PERSPECTIVES
N E U RO S C I E N C E
Astrocyte
Cellular Interactions
in the Stem Cell Niche
Neuron
NSC
Andrew E. Wurmser, Theo D. Palmer, Fred H. Gage
he ability of stem cells to both selfrenew and differentiate into many different cell types enables these versatile cells to generate and repair tissues and
organs. Yet studies of the fruit fly
Drosophila and of mammalian skin, intestine, bone marrow, and brain reveal that
these inherent stem cell features are tightly
regulated by the cells and proteins that constitute the extracellular environment (or
“niche”) that stem cells inhabit (1). On page
1338 of this issue, Shen et al. (2) take an important step forward in our understanding of
the stem cell niche. They show that endothelial cells (ECs) that are enriched in the
niche occupied by neural stem cells (NSCs)
regulate NSC proliferation and induce these
stem cells to become neurons in vitro.
It is well established that NSCs are not
randomly distributed throughout the brain,
but rather are concentrated around blood
vessels (see the figure) (3–5). This location places NSCs in close proximity to the
ECs that line blood vessels, facilitating
communication between these two cell
types (3–6). To test the degree of intercellular communication between NSCs and
ECs, Shen et al. cultured NSCs and monitored changes in their behavior when ECs
were brought into close proximity (2).
These investigators maintained cultures of
mouse embryonic NSCs (derived from the
cerebral cortex of 10- to 11-day-old mouse
embryos) by adding fibroblast growth factor–2. Under these conditions, NSCs proliferated slowly and many of them exited
the cell cycle, choosing to differentiate instead (2). However, when NSCs were cocultured with ECs their proliferation rate
doubled, resulting in the formation of large
interconnected sheets of undifferentiated
cells. A clever aspect of Shen et al.’s strategy was to introduce ECs into NSC cultures by means of transwell inserts. The
pores of the transwells were too small to
allow cell-cell contact between NSCs and
CREDIT: KATHARINE SUTLIFF/SCIENCE
T
A. E. Wurmser and F. H. Gage are in the Laboratory
of Genetics, Salk Institute, La Jolla, CA 92037, USA.
T. D. Palmer is in the Department of Neurosurgery,
Stanford University, Stanford, CA 94305, USA. Email: [email protected]
ECs, but were large enough to enable sigEC
naling factors secreted by ECs to diffuse
into the NSC cultures. Remarkably, the removal of transwells containing ECs trigCapillary
gered the coordinated differentiation of
proliferating NSCs into neurons.
Only 9% of NSCs unexposed to
Song et al.
Shen et al.
ECs expressed mature neuronal
markers, compared with 31 to 64%
Additional
of NSCs exposed to the EC transsignals from
EC
Astrocyte
the niche?
wells. This trend also was observed with cultured NSCs derived from the subventricular
Proliferative/
liferative/
if
Proliferative/
roliferativ
fera
er
zone of adult mouse brain (2).
Differentia
Differentiation
re
neu
urogenic
ro
neurogen
neurogenic
roge
o
to
o astrocy
astrocytes
tr
Thus, signaling molecules secretsiignal(s)
na
signal(s)
n
nal(
)
ed by ECs induced a shift in the
mixed population of proliferating
and differentiating NSCs, pushing
them toward self-renewal while simultaneously priming them for the
production of neurons.
NSC
The neurogenic effects of ECs
In
ncreased
rea
could not be mimicked by fibrobNSC
neuron
ro
self-renew
self-renewal
-re
lasts or by vascular smooth muscle
production
roduction
ct
cells (2), indicating that not all cell
types alter the NSC differentiation
Differentiation
to other cell
profile. The new work expands the
types?
importance of ECs beyond their traditional role as structural components of blood vessels. ECs are
known to enhance neurogenesis,
Neurons
NSCs
possibly through the secretion of
brain-derived neurotrophic factor, Neurogenesis in the NSC niche. (Top) The niche inand to induce astrocyte differentia- habited by mammalian adult NSCs is situated close to
tion within the optic nerve (5, 7). blood vessels. NSCs interact with ECs that line the
But ECs are not limited to instruct- blood vessels and with astrocytes. (Bottom) Both ECs
ing the neural lineages—they also and astrocytes secrete signaling molecules that influpromote formation of pancreatic ence neighboring NSCs to proliferate and to differentiand liver tissue independently of ate into neurons.
their ability to form vasculature (8,
9). Indeed, ECs may be instructive for many and neuron production remains an impordifferent cell types and tissues. It remains to tant goal. Such information would circumbe determined whether ECs in different tis- vent the need for NSC-EC coculture and
sues or at different developmental time pe- potentially could facilitate the production
riods display variations in their profiles of of neurons for future clinical applications.
secreted signaling molecules. Cumulatively, Identification of such factors should help to
these studies provide evidence that ECs are elucidate whether ECs boost the neurogenic
important tissue architects, specifying the potential of NSCs directly or by enhancing
fates of many different neighboring cell the survival of differentiating neurons.
types including NSCs.
Most noteworthy are the potential impliElucidating the signaling molecules se- cations of the Shen et al. work. First, if the
creted by ECs that elicit NSC proliferation influence of ECs on NSCs in vitro is con-
www.sciencemag.org
SCIENCE
VOL 304
28 MAY 2004
1253
PERSPECTIVES
served in an in vivo setting, a new therapeutic avenue for neuronal induction may be on
the horizon. Moreover, NSCs in vivo are
likely to be influenced by a convergence of
signals from many neighboring cell types.
Astrocytes, for instance, enhance the proliferative and neurogenic properties of NSCs
by a factor of 2 and 6, respectively (10).
Simultaneously combining the effects of
ECs and astrocytes, if technically feasible,
could elevate the production of neurons from
NSCs beyond what is observed individually
with either cell type (see the figure).
Researchers use a wide range of techniques to isolate and grow NSCs in vitro.
It will be interesting to see whether the
neurogenic and proliferative effects of ECs
on NSCs cultured by the Shen et al.
method are observed with NSCs grown in
different culture systems. The Shen et al.
work establishes EC-NSC coculture as an
important tool for promoting NSC selfrenewal and differentiation into neurons.
With more extensive study, this method ultimately may prove to be useful for cell replacement therapy.
References
1. E. Fuchs et al., Cell 116, 769 (2004).
2. Q. Shen et al., Science 304, 1338 (2004); published
online 1 April 2004 (10.1126/science.1095505).
3. T. D. Palmer et al., J. Comp. Neurol. 425, 479 (2000).
4. A. Capela, S. Temple, Neuron 35, 865 (2002).
5. A. Louissaint et al., Neuron 34, 945 ( 2002).
6. W. Risau, Nature 386, 671 (1997).
7. H. Mi et al., J. Neurosci. 21, 1538 (2001).
8. E. Lammert, O. Cleaver, D. Melton, Science 294, 564
(2001); published online 22 September 2001 (10.1126/
science.1064344).
9. K. Matsumoto, H. Yoshitomi, J. Rossant, K. S. Zaret,
Science 294, 559 (2001); published online 22
September 2001 (10.1126/science.1063889).
10. H. Song et al., Nature 417, 39 (2002).
D E V E L O P M E N TA L B I O L O G Y
prey, N. vectensis burrows through the mud
of brackish lagoons and estuaries in North
America and southern England. A cross
section of the adult shows not radial symmetry, as dogma would predict, but a plane
of bilateral symmetry (known as the “directive axis”) that traverses the pharynx at
Peter Holland
right angles to the primary oral-aboral
(mouth-foot) body axis. This is not an oddea anemones, corals, and jellyfish have ber of the basal cnidarian class Anthozoa. ity of Nematostella, but a characteristic of
had a much-maligned history. Aristotle This sea anemone has unusual habits that sea anemones. The bilateral symmetry of
considered them “zoophytes”: neither would have confused Aristotle, Bonnet, anthozoans (and not other cnidarians) was
animals nor plants, yet having characteris- and the early naturalists. Whereas most sea noted by Hyman (3), an influential invertetics of both. Similar views persisted for anemones stick to rocks, looking rather brate zoologist of the 20th century, but
more than 2000 years. For example, Charles like flowers yet capturing and devouring failed to become a part of common zoologBonnet’s 1799 Scala
ical knowledge. If sea
Naturae or Chain of
anemones possess bilatEnhanced online at
eral symmetry, is it howww.sciencemag.org/cgi/ Being (1) placed sea
content/full/304/5675/1255 anemones just above
mologous to our own biplants and below
lateral symmetry or did
tapeworms. He thought corals even more
it arise by convergent
primitive, situated far below plants and only
evolution? In other
marginally above asbestos. Zoologists may
words, did bilateral symshudder at such overtly hierarchical classifimetry originate earlier
cation schemes, yet the dominant view toin our ancestry than is
day is not that different. Most animals—
commonly believed or
from worms to whales, and flies to foxes—
did anthozoans evolve
comprise an evolutionary lineage called the
from a radial ancestor
Bilateria. Sometimes loosely termed “highand develop bilaterality
er animals,” bilaterians have front and rear
independently?
ends (anterior and posterior), an up-andHomology can be indown axis (dorsal and ventral), and a plane
vestigated by compariof mirror symmetry running between the
son of gene-expression
left and right sides (bilateral symmetry). In
patterns. If the same
contrast, sea anemones and their kin (phygenes are used to conlum Cnidaria) diverged earlier in animal
trol development of a
evolution, supposedly before the invention
structure (or axis) in two
of bilateral symmetry. These animals exhibmorphologically differit a simpler form of symmetry termed “radient animals, this lends
al” symmetry. On page 1335 of this issue,
support to the hypotheFinnerty and colleagues (2) present evisis of homology aldence suggesting it is time to rethink the orithough it does not prove
gins of bilateral symmetry.
it. The Hox gene cluster
Finnerty and co-workers report their
is the canonical set of
study of the starlet sea anemone, Nemagenes implicated in contostella vectensis (see the figure), a mem- Stars and stripes. The starlet sea anemone, Nematostella vectensis, trol of the anterior-posis a cnidarian that is not radially symmetrical but exhibits bilateral terior axis in bilaterians.
symmetry (2). Bilateral symmetry may have evolved before the split As for the dorsoventral
The author is in the Department of Zoology,
between bilaterians and cnidarians. [Adult N. vectensis are ~2 cm in axis, the dpp or bone
University of Oxford, Oxford, OX1 3PS, UK. E-mail:
morphogenetic protein
[email protected]
length.]
The Ups and Downs
of a Sea Anemone
CREDIT: JOHN R. FINNERTY/BOSTON UNIVERSITY
S
www.sciencemag.org
SCIENCE
VOL 304
28 MAY 2004
1255