Download Neural stem cells and cell replacement therapy

Clinical Science (2005) 108, 13–22 (Printed in Great Britain)
R
E
V
I
E
W
Neural stem cells and cell replacement
therapy: making the right cells
Angela BITHELL and Brenda P. WILLIAMS
Institute of Psychiatry, Department of Psychological Medicine, PO Box 52, De Crespigny Park, London SE5 8AF, U.K.
A
B
S
T
R
A
C
T
The past few years have seen major advances in the field of NSC (neural stem cell) research with
increasing emphasis towards its application in cell-replacement therapy for neurological disorders.
However, the clinical application of NSCs will remain largely unfeasible until a comprehensive
understanding of the cellular and molecular mechanisms of NSC fate specification is achieved.
With this understanding will come an increased possibility to exploit the potential of stem cells
in order to manufacture transplantable NSCs able to provide a safe and effective therapy for
previously untreatable neurological disorders. Since the pathology of each of these disorders is
determined by the loss or damage of a specific neural cell population, it may be necessary to
generate a range of NSCs able to replace specific neurons or glia rather than generating a generic
NSC population. Currently, a diverse range of strategies is being investigated with this goal in
mind. In this review, we focus on the relationship between NSC specification and differentiation
and discuss how this information may be used to direct NSCs towards a particular fate.
INTRODUCTION
Considerable media attention has brought stem cell
research into the spotlight, not least because of the
exciting possibilities of its application in cell replacement therapies for neurological disorders using NSCs
(neural stem cells). Despite the obvious benefits promised
by this field and some encouraging preliminary studies,
in reality there still remains a gulf between theory and
practice.
Leaving aside the highly contentious ethical debate
surrounding the procurement and use of ES (embryonic
stem) cells to generate NSCs, as opposed to adult sources
of stem cells such as those obtained from bone marrow,
we still lack a profound understanding of the basic biology of NSCs and how they can be manipulated to provide consistent and effective results in cell-replacement
strategies. NSCs are defined by three main characteristics:
they can self-renew, give rise to all of the major neural cell
types, i.e. neurons, oligodendrocytes and astrocytes, and
can repopulate a damaged region. The first of these
characteristics ensures that a pool of NSCs is maintained
at the same time as more restricted progeny are generated,
and their ability to self-renew provides a convenient approach by which to expand the initial NSC population.
Despite the use of primary fetal tissue as proof-ofprinciple in a number of clinical studies, including treatment of PD (Parkinson’s disease) patients [1,2], it is not
a viable option for large-scale therapeutic application
due to a lack of available tissue and ethical considerations.
Therefore a readily expandable NSC population, possessing an innate ability to make the three major neural cell
types and repair damage following injury or disease, is an
obvious attraction.
Key words: cell replacement therapy, embryonic stem cell, neural stem cell, neurological disorder.
Abbreviations: AP, anterior-posterior; BMP, bone morphogenetic protein; CNS, central nervous system; DA, dopaminergic; DV,
dorso-ventral; EB, embryoid body; EGF, epidermal growth factor; EGFP, enhanced green fluorescent protein; ES, embryonic stem;
FGF2, fibroblast growth factor 2; MCAO, middle cerebral artery occlusion; MN, motor neuron; NPC, neural progenitor cell; NRP,
neuronally-restricted progenitor cell; NSC, neural stem cell; PD, Parkinson’s disease; PI, positional identity; RA, retinoic acid; SGL,
subgranular layer; Shh, sonic hedgehog; SVZ, sub-ventricular zone; TF, transcription factor.
Correspondence: Dr Brenda P. Williams (email [email protected]).
C
2005 The Biochemical Society
13
14
A. Bithell and B. P. Williams
Table 1 Comparison of possible sources of NSCs
The table illustrates some of the important pros and cons of each of the four stem cell types able to give rise to neural cells. ICM, inner cell mass.
Stem cell type
Source
Pros
Cons
ES cell-derived NSCs
ICM of blastocyst
Embryonic NSCs
Embryonic CNS
Tumorigenicity, ethical considerations,
purification (markers), and require neural specification
Long-term maintenance and ethical considerations
Adult NSCs
SVZ and SGL of hippocampus
Pluripotent, unlimited proliferation and
stable karyotype
Neural lineage-committed, non-tumorigenic,
and regionally specific
Neural lineage-committed
Non-NSCs
Bone marrow, skin,
umbilical cord, blood etc.
No ethical considerations, plentiful/accessible
supply, and generate autologous cells
Nevertheless, NSCs for clinical applications must
possess a large number of specific qualities. These cells
must be readily available, expandable and karyotypically
stable in vitro, be able to differentiate in a predictable
and appropriate manner, be able to generate functional
neural cells that integrate appropriately and consequently
mediate functional repair. In short, NSCs must be predictable, efficient, effective and, of course, safe. It is thus
profoundly important to choose the correct starting cell
population.
CHOOSING A STARTING CELL POPULATION:
MAKING NSCs
To date, many types of stem cells have been studied
for their effectiveness in neural replacement strategies,
including ES cells, NSCs and a range of non-NSCs. In this
review, we will discuss the advantages and disadvantages
of each (summarized in Table 1) as well as recent experimental evidence that highlights their potential use for
clinical application.
ES cells
ES cells are pluripotent stem cells from which all
tissues of the developing embryo are derived and this
pluripotency, together with a capacity for unlimited selfrenewal, makes them an attractive source of cells for
NSC research. ES cells, first grown in vitro in the early
1980s [3,4], can readily be isolated from the ICM (inner
cell mass) of a blastocyst, expanded in vitro with LIF
(leukaemia inhibitory factor) and, subsequently, directed
down a range of specific lineage pathways, usually via
aggregates termed EBs (embryoid bodies), by exposure
to appropriate exogenous signalling molecules (for a
comprehensive review, see [5]). The generation of highly
enriched populations of NSCs from ES cells in vitro
has been the focus of concentrated efforts by many
research groups. Commonly, following growth as EBs,
cells are specified to adopt an NSC or an NPC (neural
progenitor cell) fate by exposure to RA (retinoic acid) or
by maintenance in chemically defined minimal medium
C
2005 The Biochemical Society
Long-term maintenance, restricted potential?, and
limited availability
Require neural specification, and restricted potential?
in the presence of FGF2 (fibroblast growth factor 2), a
potent mitogen. This latter process is selective for the
survival of NPCs [6,7]. However, NSCs have also been
derived from ES cells grown in co-culture with BMSCs
(bone marrow-derived stromal cells) without the need for
EB intermediates [8].
ES-derived NSCs or NPCs (more restricted multipotential progenitor cells) can then be manipulated further by cell culture conditions to generate different types
of neurons and glia. This step is particularly pertinent in
terms of generating NSCs for transplantation. The ability
to simply generate neurons or glia is not sufficient; it is
necessary for NSCs to be able to generate specific neurons
or glia in a predictable and efficient manner. Several
groups report that ES-derived NSCs commonly give rise
to GABAergic and some glutamatergic neurons as well
as astrocytes and occasionally oligodendrocytes [6,9], as
do directly isolated NSCs [10,11]. However, commonly
encountered neurodegenerative diseases including PD
and ALS (amyotrophic lateral sclerosis) are characterized
by a selective loss of DA (dopaminergic) and cholinergic
neurons respectively, and thus replacement strategies
must focus on NSCs able to generate these particular
neurons in sufficient quantities. In attempts to manipulate
the identity of ES cell-derived neurons or glia, a range of
different cell culture conditions have been utilized with
varying levels of success. For example, the addition of
Shh (sonic hedgehog), FGF8 and AA (ascorbic acid) to
culture medium at specific time points during in vitro
neural specification of ES cells increases the number of
DA and serotonergic neurons generated [12]. However,
exposing these same cells to RA and Shh significantly
increases the number of MNs (motor neurons) produced
[13]. The significance of this manipulation of neuronal
fate determination by extracellular signalling molecules
will be discussed in more detail later.
Despite the obvious benefits of using ES cells to
generate NSCs, there are associated caveats. Primarily,
it is necessary to restrict these cells to a neural lineage
[14]. With ES cells this requirement is acutely important
since any remaining non-neural pluripotent stem cells
could give rise to teratomas upon transplantation, and
Neural stem cells and cell replacement therapy
have been shown to do so in rodents [7]. This represents
something of a dilemma as one of the problems in the
field of NSC research remains the lack of a definitive
NSC marker. Indeed, even the once widely accepted
notion of markers that define differentiated neural
cell types has been challenged [15]. A vast array of
positive and negative selective markers are known that
allow for the enrichment of NSCs from non-neural or
neural cell populations, including LeX (Lewis X; [16]),
the intermediate filament protein Nestin, PNA (peanut
agglutinin), HSA (heat stable antigen; [17]) and the TFs
(transcription factors) Musashi1 [18,19] and Sox1 [20].
However, none of these markers can be used to exclusively identify NSCs since many are found in more
restricted NPCs or indeed in non-neural cell types.
Although highly enriched populations of ES-cell derived
NSCs can be purified by combining the use of these
markers with careful manipulation of cell culture conditions, at present the possibility of residual pluripotent ES
cells poses an inherent and unacceptable risk of tumour
formation.
NSCs
An alternative to using ES cell-derived NSCs is to
directly isolate NSCs. These cells have already undergone
several steps of lineage commitment to produce stem cells
destined to give rise to neurons, oligodendrocytes and
astrocytes. Since NSCs are more lineage-restricted than
ES cells, they represent less of a risk for tumour formation
following transplantation. Indeed, from the many animal
transplantation studies carried out, there is little evidence
of their spontaneous tumorigenicity.
NSCs can be isolated from either the embryonic
CNS (central nervous system) or from very restricted
neurogenic regions now known to persist in two NSC
niches in the adult brain: one in the SVZ (sub-ventricular
zone) lining the lateral ventricles [21–23] and the other
in the dentate gyrus of the hippocampus [24], although
the multipotentiality of the latter remains a contentious
issue [23].
In addition, more restricted multipotential NPCs with
the capacity to give rise to both neuronal and glial progeny in vitro have been isolated from numerous adult
brain regions in both rodents and humans [25–29]. It
remains an important consideration whether or not it is
necessary to begin with classically defined NSCs for cellreplacement therapy when more restricted NPCs, that
already contain the desired regional specification, may be
a more suitable alternative (for review, see [30]). Indeed,
lineage-restricted NPCs have been shown to be able
to integrate into the host system [31] and to elicit repair in animal models, including PD [32] and multiple
sclerosis [33].
Similarly to ES-derived NSCs, NSCs directly isolated
from the CNS also have the ability to self-renew and
can give rise to neurons, oligodendrocytes and astrocytes.
However, the kind of differentiated cell types that they
generate vary depending upon the developmental stage
and region from which they are isolated [11,34–36] and
the in vitro conditions in which they are grown thereafter
[11]. It has also become increasingly apparent that their
capacity to yield a variety of specific neural cell types can
be readily altered by manipulation of intrinsic and
extrinsic cues, a point discussed in more detail below.
Non-NSCs
Apart from embryonic and adult sources of ES cells
or NSCs, a large body of evidence has emerged over
the last few years showing the capacity of non-NSCs,
including bone marrow [37], skin [38] and umbilical cord
[39,40] stem cells, to generate various types of neurons
and glia. The obvious attraction of the use of these cells
is that they are in ready supply and eliminate many
of the ethical considerations that complicate our use of
embryonic-derived cells. Furthermore, they raise the
possibility of generating a large number of autologous
NPCs for transplantation, thus eliminating concerns over
host immune response to non-autologous grafted cells.
Although an in-depth discussion concerning the potential
of non-NSCs to generate neural cells for transplantation
is beyond the scope of this review, they represent a
potentially exciting resource, which is evident from the
ever-increasing number of reports revealing a heretofore
unimagined plasticity [41].
Thus there are many pathways by which we can reach
our goal: the generation of a population of neural cells for
transplantation therapy, each with its own characteristic
benefits and caveats (summarized in Table 1). In the next
section, we outline our current understanding of the basic
biology of NSCs, describing how intrinsic and extrinsic
factors act to control their proliferation, specification
and differentiation into specific types of neural cells.
We discuss how this knowledge can be exploited by
NSC researchers to manufacture NSCs with specific
characteristics.
SUPPLY AND DEMAND: ENOUGH NSCs
TO GO AROUND
As discussed earlier, ES cells are often used to generate
NSCs because of their unlimited ability to proliferate
in vitro. NSCs also retain the potential to self-renew,
although their capacity to do so is finite, with rodent and
human NSCs displaying different limitations in terms of
long-term propagation [42,43]. Nevertheless, there have
been two main ways in which NSCs have been maintained
in vitro for expansion (Figure 1): (i) as free-floating
clonally derived neurospheres, grown in the presence
of the mitogens EGF (epidermal growth factor) and/or
FGF2; or (ii) as adherent immortalized NSC lines,
typically carrying an oncogene to facilitate continued
C
2005 The Biochemical Society
15
16
A. Bithell and B. P. Williams
Figure 1 Two main ways in which NSCs are maintained in vitro
NSCs can be propagated in vitro either as neurospheres or as immortalized NSC lines. Neurospheres are generated from single NSCs by proliferation with the mitogens
FGF2 and/or EGF and can generate all three major neural cell types: neurons, oligodendrocytes and astrocytes. However, being non-immortalized somatic cells, they
have a finite lifespan in vitro . In contrast, NSC lines are generated by immortalization of NSCs using oncogenes (such as v-myc) or overexpression of the catalytic
subunit of telomerase (TERT). This allows their sustained growth in vitro . These cells can also generate all three major neural cell types but, of course, these cells are
now genetically modified, a possible problem for use in human transplantation.
proliferation, again grown in the presence of FGF2
(and/or EGF). One of the reasons for the senescence of
NSCs maintained long-term in vitro is believed to be
telomere shortening [42]. Previous studies have shown
that NSC immortalization using the oncogene v-myc
causes increased telomerase activity, thus maintaining a
stable karyotype and permitting extensive perpetuation
of human NSC lines in vitro [44]. Similarly, FGF2 has
been shown to mediate a dose-dependent increase in
telomerase activity in NPCs isolated from the embryonic
mouse cortex [45]. Alternatively, it is possible to
immortalize NPCs by ectopic expression of TERT, the
catalytic subunit of human telomerase, shown to sustain
human NPC expansion in vitro [46].
In terms of their ability to promote functional repair in the damaged brain, although a number of rodent
studies using immortalized NSC lines show evidence of
functional repair [47–49], less have been reported using
neurosphere-derived cells [50]. However, in these types of
studies, the migrational properties, survival and capacity
of NSC populations to differentiate remain unpredictable
and their mode of repair is poorly understood.
The ultimate goal of NSC research for replacement
therapy is to generate large numbers of NSCs or NPCs
that will efficiently undergo site-specific differentiation
before correctly integrating into the host environment
to mediate functional recovery. How can this goal be
achieved? Can NSCs meet all of the requirements and
how will they respond to the host environment? During
embryonic development, the diverse cell types in the
brain are generated in an exquisitely precise spatiotemporal order under the control of both intrinsic and extrinsic factors. Can we imitate the way in which neural cell
C
2005 The Biochemical Society
fate is determined in the developing embryo to produce
‘designer’ NSCs, programmed to adopt particular cell
fates?
RIGHT CELL, RIGHT PLACE:
LESSONS FROM DEVELOPMENT
Patterning along the neuroaxis during embryogenesis
is initiated by gradients of specific diffusible molecules
secreted from signalling centres that are responsible
for the induction of PI (positional identity). These
factors interact both synergistically and repressively
with one another in a concentration-dependent manner
to switch on cascades of downstream TFs that define
specific populations of NPCs in order to generate
the broad range of cells that constitute the adult
brain (for in depth reviews, see [51–57], and research
articles, see [11,58]). This is illustrated in Figure 2,
with two opposing gradients of diffusible signalling
factors initiating a concentration-dependent expression
of different TFs along the DV (dorso-ventral) axis [this
is equally applicable to patterning of the AP (anteriorposterior) axis]. The repressive or inductive interactions
between TFs refine further the boundaries between
each expression domain. Each progenitor domain then
generates particular differentiated cell types (Figure 2,
cell types 1–5). In this way, specific regions composed
of specific types of neurons and glia develop along the
DV and AP axes of the CNS. Despite the distinctive
characteristics of each adult brain region, the same
underlying mechanism of generating cellular diversity is
used along the extent of the developing neuroaxis, simply
Neural stem cells and cell replacement therapy
promote functional repair following transplantation in
rodent models of PD [60].
Using this knowledge, it seems logical that isolating
NSCs from the appropriate region of the embryonic brain
or engineering NPCs with a specific PI could provide a
convenient source of pre-programmed progenitor cells
that display the requisite gene expression. It is therefore
important to determine the possibilities and limitations
of using PI to predict neural cell fate determination both
in vitro and following transplantation.
MAINTAINING PI IN VITRO
Figure 2 Simplified representation of the mechanism of DV
patterning of the neuroaxis
Two opposing gradients of diffusible signalling molecules, A (black) and B (white),
are expressed in the developing neural tube. This establishes broad overlapping
regions of specific concentrations of A and B, which can interact with one another
(black/grey/white bar). The relative concentrations of these signalling molecules
regulate the expression of specific sets of TFs along the DV axis which, in turn,
can activate or repress one another (NB this equally applies to the AP axis). This
leads to the formation of distinct and discreet regions of TF expression (solid
coloured rectangles from black to pale grey), which specify the fate of subsequent
progeny from each region, producing different types of neural cells (illustrated
as cell types 1–5). In this way, the diverse range of cell types required to
generate particular regions of the brain is obtained. Similar mechanisms establish
DV and AP patterning along the entire length of the developing neural tube, each
with its own combination of diffusible signalling molecules and region-specific TF
expression.
by using different sets of diffusible signalling molecules
that regulate the expression of different downstream
TFs.
NPCs isolated from different progenitor zones within
the embryonic brain thus contain an intrinsic specification, conferred by their TF expression profile. When
NPCs are grown in vitro they are known to retain their
PI and to generate neurons or glia appropriate for
their region and developmental stage of origin ([11,50],
and A. Bithell and B. P. Williams, unpublished work).
MAKING THE RIGHT TYPE OF NEURON
To illustrate the point above relating PI to subsequent
cell fate, NPCs from the developing midbrain express
the orphan nuclear receptor TF, Nurr1. Nurr1 is also
highly expressed in DA neurons of the midbrain, which
are lost in PD, and it has been shown to instruct midbrain
NPCs to adopt DA neuronal fates. Numerous studies
report the ability of Nurr1 overexpression in CNS
precursors [59] or ES cells [60] to increase the yield of
DA neuronal progeny and even a subsequent ability to
There are numerous caveats with using NSCs or more
restricted progenitors (NPCs) from particular regions of
the CNS to obtain predictable differentiated cell types.
Firstly, although NSCs/NPCs express regional-specific
TFs when isolated from the embryo, their subsequent
propagation and maintenance in vitro reveals that this
does not remain the case; the necessity for exogenous
mitogens for continued NSC proliferation, particularly
FGF2 (as discussed above), has been reported in
numerous studies to alter their intrinsic positional
specification [11,61,62]. A good example of this is
provided by a recent study by Hack and co-workers
[11] involving the generation of neurospheres from the
dorsal or ventral embryonic telencephalon. In vivo,
as dorsal telencephalic NPCs express dorsal-specific TFs,
including Pax6, Emx2 and Ngn1 and Ngn2, and give
rise to glutamatergic neurons and later to astrocytes [63–
66]. Conversely, ventral NPCs express ventral-specific
TFs including Gsh2, Mash1, Dlx2 and Olig2, giving rise
first to GABAergic and cholinergic neurons, and then to
oligodendrocytes later in development [67–71]. Hack and
co-workers [11] reported that, although primary NPCs
isolated from these regions express the appropriate TFs
in vitro and generate the expected types of neurons or
glia, propagation of these cells as neurospheres in the
presence of FGF2 leads to a dysregulation of regional
TF expression. The outcome of this is that dorsal
spheres ectopically express ventral factors, including
Olig2, and produce differentiated cell types with a ventral
phenotype, often GABAergic neurons, while downregulating dorsal-specific TFs. Similarly, a second study
[10], comparing human and mouse embryonic NPCs, has
shown an exclusive bias for the generation of GABAergic
neurons above other types of neuron following repeated
passaging with FGF2 and EGF.
The requirement for FGF2 (or indeed EGF) may
therefore be a barrier to maintaining the positional
specification of isolated NSCs. In particular, neurospheres have increasingly been shown to alter their
gene expression profiles following long-term passage
in vitro, although the extent to which they do this is
still a matter of contention, with conflicting findings
C
2005 The Biochemical Society
17
18
A. Bithell and B. P. Williams
Table 2 Comparison of findings from neurosphere-based experiments to analyse retention of NSC regional specificity
Abbreviations: cHB, caudal hindbrain; CTX, cortex; Dienc, diencephalon; dM, dorsal mesencephalon; E13.5 (etc.), embryonic day 13.5; GE, ganglionic eminence; LGE,
lateral ganglionic eminence; MB, midbrain; MGE, medial ganglionic eminence; Mt, metencephalon; My, myelencephalon; rHB, rostral hindbrain; SC, spinal cord; vM,
ventral mesencephalon.
Authors
Age of isolation
Brain region
Expansion of neurospheres Main findings
Parmar et al. [35]
E13.5
CTX, LGE and MGE
EGF and FGF2 for six
passages
Cells largely retain regional specification (ventral-specific TFs Dlx,
Pbx, Meis and Nkx2.1 predominantly in LGE or MGE and not
CTX).
Hack et al. [11]
E14
CTX
EGF and FGF2 for four
Cells quickly obtain similar TF expression profiles. Most dorsalE14
GE
passages
specific TFs are down-regulated, and ventral-specific Olig1 and
Adult
SVZ
Olig2 are up-regulated.
Hitoshi et al. [72]
E14.5
CTX, GE, MB/rHB,
FGF2 for two passages,
Cells largely retain regional identity appropriate to DV and AP
and cHB
then EGF and FGF2 for
position. Only GE cells display appropriate migratory response
7 days
to local cues in GE slice cultures.
Gabay et al. [61]
E13.5
SC
EGF and FGF2; nonFGF2 up-regulated the expression of Olig2 in Olig2-derived
passaged neurospheres
neurospheres permitting oligodendrocyte differentiation. Olig2
down-regulation permits their differentiation into astrocytes.
Santa-Olalla et al. [62] 1-day intervals from CTX, GE, Dienc, dM, EGF or FGF2 for 7 days
Cells from all regions display ectopic or reduced expression of AP
E11.5–E15.5
vM, Mt, My
and DV markers. Pax6 is up-regulated in all cells, irrespective
and SC
of origin.
either supporting [35,72] or refuting [11,61,62] sustained
regional specification. This is despite the similarities
between starting neural cell populations and culture
conditions in these studies, the main findings of which are
summarized in Table 2. To fully understand the influence
of FGF2 (and EGF), future work must first aim to
reconcile these disparate observations.
Other NSC handling methods also have a dramatic
effect on the repertoire of progeny obtained; these include
cell density and passage number. Furthermore, the age
from which NSCs are isolated affects their intrinsic
capacity to differentiate into particular cell types; a
mechanism established in vivo to generate first one cell
type followed by another in a temporal manner from a
common progenitor. It is therefore essential that we
carefully consider the influence in vitro manipulation
has on PI and cell fate when generating cells for transplantation therapy.
INTRINSIC AND EXTRINSIC FACTORS ABLE
TO ‘PROGRAMME’ NSCs
We already know that regions of the developing brain
are patterned by gradients of different secreted and diffusible signalling molecules and that the TFs regulated
downstream of these molecules, responsible for conferring PI, then specify NPC fate. As discussed above, it has
been suggested recently that mitogens utilized in vitro,
particularly FGF2, might be responsible for altering the
expression of PI genes in NSCs, thus affecting their
capacity to generate particular cell types. Although this
C
2005 The Biochemical Society
may be an impediment to using these in vitro systems to
accurately investigate the basic biology of NSC lineage
determination in vivo, it does suggest a means by which
the intrinsic patterning of an NSC can be overridden
by exposing these cells to specific diffusible signalling
molecules.
Using telencephalic development as an example, glutamatergic neurons of the developing cortex are generated
from the progenitor zones of the cortex. However, most,
if not all, GABAergic neurons of the cortex are generated
in the ventral telencephalon and must follow a tangential
pathway of migration to reach their final destination in
the cortex [73,74]. It is now known that the reason behind
this is that the development of GABAergic neurons
requires expression of TFs that are specifically expressed
in the ventral telencephalon.
The ventralization of the telencephalon requires Shh.
Conversely, members of the WNT (Wingless-related) and
BMP (bone morphogenetic protein) families are involved
in its dorsalization ([75]; for review, see [57]). A recent
study by Gulacsi and Lillien [58] showed that NPCs
from the dorsal telencephalon, which usually give rise
to glutamatergic neurons, could be made to differentiate
into GABAergic interneurons by addition of exogenous
Shh. This ability to alter the fate of dorsal NPCs is
mediated by the antagonistic effect of Shh on BMP4
and, indeed, simply blocking the action of BMP4 in
the absence of Shh was sufficient to induce GABAergic
interneuronal differentiation. Similarly, the levels of BMP
expression can alter the progeny of a Shh-responsive
ventral progenitor cell from GABAergic neurons to
oligodendrocytes [71].
Neural stem cells and cell replacement therapy
Using exogenous signalling molecules, it is thus
possible to modulate the fate of NPCs in vitro. By
investigating the effect of different molecules on the
intrinsic gene expression profile and concomitant cell fate
of an NSC or NPC, we may be able generate neural cells
that differentiate in a specific manner. That we can alter
the intrinsic PI of an NSC or NPC and, hence, its fate
also raises the possibility that a truly specified progenitor
does not exist. Indeed, the specificity of NSCs or NPCs
in vivo may merely reflect their environment.
FUNCTIONAL REPAIR: MAKING
APPROPRIATE CONNECTIONS
Although it is critical for NSC replacement to generate
the appropriate types of neurons or glia once grafted, this
is of little consequence if the cells cannot integrate into the
host system to mediate functional recovery. Currently,
although many in vitro studies show that NSCs generate
functional neurons (for example, neurons that display
action potentials) and glia, there remains relatively sparse
corresponding in vivo evidence.
A recent study by Wernig et al. [7] highlights the ability
of transplanted NSCs to functionally integrate into a
recipient host. In this study, NSCs generated in vitro
from ES cells derived from a tau:EGFP (enhanced green
fluorescent protein) knock-in mouse were transplanted
intraventricularly into E16.5 mouse embryos; all subsequent neuronal progeny could thus be identified by
EGFP expression. Transplanted EGFP-expressing NSCs
not only migrated to diverse regions of the brain but also
integrated functionally: they formed synapses with host
cells, displayed complex morphologies and illustrated
active and passive membrane properties characteristic
of functionally integrated cells. However, the neurons
displayed a range of neurotransmitter subtypes with
no particular regional specificity. Furthermore, analysis
of TF expression revealed that integrated cells showed
a mixture of appropriate and ectopic TF expression.
Although this evidence reveals that appropriate regional
PI is not a prerequisite for functional integration and that
the NSCs are able to generate a diverse range of neurons,
it remains to be shown that the connections made
between host and transplanted neurons are appropriate
and could promote repair. As alluded to previously [11],
it is also evident that NSCs are not homogeneous and
represent a population at different stages of commitment
and therefore with different capacities to respond to
environmental cues.
A more successful approach has recently been reported
using ES cells engineered to express a specific TF. In
this study [76], ES cells were taken from a transgenic
mouse expressing EGFP under control of Hb9, an MNspecific TF, and grown in vitro in conditions to
induce MN differentiation. These MNs subsequently
expressed appropriate receptors and neurotransmitter
phenotype, and when co-cultured with muscle cells
were able to display appropriate action potentials and
functional synapses. Although conducted in vitro, this
work highlights the question of whether or not it is
more beneficial to transplant cells that have already been
specified in vitro to generate the cell types required.
Transplantation of less restricted NPCs (or NSCs) puts
greater emphasis on the requirement of appropriate local
cues for correct differentiation, migration and integration. This is made evident in a study by Han et al. [31],
who reported that NRPs (neuronally restricted progenitor cells) transplanted into the spinal cord can generate
mature neurons, whereas NSCs fail to do so, even
though they can generate neurons if transplanted into
the dentate gyrus [77]. The authors [31] suggest that
this difference is due to the fact that NRPs do not
require neurogenic signals to generate neurons, having
already undergone specification to a neuronal lineage.
In contrast, unspecified NSCs require neurogenic signals
not present or that are inhibited in the normal adult spinal
cord. Instead, NSCs give rise to glial cells, particularly
astrocytes, in response to the gliogenic signals present in
the adult spinal cord.
FUNCTIONAL REPAIR WITHOUT
REPLACEMENT
Although the goal of NSC transplantation discussed
above is achieved by cell replacement, it is important to
briefly discuss the alternatives that have emerged over the
past few years.
An interesting and perhaps unexpected finding came
from a study from Veizovic et al. [48], which involved
transplantation of a conditionally immortalized mouse
NSC line, MHP36, into a rat model of ischaemic injury.
In this model, a lesion is created in one side of the
brain by MCAO (middle cerebral artery occlusion) and
ischaemic damage can be observed by an asymmetry
in the rat’s abilities to remove sticky paper placed on
to its paw. Following transplantation of MHP36 cells,
lesioned rats displayed responses similar to non-lesioned
rats, whereas lesioned sham-grafted animals displayed the
asymmetrical bias. The interesting aspect of this study
came from the discovery that the MCAO-induced lesion
in transplanted animals was significantly reduced, but this
reduction could not be entirely attributed to the transplanted cells replacing lost host cells. Instead, the authors
suggest that the transplanted NSCs acted to preserve and
rescue host cells.
How might NSCs be able to promote repair without
replacement? It is likely that at least one of the ways in
which grafted NSCs can do this is by secreting factors that
C
2005 The Biochemical Society
19
20
A. Bithell and B. P. Williams
are conducive to host cell survival, such as neurotrophic
factors. Furthermore, as discussed earlier, there are regions of the adult brain that contain NSCs and continue to
produce neurons in the adult [the SVZ and the SGL
(subganular layer) of the hippocampus], as well as other
regions from which NPCs have been isolated. Transplanted NSCs may be able to stimulate de novo production of neural cells from these remaining progenitors in
the host brain.
This latter point has fuelled a large amount of research
into the innate ability of the adult brain to self-repair.
Although NPCs and NSCs are known to reside in the
adult brain, with the exception of the SVZ and SGL,
these cells do not normally make new neural cell types in
the adult. We presume that signals able to influence the
fate of these progenitors are simply not expressed in
the adult. Providing these necessary signals to stimulate the brain to manufacture new cells from existing
progenitors offers a conceivable alternative to cell replacement.
FUTURE OF NSC RESEARCH
Although major advances have been made in our understanding of NSC biology, there are still enormous hurdles
before NSC transplantation can ever become commonplace. Specific questions that we need answers for are:
which cells should we use as a starting point? How far
should we restrict an NSC to a specific cell fate before
transplantation? Can we programme cells in vitro to
generate predictable cell types in vivo and how important
is regional specific gene expression for site-specific
differentiation and appropriate functional integration?
We must also not forget that, although human and rodent
NSCs and NPCs share many characteristics, evidence is
plentiful that there are also many differences and a direct
comparison cannot always be made.
Although many questions still remain unanswered,
the wealth of rapidly emerging evidence in this field is
encouraging. Should we be able to gain the necessary
depth of understanding of NSC specification, it is not
inconceivable that routine NSC therapy will be a reality
in the future. In the meantime, we continue to learn more
about the elegant ways in which the complexity of the
adult brain is established from NPCs in the developing
embryo.
ACKNOWLEDGMENTS
We would like to thank Dr Colin Campbell for his helpful
comments on this manuscript. A. B. and B. P. W. are
funded by the BBSRC (Biotechnology and Biological
Sciences Research Council), MRC (Medical Research
Council) and Stanley Foundation.
C
2005 The Biochemical Society
REFERENCES
1 Kordower, J. H., Freeman, T. B., Snow, B. J. et al. (1995)
Neuropathological evidence of graft survival and striatal
reinnervation after the transplantation of fetal
mesencephalic tissue in a patient with Parkinson’s disease.
N. Engl. J. Med. 332, 1118–1124
2 Piccini, P., Brooks, D. J., Bjorklund, A. et al. (1999)
Dopamine release from nigral transplants visualized
in vivo in a Parkinson’s patient. Nat. Neurosci. 2,
1137–1140
3 Evans, M. J. and Kaufman, M. H. (1981) Establishment in
culture of pluripotential cells from mouse embryos.
Nature (London) 292, 154–156
4 Martin, G. R. (1981) Isolation of a pluripotent cell line
from early mouse embryos cultured in medium
conditioned by teratocarcinoma stem cells. Proc. Natl.
Acad. Sci. U.S.A. 78, 7634–7638
5 Rippon, H. J. and Bishop, A. E. (2004) Embryonic stem
cells. Cell Prolif. 37, 23–34
6 Okabe, S., Forsberg-Nilsson, K., Spiro, A. C., Segal, M.
and McKay, R. D. (1996) Development of neuronal
precursor cells and functional postmitotic neurons from
embryonic stem cells in vitro. Mech. Dev. 59, 89–102
7 Wernig, M., Benninger, F., Schmandt, T. et al. (2004)
Functional integration of embryonic stem cell-derived
neurons in vivo. J. Neurosci. 24, 5258–5268
8 Kawasaki, H., Mizuseki, K., Nishikawa, S. et al. (2000)
Induction of midbrain dopaminergic neurons from ES
cells by stromal cell-derived inducing activity. Neuron 28,
31–40
9 Strubing, C., Ahnert-Hilger, G., Shan, J., Wiedenmann, B.,
Hescheler, J. and Wobus, A. M. (1995) Differentiation of
pluripotent embryonic stem cells into the neuronal lineage
in vitro gives rise to mature inhibitory and excitatory
neurons. Mech. Dev. 53, 275–287
10 Jain, M., Armstrong, R. J., Tyers, P., Barker, R. A. and
Rosser, A. E. (2003) GABAergic immunoreactivity is
predominant in neurons derived from expanded human
neural precursor cells in vitro. Exp. Neurol. 182, 113–123
11 Hack, M. A., Sugimori, M., Lundberg, C., Nakafuku, M.
and Gotz, M. (2004) Regionalization and fate specification
in neurospheres: the role of Olig2 and Pax6.
Mol. Cell. Neurosci. 25, 664–678
12 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
13 Wichterle, H., Lieberam, I., Porter, J. A. and Jessell, T. M.
(2002) Directed differentiation of embryonic stem cells
into motor neurons. Cell (Cambridge, Mass.) 110, 385–397
14 Lang, K. J., Rathjen, J., Vassilieva, S. and Rathjen, P. D.
(2004) Differentiation of embryonic stem cells to a neural
fate: a route to re-building the nervous system?
J. Neurosci. Res. 76, 184–192
15 Alvarez-Buylla, A., Seri, B. and Doetsch, F. (2002)
Identification of neural stem cells in the adult vertebrate
brain. Brain Res. Bull. 57, 751–758
16 Capela, A. and Temple, S. (2002) LeX/ssea-1 is expressed
by adult mouse CNS stem cells, identifying them as
nonependymal. Neuron 35, 865–875
17 Rietze, R. L., Valcanis, H., Brooker, G. F., Thomas, T.,
Voss, A. K. and Bartlett, P. F. (2001) Purification of a
pluripotent neural stem cell from the adult mouse brain.
Nature (London) 412, 736–739
18 Sakakibara, S., Imai, T., Hamaguchi, K. et al. (1996)
Mouse-Musashi-1, a neural RNA-binding protein highly
enriched in the mammalian CNS stem cell. Dev. Biol. 176,
230–242
19 Kaneko, Y., Sakakibara, S., Imai, T. et al. (2000) Musashi1:
an evolutionally conserved marker for CNS progenitor
cells including neural stem cells. Dev. Neurosci. 22,
139–153
20 Aubert, J., Stavridis, M. P., Tweedie, S. et al. (2003)
Screening for mammalian neural genes via
fluorescence-activated cell sorter purification of neural
precursors from Sox1-GFP knock-in mice. Proc. Natl.
Acad. Sci. U.S.A. 100, 11836–11841
Neural stem cells and cell replacement therapy
21 Chiasson, B. J., Tropepe, V., Morshead, C. M. and
van der Kooy, D. (1999) Adult mammalian forebrain
ependymal and subependymal cells demonstrate
proliferative potential, but only subependymal cells
have neural stem cell characteristics. J. Neurosci. 19,
4462–4471
22 Doetsch, F., Caille, I., Lim, D. A., Garcia-Verdugo, J. M.
and Alvarez-Buylla, A. (1999) Subventricular zone
astrocytes are neural stem cells in the adult mammalian
brain. Cell (Cambridge, Mass.) 97, 703–716
23 Seaberg, R. M. and van der Kooy, D. (2002) Adult rodent
neurogenic regions: the ventricular subependyma contains
neural stem cells, but the dentate gyrus contains restricted
progenitors. J. Neurosci. 22, 1784–1793
24 Johansson, C. B., Svensson, M., Wallstedt, L., Janson,
A. M. and Frisen, J. (1999) Neural stem cells in the adult
human brain. Exp. Cell Res. 253, 733–736
25 Pagano, S. F., Impagnatiello, F., Girelli, M. et al. (2000)
Isolation and characterization of neural stem cells from the
adult human olfactory bulb. Stem Cells 18, 295–300
26 Roy, N. S., Wang, S., Jiang, L. et al. (2000) In vitro
neurogenesis by progenitor cells isolated from the adult
human hippocampus. Nat. Med. 6, 271–277
27 Roy, N. S., Benraiss, A., Wang, S. et al. (2000) Promotertargeted selection and isolation of neural progenitor cells
from the adult human ventricular zone. J. Neurosci. Res.
59, 321–331
28 Arsenijevic, Y., Villemure, J. G., Brunet, J. F. et al. (2001)
Isolation of multipotent neural precursors residing in
the cortex of the adult human brain. Exp. Neurol. 170,
48–62
29 Nunes, M. C., Roy, N. S., Keyoung, H. M. et al. (2003)
Identification and isolation of multipotential neural
progenitor cells from the subcortical white matter of the
adult human brain. Nat. Med. 9, 439–447
30 Lindvall, O., Kokaia, Z. and Martinez-Serrano, A. (2004)
Stem cell therapy for human neurodegenerative disorders:
how to make it work. Nat. Med. 10, S42–S50
31 Han, S. S., Kang, D. Y., Mujtaba, T., Rao, M. S. and
Fischer, I. (2002) Grafted lineage-restricted precursors
differentiate exclusively into neurons in the adult spinal
cord. Exp. Neurol. 177, 360–375
32 Wang, X., Lu, Y., Zhang, H. et al. (2004) Distinct efficacy
of pre-differentiated versus intact fetal mesencephalonderived human neural progenitor cells in alleviating rat
model of Parkinson’s disease. Int. J. Dev. Neurosci. 22,
175–183
33 Totoiu, M. O., Nistor, G. I., Lane, T. E. and Keirstead,
H. S. (2004) Remyelination, axonal sparing, and locomotor
recovery following transplantation of glial-committed
progenitor cells into the MHV model of multiple sclerosis.
Exp. Neurol. 187, 254–265
34 Nery, S., Fishell, G. and Corbin, J. G. (2002) The caudal
ganglionic eminence is a source of distinct cortical and
subcortical cell populations. Nat. Neurosci. 5, 1279–1287
35 Parmar, M., Skogh, C., Bjorklund, A. and Campbell, K.
(2002) Regional specification of neurosphere cultures
derived from subregions of the embryonic telencephalon.
Mol. Cell. Neurosci. 21, 645–656
36 Horiguchi, S., Takahashi, J., Kishi, Y. et al. (2004) Neural
precursor cells derived from human embryonic brain retain
regional specificity. Neurosci. Res. 75, 817–824
37 Hermann, A., Gastl, R., Liebau, S. et al. (2004) Efficient
generation of neural stem cell-like cells from adult
human bone marrow stromal cells. J. Cell Sci. 117,
4411–4422
38 Joannides, A., Gaughwin, P., Schwiening, C. et al. (2004)
Efficient generation of neural precursors from adult human
skin: astrocytes promote neurogenesis from skin-derived
stem cells. Lancet 364, 172–178
39 Jang, Y. K., Park, J. J., Lee, M. C. et al. (2004) Retinoic
acid-mediated induction of neurons and glial cells from
human umbilical cord-derived hematopoietic stem cells.
J. Neurosci. Res. 75, 573–584
40 McGuckin, C. P., Forraz, N., Allouard, Q. and
Pettengell, R. (2004) Umbilical cord blood stem cells can
expand hematopoietic and neuroglial progenitors in vitro.
Exp. Cell Res. 295, 350–359
41 Sanchez-Ramos, J. R. (2002) Neural cells derived from
adult bone marrow and umbilical cord blood.
J. Neurosci. Res. 69, 880–893
42 Ostenfeld, T., Caldwell, M. A., Prowse, K. R., Linskens,
M. H., Jauniaux, E. and Svendsen, C. N. (2000) Human
neural precursor cells express low levels of telomerase
in vitro and show diminishing cell proliferation with
extensive axonal outgrowth following transplantation.
Exp. Neurol. 164, 215–226
43 Ostenfeld, T., Joly, E., Tai, Y. T. et al. (2002) Regional
specification of rodent and human neurospheres.
Brain Res. Dev. Brain Res. 134, 43–55
44 Villa, A., Navarro-Galve, B., Bueno, C., Franco, S., Blasco,
M. A. and Martinez-Serrano, A. (2004) Long-term
molecular and cellular stability of human neural stem cell
lines. Exp. Cell Res. 294, 559–570
45 Haik, S., Gauthier, L. R., Granotier, C. et al. (2000)
Fibroblast growth factor 2 up regulates telomerase activity
in neural precursor cells. Oncogene 19, 2957–2966
46 Bai, Y., Hu, Q., Li, X., Wang, Y., Lin, C., Shen, L. and
Li, L. (2004) Telomerase immortalization of human neural
progenitor cells. Neuroreport 15, 245–249
47 Grigoryan, G. A., Gray, J. A., Rashid, T., Chadwick, A.
and Hodges, H. (2000) Conditionally immortal
neuroepithelial stem cell grafts restore spatial learning in
rats with lesions at the source of cholinergic forebrain
projections. Restor. Neurol. Neurosci. 17, 1
48 Veizovic, T., Beech, J. S., Stroemer, R. P., Watson, W. P. and
Hodges, H. (2001) Resolution of stroke deficits following
contralateral grafts of conditionally immortal
neuroepithelial stem cells. Stroke 32, 1012–1019
49 Modo, M., Hopkins, K., Virley, D. and Hodges, H. (2003)
Transplantation of neural stem cells modulates
apolipoprotein E expression in a rat model of stroke.
Exp. Neurol. 183, 320–329
50 Akiyama, Y., Honmou, O., Kato, T., Uede, T., Hashi, K.
and Kocsis, J. D. (2001) Transplantation of clonal neural
precursor cells derived from adult human brain establishes
functional peripheral myelin in the rat spinal cord.
Exp. Neurol. 167, 27–39
51 McConnell, S. K. (1995) Strategies for the generation of
neuronal diversity in the developing central nervous
system. J. Neurosci. 15, 6987–6998
52 Lumsden, A. and Krumlauf, R. (1996) Patterning the
vertebrate neuraxis. Science (Washington, D.C.) 274,
1109–1115
53 Tanabe, Y. and Jessell, T. M. (1996) Diversity and pattern in
the developing spinal cord. Science (Washington, D.C.)
274, 1115–1123
54 Jessell, T. M. (2000) Neuronal specification in the spinal
cord: inductive signals and transcriptional codes.
Nat. Rev. Genet. 1, 20–29
55 Marı́n, O. and Rubenstein, J. L. (2001) A long, remarkable
journey: tangential migration in the telencephalon.
Nat. Rev. Neurosci. 2, 780–790
56 Schuurmans, C. and Guillemot, F. (2002) Molecular
mechanisms underlying cell fate specification in the
developing telencephalon. Curr. Opin. Neurobiol. 12,
26–34
57 Campbell, K. (2003) Dorsal-ventral patterning in the
mammalian telencephalon. Curr. Opin. Neurobiol. 13,
50–56
58 Gulacsi, A. and Lillien, L. (2003) Sonic hedgehog and bone
morphogenetic protein regulate interneuron development
from dorsal telencephalic progenitors in vitro. J. Neurosci.
23, 9862–9872
59 Kim, J. Y., Koh, H. C., Lee, J. Y. et al. (2003) Dopaminergic
neuronal differentiation from rat embryonic neural
precursors by Nurr1 overexpression. J. Neurochem. 85,
1443–1454
60 Kim, J. H., Auerbach, J. M., Rodriguez-Gomez, J. A. et al.
(2002) Dopamine neurons derived from embryonic stem
cells function in an animal model of Parkinson’s disease.
Nature (London) 418, 50–56
61 Gabay, L., Lowell, S., Rubin, L. L. and Anderson, D. J.
(2003) Deregulation of dorsoventral patterning by FGF
confers trilineage differentiation capacity on CNS stem
cells in vitro. Neuron 40, 485–499
C
2005 The Biochemical Society
21
22
A. Bithell and B. P. Williams
62 Santa-Olalla, J., Baizabal, J. M., Fregoso, M.,
del Carmen Cardenas, M. and Covarrubias, L. (2003)
The in vivo positional identity gene expression code is
not preserved in neural stem cells grown in culture.
Eur. J. Neurosci. 18, 1073–1084
63 Stoykova, A., Treichel, D., Hallonet, M. and Gruss, P.
(2000) Pax6 modulates the dorsoventral patterning of the
mammalian telencephalon. J. Neurosci. 20, 8042–8050
64 Toresson, H., Potter, S. S. and Campbell, K. (2000) Genetic
control of dorsal-ventral identity in the telencephalon:
opposing roles for Pax6 and Gsh2. Development 127,
4361–4371
65 Sauvageot, C. M. and Stiles, C. D. (2002) Molecular
mechanisms controlling cortical gliogenesis.
Curr. Opin. Neurobiol. 12, 244–249
66 Schuurmans, C., Armant, O., Nieto, M. et al. (2004)
Sequential phases of cortical specification involve
Neurogenin-dependent and -independent pathways.
EMBO J. 23, 2892–2902
67 Anderson, S. A., Qiu, M., Bulfone, A., Eisenstat, D. D.,
Meneses, J., Pedersen, R. and Rubenstein, J. L. (1997)
Mutations of the homeobox genes Dlx-1 and Dlx-2 disrupt
the striatal subventricular zone and differentiation of late
born striatal neurons. Neuron 19, 27–37
68 Casarosa, S., Fode, C. and Guillemot, F. (1999) Mash1
regulates neurogenesis in the ventral telencephalon.
Development 126, 525–534
69 Marı́n, O., Anderson, S. A. and Rubenstein, J. L. R. (2000)
Origin and molecular specification of striatal interneurons.
J. Neurosci. 20, 6063–6076
70 Yun, K., Potter, S. and Rubenstein, J. L. (2001) Gsh2 and
Pax6 play complementary roles in dorsoventral patterning
of the mammalian telencephalon. Development 128,
193–205
71 Yung, S. Y., Gokhan, S., Jurcsak, J., Molero, A. E.,
Abrajano, J. J. and Mehler, M. F. (2002) Differential
modulation of BMP signaling promotes the elaboration of
cerebral cortical GABAergic neurons or oligodendrocytes
from a common sonic hedgehog-responsive ventral
forebrain progenitor species. Proc. Natl. Acad. Sci. U.S.A.
99, 16273–16278
72 Hitoshi, S., Tropepe, V., Ekker, M. and van der Kooy, D.
(2002) Neural stem cell lineages are regionally specified,
but not committed, within distinct compartments of the
developing brain. Development 129, 233–244
73 Anderson, S. A., Mione, M., Yun, K. and Rubenstein,
J. L. R. (1999) Differential origins of neocortical
projection and local circuit neurons; role of Dlx genes
in neocortical interneuronogenesis. Cereb. Cortex 9,
646–654
74 Anderson, S. A., Marı́n, O., Horn, C., Jennings, K. and
Rubenstein, J. L. (2001) Distinct cortical migrations from
the medial and lateral ganglionic eminences. Development
128, 353–363
75 Gunhaga, L., Marklund, M., Sjodal, M., Hsieh, J. C.,
Jessell, T. M. and Edlund, T. (2003) Specification of dorsal
telencephalic character by sequential Wnt and FGF
signaling. Nat. Neurosci. 6, 701–707
76 Miles, G. B., Yohn, D. C., Wichterle, H., Jessell, T. M.,
Rafuse, V. F. and Brownstone, R. M. (2004) Functional
properties of motoneurons derived from mouse embryonic
stem cells. J. Neurosci. 24, 7848–7858
77 Shihabuddin, L. S., Horner, P. J., Ray, J. and Gage, F. H.
(2000) Adult spinal cord stem cells generate neurons after
transplantation in the adult dentate gyrus. J. Neurosci. 20,
8727–8735
Received 20 September 2004; accepted 5 October 2004
Published as Immediate Publication 5 October 2004, DOI 10.1042/CS20040276
C
2005 The Biochemical Society