursor factory, and once the cells start to differentiate, they
are no longer stem cells. In this vision, the guided
differentiation of stem cells, and their production at scale,
becomes an ‘‘enabling technology.’’
Of these five requirements above, none is impossible. None
of them exists yet, either, but the record of human innovation
20
CHAPIN AND STEDMAN
Fig. 3. This figure presents the pathways most affected in mouse ESC’s.
The pathways were those identified in the National Academies’ book (NRC,
2000) as critical to normal development. The right column presents the gene
expression level in day 5 control embryoid bodies (clumps of cells
differentiating from ESC’s), and the column on the left presents the gene
expression responses after 5 days of treatment with 0.5 ng/ml retinoic acid. The
cells were grown under the conditions in Paquette et al. (2008) and applied to
TaqMan Low Density Arrays from Applied Biosystems (Foster City, CA).
Green represents lowest expression and red represents the highest expression.
These data are simply emblematic of a first step toward assessing the activity in
pathways, going one step beyond individual genes.
Fig. 2. These figures plot major events (paradigm shifts) over the course of
time (Y-axis) against the time before the present across the X-axis. Figure 2A
shows one set of key events. To demonstrate that this apparent speeding up of
paradigm shifts is not simply due to which major events one chooses, Figure
2B depicts a different and more comprehensive set of paradigm-shifting events;
the same trend is evident in both figures. From this (and much other material in
Kurzweil’s book The Singularity is Near), it is clear that the rate of all change is
logarithmic, not linear. The impact for us? Change will happen sooner than we
expect based on a very great deal of history.
and Ray Kurzweil’s strong documentation of the logarithmic
speed of advance in every area (Fig. 2, and Kurzweil, 2005)
give us great confidence that these requirements are achievable
within the next 5–8 years.
Now, in contrast to using them as a factory to make other
cells, can they be exploited for what they naturally and
spontaneously want to do? The answer is ‘‘Yes.’’ What do stem
cells do? They divide and differentiate. This is what normally
happens in an embryo, so it is hardly surprising, then, that stem
cells are being used ‘‘as stem cells’’ to help predict
developmental toxicity. We would not reprise the whole
process by which the European Commission for the Validation
of Alternative Methods evaluated and ‘‘validated’’ the
embryonic stem cell test because that is well documented
already (Buesen et al., 2009; Genschow et al., 2004; Paquette
et al., 2008; Peters et al., 2008; Seiler et al., 2004). It is enough
to know that stem cells are already being exploited for doing
what comes naturally. It is the possibility of using them to
produce cells and cell culture models that do not yet exist
which is so alluring.
How does one measure responses in the stem cells
themselves? One way is to do what the tox testing vision
suggests and focus on pathways and networks of genes, with the
pattern of response relating to an in vivo toxicity profile. This is
just now being explored in stem cells. Figure 3 is an example of
such efforts. In our own laboratory, we have started to evaluate
the effects of teratogens and non-teratogens on the 17 signaling
pathways identified by a National Academy sub-committee in
2000 (NRC, 2000). This is simply one way of showing gene
ontologies, grouped by cellular function. Assessing the changes
due to an exposure and then assembling those data into
‘‘knowledge’’ is when the mental plow must be set deep and
(for the time being) progress is slow. This step of synthesis and
assembly is another opportunity for improvement, as it will be
required of all network and pathway analyses.
It is clear that stem cells can play a significant role in the
realization of the tox testing vision. Since we have some
21
STEM CELLS AND THE TOXICITY TESTING VISION
experience in working with in vitro systems and using them to
predict toxicity, we would like to turn now to some
considerations of the vision itself.
This tox testing vision invokes the power of pathways and
asserts that knowing which pathways are disrupted will help
separate toxic compounds from nontoxic ones. Being long-time
students of scientific progress, we are certain that reality will be
much more complicated than this. There is still much to learn
about how pathways relate to phenotype and a toxicity seen
in vivo. And even leaving all that aside, there is much more to
learn about how physiology works and how organs and tissues
interact to maintain biochemical stability in the gale of exogenous
chemicals. Even with all we know, we realize that our ignorance
is profound and deep. On the other hand, it may be true that in the
final version of this testing scenario, we would not need to know
how a toxicity will manifest but would only need to know which
tissue doppels in vitro are sufficiently affected to pass over the
threshold of change into toxicity. Perhaps, we will not need to
reconstruct all the steps leading from reduced neuronal steroid
sensitivity to increased ovarian steroid output to altered estrous
cycle (persistent estrus) to infertility; eventually, it may be that
seeing the neuronal change will be enough to flag a compound as
potentially toxic and lead to its testing in animals. Indeed, in
a perverse way, this testing system might actually help shortcircuit many future mechanism-of-action studies because the
animal tests would be triggered by an effect noted in a certain
tissue doppel in vitro, which may or may not be the tissue that
manifests the toxicity in vivo. Assuming few false positives
(always a dangerous assumption), we could see this leading to
a small revolution in our understanding of how cells and tissues
link together physiologically in vivo. So there is much to learn
about how to relate pathways to phenotype in vivo, even while we
grant that this may not be critical to implementing the pathwaysbased testing vision.
In addition, the body does automatically what a testing
structure will have to do consciously: integrate a number of
physiological processes in series. Every in vitro assay is an
approximation of what happens in vivo, and it is never 100%
predictive. Modeling the numerous events produced by an
exogenous compound in vivo could build error upon error. For
example, to turn a pathways-based hazard identification into
a risk assessment for free-range humans, we will likely be
modeling things like uptake across an epithelial barrier,
metabolism, excretion, and an organ-specific toxicity response.
For the sake of argument, let us simplify and model only those
four processes. If each prediction has an 85% chance of being
correct (and in our experience, this is greater than the
performance of most predictive assays), then we have
a (0.85)4 ¼ 52% chance of running something through all
four models in series and correctly predicting what will be seen
in vivo. This seems hardly worth all the effort to set up this
massive process. We remain hopeful that an approach will be
developed which will not just amplify the weaknesses of using
multiple assays in series.
All the foregoing is why we caution against overoptimism.
We should not confuse perspicacity with propinquity. That is,
just because we can ‘‘see’’ an object (through the powerful
telescope of the tox testing vision) does not mean that this object
is ‘‘close.’’ ‘‘Seeing it’’ is different from ‘‘being there,’’ and there
are many hurdles to cross before the vision is realized.
Nonetheless, we remain strong supporters of such a vision. It
provides a guide and a framework that allows the immediate
application of any advance in understanding or insight into
a mechanism of toxicity or even basic biology and thus helps
drive all science forward. It should (eventually) help reduce
animal use while generating more data, which are two noble
goals. And it is a deeply compelling vision, rooted in the
assumption that more knowledge, rationally assembled and
compassionately used, can improve the human condition. It is
a challenge worthy of the greatest minds among us.
REFERENCES
Andersen, M. E., and Krewski, D. (2009). Toxicity testing in the 21st century:
Bringing the vision to life. Toxicol. Sci. 107(2), 324–330.
Banas, A., Quinn, G., Yamamoto, Y., Teratani, T., and Ochiya, T. (2006). Stem
cells into liver—basic research and potential clinical applications. Adv. Exp.
Med. Biol. 585, 3–17.
Barbieri, T., Klivenyi, P., Calingasan, N. Y., Lee, H., Kawamata, H.,
Loonam, K., Perrier, A. L., Bruses, J., Rubio, M. E., Topf, N., et al.
(2003). Neural subtype specification of fertilization and nuclear transfer in
embryonic stem cells and application to parkinsonian mice. Nat. Biotechnol.
21, 1200–1207.
Beqqali, A., Eldik, W., Mummery, C., and Passier, R. (2009). Human stem
cells as a model for cardiac differentiation and disease. Cell. Mol. Life Sci.
66(5), 800–813.
Bielby, R. C., Boccaccini, A. R., Polak, J. M., and Buttery, L. D. K. (2004). In
vitro differentiation and in vivo mineralization of osteogenic cells derived
from human embryonic stem cells. Tissue Eng. 10, 1518–1525.
Buesen, R., Genschow, E., Slawik, B., Visan, A., Spielmann, H., Luch, A., and
Seiler, A. (2009). Embryonic stem cell test remastered: Comparison between
the validated EST and the new molecular FACS-EST for assessing
developmental toxicity in vitro. Toxicol. Sci. 108(2), 389–400.
Ebert, A. D., Yu, J., Rose, F. F., Mattis, V. B., Lorson, C. L., Thomson, J. A.,
and Svendsen, C. N. (2009). Induced pluripotent stem cells from a spinal
muscular atrophy patient. Nature 457(7277), 277–280.
Genschow, E., Spielmann, H., Scholz, G., Pohl, I., Seiler, A., Clemann, N.,
Bremer, S., and Becker, K. (2004). Validation of the embryonic stem cell test
in the international ECVAM validation study on three in vitro embryotoxicity tests. Altern. Lab. Anim. 32(3), 209–244.
Godier, A. F. G., Marolt, D., Gerecht, S., Tajnsek, U., Martens, T. P., and
Vunjak-Navakovic, G. (2008). Engineered microenvironments for human
stem cells. Birth Defects Res. C 84, 335–347.
Graf, T., and Stadtfeld, M. (2008). Heterogeneity of embryonic and adult stem
cells. Cell Stem Cell 3, 480–483.
Hamazaki, T., Iiboshi, Y., Oka, M., Papst, P. J., Meacham, A. M., Zon, L. I.,
and Terada, N. (2001). Hepatic maturation in differentiating embryonic stem
cells in vitro. FEBS Lett. 497, 15–19.
Hoffman, W. (2008). Regeneration of the gastric mucosa and its glands from
stem cells. Curr. Med. Chem. 15, 3133–3144.
Hoogendoorn, R. J. W., Lu, Z. F., Kroeze, R. J., Bank, R. A., Wuisman, P. I.,
and Helder, M. N. (2008). Adipose stem cells for intervertebral disc
22
CHAPIN AND STEDMAN
regeneration: Current status and concepts for the future. J. Cell. Mol. Med.
12(6A), 2205–2216.
Kim, J.-H., Auerbach, J. M., Rodriguez-Gomez, J. A., Velasco, I., Gavin, D.,
Lumelsky, N., Lee, S.-H., Nguyen, J., Sanchez-Pernaute, R., Bankiewicz, K.,
et al. (2002). Dopamine neurons derived from embryonic stem cells function
in an animal model of Parkinson’s disease. Nature 418, 50–56.
Kramer, J., Hegert, C., Guan, K., Wobus, A. M., Müller, P. K., and
Rohwedel, J. (2000). Embryonic stem cell-derived chondrogenic differentiation in vitro: Activation by BMP-2 and BMP-4. Mech. Dev. 92, 193–205.
Kurzweil, R. (2005). In The Singularity Is Near, Viking Press: New York, pp. 652.
Lee, J., Lilly, G. D., Doty, R. C., Podsiadlo, P., and Kotov, N. A. (2009). In vitro
toxicity testing of nanoparticles in 3D cell culture. Small 5(10), 1213–1221.
Lee, S.-H., Lumelsky, N., Studer, L., Auerbach, J. M., and McKay, R. D.
(2000). Efficient generation of midbrain and hindbrain neurons from mouse
embryonic stem cells. Nat. Biotechnol. 18, 675–679.
Lumelsky, N., Blondel, O., Laeng, P., Velasco, I., Ravin, R., and McKay, R.
(2001). Differentiation of embryonic stem cells to insulin-secreting structures
similar to pancreatic islets. Science 292, 1389–1394.
Maherali, N., and Hochedlinger, K. (2008). Guidelines and techniques for the
generation of induced pluripotent stem cells. Cell Stem Cell 3(6), 505–605.
National Research Council (NRC). (2000). Scientific Frontiers in Developmental Toxicology and Risk Assessment. The National Research Council
Committee on Developmental Toxicology, The National Academies Press,
Washington, DC, pp. 327.
National Research Council (NRC). (2007). Toxicity Testing in the 21st Century.
A Vision and a Strategy. The National Research Council Committee on
Toxicity Testing and Assessment of Environmental Agents, p. 196. The
National Academies Press, Washington, DC.
Paquette, J., Kumpf, S., Streck, R., Thomson, J., Chapin, R., and Stedman, D.
(2008). Assessment of the embryonic stem cell test and application and use
in the pharmaceutical industry. Birth Defects Res. B Dev. Reprod. Toxicol.
83(2), 104–111.
Peters, A., Steemans, M., Hansen, E., Mesens, N., Verheyen, G., and
Vanparys, P. (2008). Evaluation of the embryotoxic potency of compounds
in a newly revised high throughput embryonic stem cell test. Toxicol. Sci.
105(2), 342–350.
Raikwar, S., and Zavazava, N. (2008). Insulin-producing cells derived from
embryonic stem cells: Are we there yet? J. Cell. Physiol. 218, 256–263.
Seiler, A., Visan, A., Buesen, R., Genschow, E., and Spielmann, H. (2004).
Improvement of an in vitro stem cell assay for developmental toxicity: The
use of molecular endpoints in the embryonic stem cell test. Reprod. Toxicol.
18(2), 231–240.
Slack, J. M. W. (2008). Origin of stem cells in organogenesis. Science 322,
1498–1501.
Soto-Gutierrez, A., Basma, H., Navarro-Alvarez, N., Uygun, B. E.,
Yarmush, M. L., Kobayashi, N., and Fox, I. J. (2008). Differentiating stem
cells into liver. Biotechnol. Genet. Eng. Rev. 25, 149–164.
Takahashi, K., and Yamanaka, S. (2006). Induction of pluripotent stem cells
from mouse embryonic and adult fibroblast cultures by defined factors. Cell
126, 663–676.
Vittet, D., Prandini, M. H., Berthier, R., Schweitzer, A., Martin-Sisteron, H.,
Uzan, G., and Dejana, E. (1996). Embryonic stem cells differentiate in vitro
to endothelial cells through successive maturation steps. Blood 88,
3424–3431.
Vogel, G. (2008). Breakthrough of the year: Reprogramming cells. Science
322(5909), 1766–1767.
Wichterle, H., Lieberam, I., Porter, J. A., and Jessell, T. M. (2002). Directed
differentiation of embryonic stem cells into motor neurons. Cell 110,
385–397.
Wobus, A. M., Guan, K., Yang, H.-T., and Boehler, K. R. (2002). Embryonic
stem cells as a model to study cardiac, skeletal muscle and vascular smooth
muscle cell differentiation. Methods Mol. Biol. 185, 127–156.
Zhao, R., and Daley, G. Q. (2008). From fibroblasts to iPS cells: Induced
pluripotency by defined factors. J. Cell. Biochem. 105, 949–955.