Tải bản đầy đủ - 0trang
4 From a Flat Neuroectodermal Sheet to a Complex Three-Dimensional Structure: Morphogenetic Transformations Leading to CNS Shaping
Principles of Early Vertebrate Forebrain Formation
acquires a speciﬁc shape and organization, largely under the control of the same
fundamental GRNs that implement regional fate.
At the beginning of CNS morphogenesis the flat and naïve neuroectoderm
extends along the AP axis and, at the same time, its cells compact and intercalate
medio-laterally. This process is known as “convergent extension” and culminates
with the formation of the neural plate (reviewed in Lowery and Sive 2004), which
then folds and gives rise to the neural tube. In mammals and birds, this folding
normally occurs with a speciﬁc sequence and involves the apical constriction of
neural plate cells at speciﬁc points, which favours the longitudinal bending of the
neural tube. Both convergent extension movements and neural plate folding require
the Wnt/planar cell polarity pathway (PCP), which controls similar processes in
other embryonic tissues. In the absence of Wnt/PCP function, convergent extension
is severely impaired, resulting in a wider neural plate that cannot fold (reviewed in
Sokol 2015). This is because, normally, the Wnt/PCP pathway promotes an
actomyosin-dependent contraction of neuroepithelial cells trough the upregulation
of Rho kinase at the apical adherens junctions, which ultimately favours folding and
bending of the neural plate (Nishimura et al. 2012).
In teleost ﬁshes neural tube formation is mechanistically different, because
neural plate cells compact at the midline and form a neural rod. Neuroepithelial
cells in the rod then undergo cell division at the midline, after which the daughter
cells rearrange at both sides of the midline and orient their apical sides towards the
centre of the rod, establishing the apical (luminal) side of the forming neural tube
(Ciruna et al. 2006; Tawk et al. 2007). The molecular mechanisms involved in this
neuroepithelial condensation and lumen formation are currently unclear, but likely
involve the control of cytoskeleton dynamics and extracellular matrix remodelling
(Araya et al. 2014; Buckley and Clarke 2014; Clarke 2009).
While most of the neural plate gives rise to an almost straight tube, the anterior
part of the neural primordium undertakes a more complex reorganisation according
to the convoluted structure of the brain. The most notable initial rearrangement of
forebrain morphogenesis consists in the bulging of the eye primordia from its lateral
walls. Recent studies have exploited the advantages of teleost ﬁsh to image this
event in vivo, showing that, at least in this species, cells fated to become eye
precursors are highly cohesive and strictly segregated from those of the surrounding
domains (Cavodeassi and Houart 2012). Both phenomena, cohesion and segregation, are promoted by the combined function of Cxcr4, the Eph/Ephrin and the
Wnt/noncanonical signalling pathways (Bielen and Houart 2012; Cavodeassi et al.
2005, 2013). As they evaginate, the eye-ﬁeld cells extensively rearrange, intercalate
among each other (Ivanovitch et al. 2013; Rembold et al. 2006) and, at the same
time, elongate and polarise to establish the tight neuroepithelial structure of the eye
primordia, also known as optic vesicles (Ivanovitch et al. 2013). This acquisition of
apico-basal polarity of the optic vesicle cells, marked by the onset of pard6cb
expression, an apical polarity marker, and by the localised accumulation of
Laminin1 around the eye primordium, is required for an accurate vesicle evagination as interference with polarization events disrupts proper optic vesicle formation (Ivanovitch et al. 2013).
F. Cavodeassi et al.
As optic vesicles form, telencephalic cells also undergo morphogenetic changes
with at least two differences from eye precursors: they converge fast towards the
midline and polarize later (Rembold et al. 2006). This differential cell migration is
effected, at least in part, by the cadherin-like molecule Nlcam, which is respectively
expressed at high and low levels in the telencephalic and eye precursors (Brown
et al. 2010). Increasing Nlcam expression in the eye ﬁeld makes its cells converging
towards the midline instead of evaginating, thus behaving like telencephalic progenitors. Polarisation and migration differences between eye and telencephalic cells
are ultimately controlled by the GRN that directs their respective fate. For example,
Rx3, one of the “kernels” for eye speciﬁcation and morphogenesis, regulates the
expression of pard6 and nlcam (Brown et al. 2010; Ivanovitch et al. 2013), as well
as that of cxcr4a, ephA4/B4 and the Wnt/noncanonical receptor fzd5, therefore
forming a GRN necessary to control the segregation of the eye ﬁeld from the
surrounding neural plate territories (Fig. 11.3) (Bielen and Houart 2012;
Cavodeassi et al. 2013 and our unpublished results; see also Chap. 9).
Forebrain morphogenesis in organisms others than teleost ﬁshes has not been
analysed in detail mainly due to imaging limitations. It is thus unclear whether
similar mechanisms operate across species. As neural tube formation in birds and
mammals involves neural plate folding instead of rod cavitation, it is possible that
cell speciﬁcation/polarization/movement occur with a different sequence and certainly at a different pace. Nevertheless, mouse embryonic stem cells have the
spontaneous ability to generate optic cups, when cultured in the presence of the
appropriate factors and in a three-dimensional matrix rich in Laminin1 (Eiraku et al.
2011). This process involves the polarisation of the actin cytoskeleton in the
evaginating cells (Eiraku et al. 2011), suggesting some mechanistic conservation
across species. Our increasing ability to reproduce organ morphogenesis in vitro
constitutes an excellent opportunity to fully understand conserved and divergent
aspects of vertebrate forebrain formation (reviewed in Sasai et al. 2012).
Fig. 11.3 A basic and simpliﬁed GRN involved in eye ﬁeld morphogenesis. The network has
been depicted with BioTapestry software (Longabaugh et al. 2005) using information from
zebraﬁsh. Rx3 is represented as a “kernel” gene that coordinates eye morphogenesis and its
segregation from the telencephalic and diencephalic territories
Principles of Early Vertebrate Forebrain Formation
Post-transcriptional Control: The Role of miRNAs
The networks described so far are linked to the transcriptional control of gene
expression. However, there is mounting evidence that post-transcriptional mechanisms are additional essential pieces of the regulatory landscape in both embryonic
development and tissue homeostasis. The mechanism that is currently receiving
most attention implicates the activity of microRNAs, a family of non-coding RNAs.
miRNAs are single-stranded RNAs of around *22 nucleotides, the generation of
which involves the transcription of their respective genes into a primary transcript
that is processed in the nucleus into an approximate 70 nucleotides long
pre-miRNA (Lee et al. 2002). This pre-miRNA is further sequentially processed
into a mature double-stranded RNA by two endoribonucleases, known as Drosha
and Dicer (Hutvagner et al. 2001; Lee et al. 2003). One of the strands of the mature
miRNA is then incorporated into a silencing ribonucleo-protein complex called
RISC (RNA-induced silencing complex) that binds to the complementary seed
sequences present in the 3′ UTR of the multiple target mRNAs (Hammond et al.
2000; Schwarz et al. 2003). Once bound to the target mRNAs, the catalytic activity
of the RISC, known as Argonaute, destabilises the mRNAs or inhibits their
translation (Makeyev et al. 2007; Valencia-Sanchez et al. 2006), thereby controlling
the amount of the corresponding protein ﬁnally available in a tissue.
miRNAs show very dynamic expression patterns from early embryogenesis to
adult tissues in most organs and species so far analysed (Kapsimali et al. 2007).
Their function is essential for early development as demonstrated by the evolutionary conserved existence of both maternal and zygotic Dicer transcripts, in the
absence of which the formation of mature miRNAs is abolished. Zygotic dicer null
mutant mice show an embryonic lethal phenotype at gastrulation stages (Bernstein
et al. 2003), whereas in absence of maternal Dicer mouse zygotes do not complete
the ﬁrst cell division (Tang et al. 2007). In zebraﬁsh instead loss of function of both
maternal and zygotic Dicer leads to abnormal gastrulation and severe alterations in
morphogenesis (Giraldez et al. 2005), whereas zygotic dicer mutants develop
normally until late larva stages, suggesting that the requirement of miRNAs
function during gastrulation is not conserved in teleosts (Wienholds et al. 2003).
About 70 % of all known miRNAs are expressed in the CNS (Diaz et al. 2014),
although most of them seem to play roles at late stages of CNS differentiation
(Kawase-Koga et al. 2009), when the ﬁne regulation of mRNA products is perhaps
most necessary to generate neuronal diversity. A few studies have nevertheless
identiﬁed miRNAs involved in early stages of CNS speciﬁcation, patterning and
morphogenesis. miR-96, miR-290–295 and miR-200 promote ectoderm versus
neuroectoderm fate speciﬁcation by limiting the amount of the neural TF Pax6 (Du
et al. 2013; Kaspi et al. 2013), which, in turn, activates miR-135b. The latter then
down-regulates TGF-b/BMP signalling and therefore locks neuroectodermal fate
(Bhinge et al. 2014). Components of the Wnt/b-catenin pathway that, as we have
already mentioned, controls all these events, are also conserved miRNAs targets.
For instance, miR-34 targets b-catenin in Xenopus, thereby regulating the
F. Cavodeassi et al.
expression of downstream genes and the correct establishment of axis polarity (Kim
et al. 2011). An additional example of how miRNAs impact on AP patterning is
provided by their role in reﬁning and maintaining the function of secondary
organisers, such as in the case of the MHB. miR-9 is expressed around this organizing boundary, where it targets several transducers of the FGF signalling pathway
(Leucht et al. 2008), limiting its signalling effects. As miRNAs usually bind several
distinct mRNAs, miR-9 seems also to reduce the activity of neurogenic genes at the
MHB, maintaining this territory in an undifferentiated state essential for its function
as an organiser (Leucht et al. 2008).
Quite likely miRNAs participate in the regionalization of the entire forebrain
primordium, but at the moment, most studies have focused on the eye primordium,
which expresses several miRNAs and is severely affected by Dicer inactivation.
Dicer deletion causes microphthalmia (reduction of the eye size) affecting the lens
placode, the neural retina, the pathﬁnding of the retinal ganglion cell axons as well
as the pigmentation and adhesion of the Retinal Pigment Epithelium (RPE), which,
in turn, affect photoreceptors’ maturation (Conte et al. 2013; Ohana et al. 2015).
Besides the general demonstration that mRNA silencing is relevant for eye speciﬁcation, knock-down/out studies are beginning to delineate the speciﬁc function of
each miRNA in eye formation. Among them, miR-124 and miR-204 are particularly important. miR-124 maintains optic vesicle cell proliferation at early stages of
development by turning off the proneural gene neuroD1. This early function prevents the onset of neurogenesis (Liu et al. 2011). Later on miR-124 promotes
differentiated cone photoreceptor survival by targeting the TF Lhx2 (Sanuki et al.
2011). miR-204 instead modulates the levels of the TF Meis2, which is upstream of
Pax6 in the GRNs controlling morphogenesis and speciﬁcation of both the lens and
the retina (Conte et al. 2010). Consistent with the general observation that miRNAs
have rather heterogeneous targets, slightly later, miR-204 targets EphB2 and EfnB3
(Conte et al. 2014), a signalling system implicated in retinal ganglion cell axon
pathﬁnding, as well as effector genes of RPE differentiation (Adijanto et al. 2012).
Many more studies are needed to fully understand how miRNAs contribute to
forebrain development. Nevertheless, it is becoming apparent that many miRNAs
can contribute to the regulation of the same process and also that each miRNA, is
recurrently used during development for different purposes, further contributing to
diversify the GRNs that lead to a mature forebrain.
In conclusion, in this chapter we have provided a general and simpliﬁed view of the
principles that govern early forebrain development. This information derives from a
huge number of studies based on experimental manipulations of gene activity in
different vertebrate species, of which unfortunately we could not give a full account
here. These studies have been facilitated by the sequencing of several genomes,
which also led to the identiﬁcation of a large number of non-coding RNAs, as well