6
NEURAL
CONTROL OF POST-PHYLOTYPIC DEVELOPMENT
Perhaps the most interesting thing about having a hormonal
regulation of development is that development comes under
the control of the central nervous system.
H.F. Nijhout
After
the phylotypic stage, with the formation of the Bauplan of
the phylum and the development of the operational CNS in
metazoans, the reserve of parentally provided epigenetic
information is exhausted. At this juncture, the embryo
emerges as an informationally self-reliable organism,
capable of generating the huge amount of epigenetic
information necessary for determining the complex spatial
arrangement of the myriad of cells for building
supracellular structures in the process of organogenesis and
morphogenesis. Adequate experimental evidence shows that the
embryonic and young CNS takes over the post-phylotypic
development up to adulthood: the CNS becomes the exclusive
source of the epigenetic information that via signal
cascades is transmitted for the development of tissues,
organs and animal morphology in general.
Neural Control of Post-phylotypic
Development
Early stages
of development up to the formation of gastrula and neurula may
considerably vary between species and other taxa of a phylum.
But, after that point in their embryonic development all of
them converge to a morphologically common stage known as
phylotypic stage, characterized by a shared Bauplan. After the
phylotypic stage, organisms of various taxa may again diverge
considerably from the common phylotypic morphology.
Before the
phylotypic stage only a few inductive events take place based
on global interactions (such as the establishment of body
axes, formation of mesoderm, gastrulation, and induction of
the nervous system), and all of them are epigenetically
determined by parental epigenetic information (cytoplasmic
factors). But, after that stage of morphological convergence
multiple inductions take place all over the embryo, leading to
formation of organs, organ systems, and other metazoan
structures. Clearly, parentally inherited concentration
gradients that might have made the early development possible
are no longer suitable to determine the spatially complex
patterns of organs and systems of organs. Their development
implies the presence and use of an enormous amount of
positional information which cannot be provided by
concentration gradients of morphogens.
The post-phylotypic
development is defined here in the broadest sense so as to
comprise all the morphological, physiological, and behavioral
transformations that characterize the individual development
from the phylotypic stage to adulthood.
As already
mentioned, no matter how different developmental pathways in
the early embryonic development may be, all of them converge
to a common Bauplan at the phylotypic stage (extended germ
band stage in insects and tail bud stage in vertebrates). At
this juncture, any effect of parentally provided epigenetic
information (cytoplasmic factors) on the embryonic development
terminates.
What is the
origin of the huge amount of information necessary for the
postphylotypic development after the exhaustion of parental
epigenetic information at the phylotypic stage? The only known
important event coincident with the exhaustion of the
parentally provided epigenetic information (parental
cytoplasmic factors) that might be relevant to the above
question is probably the development of the operational CNS
and the related ICS (integrated control system).
Neuroblasts in
the neural tube are the first fully differentiated cells to
appear in Drosophila embryos and the nervous system is
the first organ system to develop and function in metazoan
embryos. Temporally, it precedes or coincides with the
beginning of embryonic inductions and organogenesis. Formation
of somites, limb buds, gonads, kidney, lung and heart take
place only after, and in relation to, the formation of the CNS
(Hall, 1998f).
Before
presenting the representative evidence on the role of the CNS
as the originator of inductive signals for metazoan
organogenesis and morphogenesis, I will briefly review the
postphylotypic development in invertebrates and vertebrates,
focusing on the phenomenon of metamorphosis, a widespread mode
of individual development in invertebrates (and vertebrate
amphibians) that represents an impressive manifestation of the
control of the postphylotypic development by the CNS as
controller of the ICS (integrated control system). As a form
of individual development, metamorphosis is more accessible to
observation, hence suitable for investigating the mechanisms
of individual development. It offers a simpler picture of the
flow of information from the CNS for the postphylotypic
development in metazoans.
Neural Control of
Metamorphosis in Invertebrates
All types of metamorphosis in
invertebrates and vertebrates, however different they might
appear, have in common apoptosis, programmed death of cells,
as a special morphogenetic mechanism for sculpting their
structure by getting rid of parts and organs of previous
stages and for developing the adult morphology. Hence, a
glimpse of the mechanisms of control of apoptotic processes is
necessary for understanding the nature of metamorphosis in
invertebrates.
Apoptosis in Invertebrates
Apoptosis, as a form of programmed cell death was described
first by C. Vogt (Seipp et al., 2001) in 1842. It is essential
for morphogenesis and organogenesis in all metazoans,
invertebrates and vertebrates alike. While metazoan cells
might have retained the constitutive mechanism of apoptosis,
apoptosis in metazoans is primarily a regulated phenomenon,
indispensable during the embryonic development and all the
later stages of life. It is a sculpting developmental tool for
“the elimination of superfluous cells, morphogenetic changes,
and hollowing out solid structures to form cavities or tubes.”
(Lassus et al., 2002).
The proximate agents of the apoptotic cell death are a
number of proteases (5 proteases in Drosophila),
known under the common name of caspases. The process starts
with activation of “initiator” caspases (caspases 2, 8, 9)
followed by activation of effector caspases (caspases 3, 6,
and 7), which act proteolytically on cell proteins and
enzymes responsible for cell metabolism and reproduction.
Initiator caspases cause the permeabilization of
mitochondria which simply amplify the caspase activity but
have no regulatory function in apoptosis (Lee and Baehrecke,
2001). The dead cells then are endocytically engulfed by
macrophages.
In Drosophila, products of 3 genes, rpr
(reaper), grim (genes associated with
retinoid-IFN-induced mortality),
and hid (head involution defective), are known
to act as signals for activating the downstream mechanism of
apoptosis and two other proteins, (dIAP1 and dIAP2), act as
inhibitors of apoptosis (Lassus et al., 2002). The
expression of these preapoptotic genes in Drosophila
always follows the ecdysone pulses, mediated by specific
nuclear receptors, EcRs. These pulses induce a few early
response genes, such as BR-C (Broad-Complex ),
E74, and E75, whose products (transcription
factors), in turn, activate more than 100 late response
genes. Downstream the EcR, and upstream the H99 region that
contains the rpr, grim, and hid
preapoptotic genes, is the early response gene E93 (Jiang et
al., 1997). During the Drosophila metamorphosis
ecdysone pulses direct the destruction of obsolete larval
tissues and their replacement by tissues and structures that
form the adult fly...via the precise stage- and
tissue-specific regulation of key death effector genes. (Draizen
et al., 1999)
In other words, during Drosophila metamorphosis,
ecdysone controls apoptosis by regulating expression of
preapoptotic genes (Draizen et al., 1999; Namba et al.,
1997). But it is a well-known and well-established fact that
the ecdysone pulses and ecdysone synthesis in insects are
under strict control of the CNS (figure 6.1).
Figure 6.1.
Signal
cascade for the Drosophila apoptotic pathway starts
in the CNS. Death signals induce ecdysone, Rpr, Grim, and
Hid, which lead to the activation of caspases. Dronc, is
activated by the CED-4/Apaf-1 homolog Dark, which is
required for Rpr-, Hid-, or Grim-induced cell death. Diaps
act by binding to procaspases and by preventing their
activation. Rpr, Hid, and Grim, by binding to Diap1, are
thought to disrupt inhibitor of apoptosis-caspase complexes,
leading to caspase activation. The baculovirus protein P35
acts to inhibit many caspases but has not yet been shown to
directly inhibit Dronc. In the Dronc pathway, P35 may
function by inhibiting a downstream caspase, such as Drice
(Modified from Quinn et al., 2000).An excess of cells is
always produced during the embryonic development in
metazoans, but the embryo gradually eliminates the
unnecessary cells in the processes of morphogenesis and
organogenesis. So, e.g. in response to hyperplasia induced
by experimental increase of the cycline expression,
Drosophila embryo proportionately increases apoptotic
processes. Similarly, increased bicoid expression in
Drosophila leads to the development of a larger head,
which is corrected later by apoptotic death of excess cells
(From Namba et al., 1997).
It is clear from such examples that somehow the embryo
identifies the excess cells (but no individual cell can figure
it out) and corrects the state by apoptotically eliminating
them. Furthermore, metazoans seem to control not only the
number of their cells, but by neural mechanisms they also
control the size of their body (see later on the control of
body size) by making the necessary adjustments in the number
or the size of their cells (cells of pentaploid salamanders
have five times the volume of the cells of haploid species but
both of them have approximately the same size because the
pentaploid species have 5 times less cells in their body).
What estimates the number of cells and perceives excess cells
that have to be eliminated? What sends the apoptotic message
to the organs and tissues that will undergo apoptosis? It is
clear that it is not the genes, not individual cells but a
supracellular entity capable of monitoring the status of the
embryonic organism as a whole.
Neural Control of Metamorphosis in Cnidarians
After fertilization, eggs of Hydractinia echinata start
the cleavage. The embryo has only two germ layers. Within 2-3
days from fertilization, a ciliated spindle-shaped planula
develops (Walther et al., 1996), which stops cell
proliferation and cell differentiation but is competent of
starting metamorphosis (Plickert et al., 1988). The ~1mm long
planula, consisting
of
~10,000 cells, settles before starting
metamorphosis.
The environmental cue inducing the competent larva to
enter metamorphosisze
and transform into a primary polyp is a chemical signal
released by a bacterium, but metamorphosis can be induced by a
variety of other chemical agents. The bacterial signal is
received by neurosecretory cells
at
in the anterior part of the planula.
According to B. Schwoerer et al. (1990)there
the external cue triggers secretion of an internal signal,
which is transmitted all the way to the posterior end and
“synchronizes the events of metamorphosis” (Schwoerer-Bohning
et al. 1990). Hydractinia echinata starts cell
proliferation by the 9th hour after induction of
metamorphosis, starting from the middle gastric region (Plickert
et al., 1988).
In 1981 a “morphogenetic peptide” was isolated from Hydra
attenuata and Anthopleura elegantissima that was
characterized as “head-inducing morphogen” (Schaller and
Bodenmuller, 1981). This is a neuropeptide synthesized by, and
stored in, the secretory granules of the cnidarian neurons.
Another isolated and identified peptide is the neuropeptide
metamorphosin A (MMA), which induces metamorphosis of the
treated intact larvae.
In 1981 a “morphogenetic peptide” was isolated from Hydra
attenuata and Anthopleura elegantissima that was
characterized as “head-inducing morphogen” (Schaller and
Bodenmuller, 1981). This is a neuropeptide synthesized by,
and stored in, the secretory granules of the cnidarian
neurons.
Another isolated and identified peptide is the neuropeptide
metamorphosin A (MMA), which induces metamorphosis of the
treated intact larvae.
”MMA” is among
the most potent peptides known to have a function in
controlling animal development. (Leitz et al., 1994)
The subsequent discovery of the role of GLWamide
neuropeptides in metamorphosis of Hydractinia echinata
led to the notion of the existence in these lower
invertebrates of a neuropeptide signal system controlling
metamorphosis (Schmich et al., 1998). In H. echinata,
sensory neurons secrete two antagonistically acting
neuropeptides: GLWamides (stimulators of metamorphosis) and
RFmides (inhibitors of metamorphosis) in adaptive responses
to environmental conditions (Katsukara et al., 2003).
Hydra neurons secrete another neuropeptide, the “head
activator”, which is a signal molecule (growth hormone)
(Schaller, 1976a). The neuropeptide determines the
transformation of the interstitial cells into nerve cells
(Schaller, 1976b). The head activator binds to a receptor
protein, HAB (head-activator binding protein),
Based on the experimental observation that the
neurotransmitter serotonin stimulates metamorphosis in the
planulae of the cnidarian Eudendrium racemosum, which
normally does not metamorphose, as well as in other
evidence, it is concluded that serotonin is involved in the
perception of the metamorphosis-triggering environmental
cue, through serotonin containing sensory neurons in hydroid
planulae of different species (Zega et al., 2007).
Even this glimpse of signals inducing morphogenesis and
metamorphosis in Hydra shows that all of them are
neuropeptides secreted by secretory neurons of
the neural net in adaptive responses to environmental cues.
Neural Control of
Metamorphosis in Molluscs
In molluscs
metamorphosis depends
chiefly on the activity of neurotransmitters that the
nervous system releases in response to external cues. A
metamorphosis-inducing role is demonstrated for several
neurotransmitters in this group.
The free-swimming veliger (a stage in mollusc metamorphosis)
larvae have a functioning nervous system with the rudiments
of almost all the adult ganglia and an apical sensory organ
(ASO) (Lacalli, 1994), containing, among other cells,
serotoninergic secretory neurons. These neurons receive
environmental cues, integrate and process them, and respond
by releasing neurotransmitters, such as serotonin and/or
catecholamines (epinephrine, norepinephrine, and dopamine),
GABA (gamma-amino butyric acid), etc., all of them
with proven metamorphosis-inducing action in molluscs.
Molluscan larvae exposed to these neurotransmitters are
induced to metamorphose. So, e.g., application of serotonin
induces metamorphosis in 80-100% of the competent larvae of
coenogastropod Ilyanassa obsoleta (Leise et al.,
2001). Induction of metamorphosis by high levels of NE (norepinephrine)
and DA (dopamine) in metamorphically competent Phestilla
larvae (Pires et al., 1997) and by DOPA (L-3,
4-dihydroxyphenilalanine) in Crepidula fornicata
(Pires et al., 2000), as well as inhibition of metamorphosis
after depletion of these neurotransmitters indicate that
neurotransmitters may participate in the control of
gastropod development. (Pires et al., 1997)
Summarizing, it may be said that metamorphosis in molluscs
is under direct control of neurotransmitters released by
secretory neurons.
Neural Control of Metamorphosis in Cnidarians
After fertilization, eggs of Hydractinia echinata start the
cleavage. The embryo has only two germ layers. Within 2-3
days from fertilization, a ciliated spindle-shaped planula
develops (Walther et al., 1996), which stops cell
proliferation and cell differentiation but is competent of
starting metamorphosis (Plickert et al., 1988). The ~1mm
long planula, consisting of ~of only ~10,000 cells, settles
before starting metamorphosis.
The environmental cue inducing the competent larva to enter
metamorphosisze and transform into a primary polyp is a
chemical signal released by a bacterium, but metamorphosis
can be induced by a variety of other chemical agents. The
bacterial signal is received by neurosecretory cells at in
the anterior part of the planula. According to B. Schwoerer
et al. (1990) There the external cue triggers secretion of
an internal signal, which is transmitted all the way to the
posterior end and “synchronizes the events of metamorphosis”
(Schwoerer-Bohning et al. 1990). Hydractinia echinata starts
cell proliferation by the 9th hour after induction of
metamorphosis, starting from the middle gastric region (Plickert
et al., 1988).
In 1981 a “morphogenetic peptide” was isolated from Hydra attenuata and
Anthopleura elegantissima that was
characterized as “head-inducing morphogen” (Schaller and
Bodenmuller, 1981). This is a neuropeptide synthesized by,
and stored in, the secretory granules of the cnidarian
neurons.
Another isolated and identified peptide is the neuropeptide
metamorphosin A (MMA), which induces metamorphosis of the
treated intact larvae.
”MMA” is among the most potent peptides known to have a
function in controlling animal development. (Leitz et al.,
1994)
The subsequent discovery of the role of GLWamide
neuropeptides in metamorphosis of Hydractinia echinata led
to the notion of the existence in these lower invertebrates
of a neuropeptide signal system controlling metamorphosis (Schmich
et al., 1998). In H. echinata, sensory neurons secrete two
antagonistically acting neuropeptides: GLWamides
(stimulators of metamorphosis) and RFmides (inhibitors of
metamorphosis) in adaptive responses to environmental
conditions (Katsukara et al., 2003).
Hydra neurons secrete another neuropeptide, the “head
activator”, which is a signal molecule (growth hormone)
(Schaller, 1976a). The neuropeptide determines the
transformation of the interstitial cells into nerve cells
(Schaller, 1976b). The head activator binds to a receptor
protein, HAB (head-activator binding protein), that which is
present both as a secreted protein and as a
membrane-anchored receptor (Hampe et al., 1999). The
neuropeptide is found to help regeneration of Hydra
tentacles by stimulating cell division in damaged tentacles.
It also stimulates bud formation.
Based on the experimental observation that the
neurotransmitter serotonin stimulates metamorphosis in the
planulae of the cnidarian Eudendrium racemosum, which
normally does not metamorphose, as well as in other
evidence, it is concluded that serotonin is involved in the
perception of the metamorphosis-triggering environmental
cue, through serotonin containing sensory neurons in hydroid
planulae of different species (Zega et al., 2007).
Even this glimpse of signals inducing morphogenesis and
metamorphosis in Hydra shows that all of them are
neuropeptides, i.e. peptides secreted by secretory neurons
of the neural net in adaptive responses to environmental
cues.
Neural Control of Metamorphosis in Molluscs
In molluscs as well, metamorphosis depends chiefly on the
activity of neurotransmitters that the nervous system
releases in response to external cues. A
metamorphosis-inducing role is demonstrated for several
neurotransmitters in this group.
The free-swimming veliger (a stage in mollusc metamorphosis)
larvae have a functioning nervous system with the rudiments
of almost all the adult ganglia and an apical sensory organ
(ASO) (Lacalli, 1994), containing, among other cells,
serotoninergic secretory neurons. These neurons receive
environmental cues, integrate and process them, and respond
by releasing neurotransmitters, such as serotonin and/or
catecholamines (epinephrine, norepinephrine, and dopamine),
GABA (gamma-amino butyric acid), etc., all of them with
proven metamorphosis-inducing action in molluscs.
Molluscan larvae exposed to these neurotransmitters are
induced to metamorphose. So, e.g., application of serotonin
induces metamorphosis in 80-100% of the competent larvae of
coenogastropod Ilyanassa obsoleta (Leise et al., 2001).
Induction of metamorphosis by high levels of NE (norepinephrine)
and DA (dopamine) in metamorphically competent Phestilla
larvae (Pires et al., 1997) and by DOPA (L-3,
4-dihydroxyphenilalanine) in Crepidula fornicata (Pires et
al., 2000), as well as inhibition of metamorphosis after
depletion of these neurotransmitters indicate that
neurotransmitters may participate in the control of
gastropod development. (Pires et al., 1997)
Summarizing, it may be said that metamorphosis in molluscs
is under direct control of neurotransmitters released by
secretory neurons.
Neural Control of Metamorphosis in Insects
Two major hormones, Ec (ecdysone) and JH (juvenile hormone),
control the complex processes of metamorphosis in insects.
Extensive studies carried out especially during 2-3 last
decades show that hormonal response in insects is under
strict cerebral control.
Ecdysone is a steroid hormone secreted by prothoracic gland
that, in its active form, stimulates metamorphosis and
regulates molting in insects. In the tobacco hornworm,
Manduca sexta, these hormones control metamorphosis,
especially processes of muscle degeneration and programmed
neuron death (Weeks and Truman 1986). In Drosophila, ecdysone
induces the histolysis of nearly all of the larval tissues
and differentiation and morphogenesis of the structures
composing the adult fly. (Baehrecke, 1996)
These functions are mediated by its heteromeric receptor,
EcR, which is implicated in the reorganization of imaginal
and larval tissues at the onset of metamorphosis (Li and
Bender, 2000). Let’s remember that secretion of ecdysone by
the prothoracic gland is regulated by a brain neuropeptide,
the PTTH (prothoracicotropic hormone) (in the silkworm, the
brain secretes bombyxin, an additional neuropeptide with
ecdysone-stimulating activity). Altered neuropeptide
synthesis leads to low levels of ecdysteroids, and delay or
block of metamorphosis in Drosophila mutants (Zitnan et al.,
1993). In the butterfly Precis coenia, the neuropeptide
bombyxin acts together with ecdysone to stimulate growth of
the wing imaginal discs and
The level of bombyxin in the hemolymph is modulated by the
brain in response to variation in nutrition and is part of
the mechanism that coordinates the growth of internal organs
with overall somatic growth. (Nijhout and Grunert, 2002)
Ecdysone also stimulates the growth of cuticle in each
larval stage. After specific developmental transformations,
the insect larva must shed the cuticle in order to enter the
next developmental stage. The process of cuticle shedding
requires strict coordination, transformation, and loosening
of the cuticle and a stereotyped sequence of preecdysis and
ecdysis behaviors. The specific ecdysis behavior is
regulated by the hormone ETH (ecdysis-triggering hormone)
released by the endocrine Inka cells, under stimulation of
the neurohormone EH (eclosion hormone) released by EH
neurons (figure 6.2). The absence of ETH in Drosophila null
mutants for the respective hormone leads to “incomplete ecdysis and 98% mortality at the transition from the first
to second larval instar.” When those mutants are treated
with a synthetic ETH they start metamorphosis (Yonseong et
al., 2002). ETH and crustacean cardioactive peptide (CCAP)
secreted by CCAP neurons elicit the first two motor
behaviors, the pre-ecdysis and ecdysis behaviors (Gammie and
Truman, 1997). In response to ETH, specific neurons in the
brain secrete the eclosion hormone (EH) which, in turn,
increases secretion of the neuropeptide CCAP by CCAP
neurons:
The sequential performance of the two behaviors arises from
one modulator activating the first behavior and also
initiating the release of the second modulator. The second
modulator, then turns off the first behavior while
activating the second. (Gammie and Truman, 1997).
The endocrine Inka cells of the insect epitracheal glands,
at the end of each developmental stage, respectively secrete
the pre-ecdysis- and ecdysis-triggering hormones (PETH and
ETH), to which each abdominal ganglion respond by starting,
within a few minutes, preecdysis II and I, respectively.
The initiation of preecdysis and the transition to ecdysis
are regulated by stimulatory and inhibitory factors released
within the central nervous system after the initial actions
of PETH and ETH. (Zitnan and Adams, 2000)
These hormones determine the sequential stereotyped
behaviors of cuticle shedding, but ETH secretion itself is
controlled by the eclosion neurohormone, EH (Kingan et al.,
2001), secreted by two pairs of ventromedial (VM) brain
neurons. EH released by the neurosecretory VM cells in the
brain stimulatescells in the brain stimulates the release of
ETH by Inka cells, which, via a positive feedback loop,
stimulates the neurosecretory cells to secrete more EH. This
causes activation of peptidergic neurons that release CCAP
(crustacean cardioactive peptide) (Ewer et al., 1997)
(figure 6.2).
Figure 6.2. Diagram showing the neuromodulator
pathways controlling pre-ecdysis and ecdysis behaviors.
Release of ETH from the Inka cells both initiates
pre-ecdysis and excites the ventromedial (VM) neurons
that contain eclosion hormone. Because they are part
of a positive feedback loop, the Inka cells and
VM neurons release almost all of their peptide
stores. EH release within the CNS triggers cGMP
upregulation in the Cell 27/704 group, causing the central
and peripheral (not shown) release of CCAP.
Centrally released CCAP both activates the
ecdysis motor program and terminates pre-ecdysis.
Sensory input (possibly from bristle hairs deformed by the
pressure of the old cuticle) may maintain
excitation of the Cell 27/704 group to insure
CCAP release and the continuation of ecdysis until
the cuticle is shed. Removal of the cuticle eliminates
the sensory input, resulting in the cessation of
CCAP release and of ecdysis behavior (From Gammie
and Truman, 1997).
The pre-ecdysis, but not ecdysis, can be experimentally
elicited in isolated abdomens simply by injecting the
neurohormone EH.
The idea that secretion of ETH by Inka cells is under
control of the eclosion hormone (EH) has been challenged
recently. Although EH stimulates secretion of ETH by Inka
cells, these cells secrete ETH even in the absence of EH
(Clark et al., 2004). However, the central neural control
of the initiation of the ETH secretion may be exerted in
an alternative way. Indeed, it is observed that insects
cannot initiate ecdysis behavior if their brain is
disconnected, even when EH is injected. Novicki and Weeks
(1996) assume that
Neural input from the brain to the abdominal ganglia is
necessary to initiate ecdysis. (Nowicki and Weeks, 1996)
In the tobacco hawkmoth, Manduca sexta wing expansion
and cuticle tanning takes place in the posteclosion
period. Both these processes are neurally determined by
the secretion of the neurohormone bursicon in BAG neurons
(Luan et al., 2006; Dai et al., 2007; figure 6.3).
Besides ecdysone, JH (juvenile hormone) has an essential
regulatory function in insect metamorphosis. JH is
synthesized by the endocrine glands, corpora allata (in
Drosophila by an analogous ‘ring gland’) and released into haemolymph. JH prevents expression of genes for
metamorphosis in insects, but allows expression of
ecdysteroid early- and late response genes thus making
molting possible (Riddiford, 1966).
Figure 6.3. A model for the regulation of bursicon
secretion from the BAG by NCCAP-R. In the absence of a
positive signal from NCCAP-R, the BAG are electrically
silent and do not secrete bursicon (top). However,
stimulation of PKA in NCCAP-R (presumably in the period
around eclosion) causes release of a positive (possibly
synaptic) signal (S) from these neurons, which results in
the activation of the BAG and the secretion of bursicon
(bottom). D, M, and V refer to dorsal, middle, and ventral
dispositions of the neurons within the SEG.
Abbreviations: BAG, bursicon - expressing neurons of
abdominal ganglion; NCCAP-R, a subset of NCCAP bursicon-expressing
neurons of the suboesophageal ganglion; SEG,
suboesophageal ganglion (From Luan et al., 2006).
After the discovery
that denervation of corpora allata in cockroaches induces
secretion of JH by these glands, biologists understood
that some brain substance must inhibit JH synthesis in
insects. Now we know that the regulation of JH synthesis
by corpora allata is a function of two antagonistically
acting groups of neuropeptides: allatostatins, with
inhibitory effect on the JH synthesis, and allatotropins,
which stimulate hormone secretion by corpora allata. All
the allatotropins and allatostatins isolated and
identified so far are neuropeptides synthesized by
neurosecretory cells in the insect brain.
Neural Control of
Metamorphosis in Vertebrates
Neural Control of
Metamorphosis in Amphibians
Specific neural
circuits in the tadpole’s CNS process the input of
internal stimuli, such as growth beyond a threshold and
environmental cues, such as photoperiod, rise of the water
temperature, lowering of the water level, etc. (Denver,
1997). By processing the input of those stimuli, neurons
in the non-hypothalamic brain
release
catecholamines at their synaptic contacts with the cell
bodies and dendrites of TRH (thyrotropin-releasing hormone
- N.C.)-containing neurons. (Strand, 1998e)
in the hypothalamus.
In response to these neural signals, hypothalamic neurons
secrete TRH (thyrotropin-releasing hormone), which
stimulates the tadpole pituitary to secrete the TSH
(thyroid-stimulating hormone), which in turn regulates
physiological levels of the thyroid hormones,
triiodthyronine (T3) and thyroxine (T4), secreted by the
thyroid gland.
Thyroid hormone is
indispensable for amphibian metamorphosis and is
responsible for initiating all the physiological
manifestations of metamorphosis (morphogenesis, cell
death, body restructuring, etc.) (Tata, 1998).
TH receptors mediate both early and late developmental
programs of metamorphosis as diverse as growth in the
brain, limb buds, nose and Meckel’s cartilage, remodeling
of the intestine, and death and resorption of the gills
and tail. (Tata, 1998)
The functional
versatility of the thyroid hormone during amphibian
morphogenesis derives from the crucial role its receptor (TR)
plays in gene expression by inducing chromatin remodeling
via histone acetylation/deacetylation (Li et al., 1999;
Sachs and Shi, 2000).
Mediators of the
action of TH in metamorphosis are matrix
metalloproteinases (MMPs) (Damjanovski et al., 2000) that
induce remodeling of the extracellular matrix via
integrins, as well as their membrane receptors, which in
turn send signals for expression of specific genes. These
signals make possible the apoptotic remodeling of the
intestine during Xenopus laevis metamorphosis (Shi et al.,
1998). The thyroid hormone is also involved in tail
regression of amphibian tadpoles by activating expression
of the Bax gene, which stimulates apoptosis of tail
myocytes (Sachs et al., 1997). The signal cascade that
starts in the tadpole brain looks as follows:
catecholamines in the
nonhypothalamic brain --> hypothalamic TRH (thyrotropin-releasing
hormone)--> pituitary TSH (thyroid-stimulating hormone)
--> thyroid TH (thyroid hormone) --> (MMPs) matrix
metalloproteinases -->integrins --> signals for gene
expression --> apoptotic remodeling of the Xenopus laevis
intestine.
The role of the TH in
amphibian metamorphosis derives from the hormone’s ability
to regulate expression of a great variety of genes in
various tissues via its receptors (TR-alpha and TR-beta),
which at the same time act as transcription factors. T3 is
the most active form of thyroid hormones, which by binding
its receptor, forms the active complex T3/TR that is able
to recruit one of the chromatin-remodeling enzymes. This
enables the receptor to specifically bind its DNA response
element in the target gene in the form of monomers,
homodimers, or even heterodimers with the retinoic acid.
This is known as the primary gene expression response.
Then, during the secondary gene expression response, the
products of gene expression act on a downstream set of
genes. The differential expression of genes induced by
this signal cascade during metamorphosis enables the
tadpole to specify differential developmental responses,
from cell proliferation and cell differentiation, to the
programmed cell death.
Transcription
factors, cellular enzymes, a cytoskeletal element, and
secreted signaling molecules of four general classes of
genes are activated by the T3 in Xenopus tadpoles (Denver,
1997a). R. Denver estimates that TH alone upregulates ~35
genes and downregulates ~10 genes involved in the tail
degeneration program of Xenopus laevis.
Post-phylotypic
Development in Vertebrates
The Embryonic
Central Nervous System Controls the Postphylotypic
Development in Vertebrates – Empirical Evidence
Termination of the
function of parental cytoplasmic factors at the phylotypic
stage, coincides with the beginning of organogenesis, when
the embryo enters the phase of accelerated, more complex
development requiring investment of ever-increasing
amounts of epigenetic information. Where does the enormous
information for erecting the extraordinarily complex
metazoan structure come from?
As noted earlier, the
most important event coincident with the termination of
the regulatory function of the parental cytoplasmic
factors, which might be relevant to the above question, is
probably the development of the operational CNS. Its
development occurs simultaneously with both exhaustion of
maternal cytoplasmic factors and the beginning of the more
complex stage of organogenesis. The candidacy of the CNS
for the function of the control of the post-phylotypic
development is not easy to argue. Out of many questions
arising, the most dismaying one has to answer is the
following: How could the CNS know what and how to do
during sequential stages of individual development?
My approach to the
issue will be empirical, to trace the source of
information by identifying the origin of inductive signals
rather than trying to theoretically argue the possibility
that the CNS may be in possession of, or can generate, the
epigenetic information necessary for individual
development.
In chapter 1 I
presented evidence on the role of the CNS in maintaining
and restoring homeostatic/physiological parameters and
described the CNS control of a few behavioral and
morphological traits; I have also argued and substantiated
the idea that an integrated control system (ICS), with the
CNS as its controller, is operational in metazoans. In
this chapter I will expose a representative portion of
available evidence on the CNS origin of signals for the
development of tissues and organs in the course of the
post-phylotypic development.
Despite the fact that
the possible role of the CNS in embryonic development has
not been an authentic object of biological research and
almost no experiments designed to investigate that
possibility have ever been carried out, surprisingly, the
evidence for substantiating that role seems to be
adequate.
Directed Cell
Migration
Directed cell
migration is a common and essential mechanism of
morphogenetic processes taking place during the post-phylotypic
development. It is determined by a mechanism that enables
migrating cells to detect specific cues released by other
cells, or extracellular matrix proteins, along the
migrating path towards the target sites (O’Toole, 2001).
This mechanism is based on the presence on the membrane of
migrating cells of receptor molecules that bind their
specific ligands.
Proximate causes of
production of both the membrane receptors and their
extracellular ligands are growth factors (De Felici and
Pesce, 1994; Salcedo et al., 1999) and hormones
(Caballero-Campo et al., 2002; Dominguez et al., 2003;
Kitaya et al., 2004), which, in turn, represent elements
of signal cascades originating in the CNS.
SDF-1 (stromal
cell-derived factor-1) is one of the cues that binds cell
surface chemokine receptors of the CXCR family. It
stimulates migration of endothelial cells, as well as B
and T lymphocytes, by inducing reorganization of their
actin cytoskeleton (Hall and Korach, 2003). SDF-1 also
displays chemotactic properties for migration of
megakaryocytes through bone marrow endothelial cells
(Hamada et al., 1998). SDF-1 alpha also induces T cell
migration, whereas a hypothalamic neurohormone,
somatostatin, inhibits that migration. Cyclosomatostatin,
an antagonist of the neurohormone, suppresses the action
of somatostatin (Talme et al., 2004). It is demonstrated
that, in vitro, migration of T-lymphocytes on a surface
coated with extracellular matrix (ECM) is stimulated by
the expression in those cells of MMP-9 (matrix
metalloproteinase-9), by TIMP-1 (tissue inhibitor of
matrix metalloproteinase-1), as well as by exposure to
various chemokines. Prevention of expression of MMP-9
makes T-lymphocytes unable to migrate (Ivanoff, 2003).
Stimulation of the
bovine capillary endothelial cells with VEGF (vascular
endothelial growth factor) or bFGF (basic fibroblast
growth factor) upregulates synthesis of the CXCR4 mRNA,
thus inducing migration of the endothelial cells towards
the source of SDF-1 alpha (Salcedo et al., 1999). FGF-2
induces formation of angioblasts from the splanchnic
mesoderm (Poole et al., 2001) and VEGF-A stimulates
migration of angioblasts and formation of small blood
vessels (Ash and Overbeek, 2000). In Xenopus, VEGF acts as
a chemoattractant expressing the gene for its receptor,
Flk-1 that is necessary for migration of angioblasts, from
mesoderm to the midline, where angioblasts pattern the
dorsal aorta (Cleaver and Krieg, 1998).
Migration of
neutrophil leukocytes towards inflammation sites is
induced by neuropeptides VIP (vasoactive intestinal
peptide), CGRP (calcitonin gene-related peptide),
secretoneurin, and substance P (Dunzendorfer et al.,
1998), whereas migration of eosinophil leukocytes is only
stimulated by the neuropeptide substance P (Carolan and
Casale, 1993). Migration of cancer cells towards
metastasis sites, as well, is induced by similar
mechanisms with the involvement of the same elements, such
as SDF-1 and members of the CXC receptor family (Koshiba
et al., 2000; Hwang et al., 2003; Parmo-Cabanas et al.,
2004).
The evidence
presented so far shows that SDF-1, chemokine receptors,
MMPs, and several growth factors and neuropeptides, are
the main players involved in the processes of directed
cell migration during embryonic and postembryonic
development. The guidance role of the SDF-1 is based on
its ability to bind cell surface G protein-coupled
receptors of the CXCR family. But, is there any relation
between the SDF-1 and any of the
hypothalamic-pituitary-target endocrine axes? Earlier we
have shown that 17beta-estradiol (E2), via its nuclear
receptor, induces expression of the gene coding for SDF-1
(Coser et al., 2003; Hall and Korach, 2003) and the growth
factor TGF-beta1 downregulates transcription of that gene.
Secreted proteins
that induce expression of chemokine receptors are the VEGF
(vascular endothelial growth factor) and basic fibroblast
growth factor (bFGF). Growth factors involved in the
process of directed cell migration as well are under
hormonal control: expression of the TGF-beta group is
stimulated by RA (retinoic acid) (Cupp et al., 1999),
which, in turn, is regulated by the pituitary FSH
(follicle-stimulating hormone) (Guo et al. 2001) and
gonadotropins (Demura et al., 1993; Vermot et al., 2000),
whereas estrogens induce expression of the RA receptor.
Expression of VEGF, as well, is induced by estrogens and
androgens (Rouhola et al., 1999), but is inhibited by
overexpression of estrogen receptor alpha (Ali et al.,
1999).
Similarly, it has
been demonstrated that progesterone induces transcription
of genes for chemokine receptors CXCR1, CXCR4, CXCR5, and
CCR2B in the human blastocyst (Dominguez et al., 2002) as
well as ligands of the CXCR3, Mig and IP-10 in human
endometrium (Kitaya et al., 2004). The chemokine SDF-1,
can also induce expression of its receptor, CXCR4 (Wu et
al., 2004).
In vitro
administration of progesterone and estradiol in human
endometrial stromal cells inhibits secretion of MMPs
(matrix proteinases) (Lockwood et al., 1998), and
progesterone withdrawal induces secretion of MMPs (Irwin
et al., 1996), but not the secretion of TIMPs (tissue
inhibitors of metalloproteinases) (Salomonsen et al.,
1997). Several other hormones, such as retinoic acid,
glucocorticoids, and androgens are known to inhibit the
synthesis of metalloproteinases. Mediators of the function
of these hormones are their nuclear receptors, which may:
1. Form complexes on
the DNA,
2. Induce secretion
of TIMP mRNA, which binds MMPs, or
3. Bind to elements
of MMP transcription (Schroen and Brinckerhoff, 1996).
The hormonal
regulation of the SDF-1, in the context of the central
regulation of the secretion of hormones by target
endocrine glands clearly hints the possibility of
neuroendocrine regulation of its secretion.
Somitogenesis
Formation of somites,
metameric structures of the vertebrate embryo, represents
a fundamental aspect of the vertebrate body segmentation.
Somites arise from the presomitic mesoderm (PSM) and
develop sequentially in rostral-caudal order. The
development of somites is one of the earliest events in
the post-phylotypic development of vertebrates that takes
place under the control of the neural tube, the precursor
of the CNS.
Development of
Somites and Myogenesis
In zebrafish, soon
after formation of the neural tube, a localized expression
of the genes for the hypothalamic GHRH (growth
hormone-releasing hormone) and PACAP (pituitary adenylate
cyclase activating polypeptide) occurs in the neural tube
and the cerebellum. It has been proposed that the temporal
and spatial expression of those genes suggests that:
these hormones may modulate patterning during development.
(Krueckl, et al., 2002)
In
Xenopus laevis,
segmentation begins in the PSM with rhythmic expression of
the gene for Mesp-like bHlH protein, also known as
thylacin 1, in a pattern that determines both segment
boundaries and polarity (Moreno and Kintner, 2004).
Expression of myogenic bHlH (basic helix-loop-helix) genes
is induced by signaling proteins of the Wnt family,
originating in the dorsal regions of the neural tube, and
Sonic hedgehog (Shh) secreted from both the neural tube
floor plate and the notochord (Fan and Tessier-Lavigne,
1994) (figure 6.4).
A combination of the
above factors is capable of inducing myogenesis in somites
in vitro (Munsterberg et al., 1995). Besides thylacine 1,
components of the Notch signaling pathway (expression of
Notch is closely related to the morphogenetic processes in
the CNS), are expressed in a rhythmic pattern (Kim et al.,
2000), thus determining the identity of the incipient PSM
segments. The inducer of expression of the segmentation
bHlH gene, thylacine 1, is the hormone RA (Moreno and
Kintner, 2004). This fact adds another link to the causal
chain (originating in the CNS) of the segmentation
signaling in the PSM.
Figure 6.4. A
simplified scheme of signaling molecules in newly formed
epithelial somite. Shh (gray dots), produced by notochord
(Nc) and floor plate, acts on the ventral domain of newly
formed epithelial somites, inducing sclerotome, and also
on the dorso-medial domain, inducing medial dermomyotome.
Wnt1 (black dots), produced by dorsal neural tube (NT),
acts (with Shh) on the dorso-medial domain of newly formed
somites (Sm), where Myf5 expression is observed soon after
epaxial progenitors are specified. Wnt7a (semilunar dots),
produced by dorsal ectoderm (DE), acts on the dorso-lateral
domain, where hypaxial progenitors are specified. BMP4
(black polygons), produced by lateral mesoderm (LM),
prevents MyoD activation and early differentiation in the
lateral domain of somites. Its action is counteracted by
direct binding of Noggin (gray triangles) produced by
dorsal neural tube (From Cossu and Borello, 1999).
Expression of beta-catenin
mRNA in somites is regulated by positive and negative
signals (BMP4, Shh, and Wnt/Wnt3) from the neural tube,
notochord, and lateral plate mesoderm (Schmidt et al.,
2000; figure 6.5). Within somites themselves, myogenesis,
formation of muscles, is also induced by neural tube
signals, probably directly by Wnt-1 (Stern et al., 1995)
or via growth factors, bFGF (basic fibroblast growth
factor), TGF-beta1 (transforming growth factor beta1), and
ds/1 (dorsalin) (Stern et al., 1997). Neural tube signals,
such as Wnt-1, may also induce expression of Noggin, a BMP
antagonist, which may induce somite myogenesis by allowing
expression of MyoD and Myf5 in somite cells (Reshef et
al., 1998).
Figure 6.5.
Proposed model of events that lead to stable induction of
myogenesis in the somite. (1) Wnt and Shh act together to
induce high levels of Wnt pathway components. This results
in the activation of skeletal muscle genes. (2) The Wnt
pathway is stably induced and Shh is no longer required.
Ultimately skeletal muscle differentiation becomes
autonomous (From Schmidt et al., 2000).
The sequential
anterior-posterior order of the development of somites
points to the existence in the embryo of a segmentation
clock that establishes that order. Indeed, adequate
evidence shows that the segmentation clock consists of the
Wnt/beta-catenin signaling “via a negative feedback
mechanism” (Aulehla et al., 2003).
Expression of all the
investigated MyoDs (Xu et al., 2000), including bHlH
proteins, esr9 and esr10, in Xenopus somites (Li et al.,
2003), occurs rhythmically and rhythmically is activated
the Notch signaling pathway (Li et al., 2003). Together,
these facts suggest that a segmentation clock of the
neural tube/notochord axis regulates rhythmic expression
of PSM (presomitic mesoderm) segments.
Myogenesis
During metamorphosis,
insects develop pools of myoblasts in muscle Anlagen,
which are closely associated with local nerves. Local
innervation is essential for the development of abdominal
muscles of insects in general. In Drosophila, for
instance, the patterning of segmental muscles is
determined not by myoblasts involved in muscle
development, but by the motor and sensory neurons
innervating these muscles (Lawrence and Johnston, 1986).
It is observed that abdominal myoblasts arrive at the
target sites by their association with, and migration
along, the nerves. (Currie and Bate, 1991)
Local denervation
causes a decrease in the number of muscle nuclei,
suggesting that innervation is necessary for regulating
myoblast division and survival. However, denervated
muscles undergo differentiation and establish basic muscle
patterns, a fact that may be explained by a central
regulation via the neurohormonal pathways.
There is at least one
case when the role of the innervation in the development
of muscle is of the all-or-none type. In Drosophila, male
flies have an additional, male specific muscle (msm), or
the Muscle of Lawrence, in the dorsal part of the
abdominal segment 5 (A5), which results from aggregation
of 3 to 5 adjacent muscles. In distinction from abdominal
muscles, the Muscle of Lawrence does not form at all in
the absence of innervation (denervation at the onset of
metamorphosis prevents its development). It is noteworthy
that innervation of fibers of this muscle is more
extensive than any other muscle in Drosophila (Currie and
Bate, 1995).
In denervated legs of
Manduca sexta, myoblasts cannot proliferate and move to
the proper locations for developing leg muscles, what, in
a considerable proportion of cases, leads to the
development of legs lacking muscles (Consoulas and Levine,
1997). In the same insect, denervation of respecified
muscles during metamorphosis leads to dedifferentiation of
their cells, loss of muscle fibers and, at times, to
muscle death (Bayline et al., 1998). Denervation of the
Anlage of the longitudinal flight muscle in this insect
prevents the development of that muscle, due to the
failure of myoblasts to accumulate in the muscle Anlage (Bayline
et al., 2001).
Noteworthy is another
interesting fact on the CNS control of muscle development.
Null mutations of the single-minded gene in Drosophila
are
associated with abnormal development of the ventral
oblique muscles above the central nervous system.
Investigators have concluded that this defect
is not due to the absence of single-minded expression in
muscle precursor cells and likely results from an
influence of the central nervous system on ventral muscle
development. (Lewis and Crews, 1994)
In the tobacco
hawkmoth, Manduca sexta, a LIM-only protein (DALP) induces
elimination of excess myoblasts of the intersegmental
muscles (ISMs) at the end of metamorphosis. Again, the
causal chain, the signal cascade responsible for the death
of ISMs starts in the insect’s CNS, as it is suggested
from the experimental demonstration that the death of ISMs
is triggered by a drop in the level of the molting
hormone, 20-hydroxyecdysone (Hu et al., 1999), which, as
it is well known, is under strict cerebral control.
Maturation of
oviposition properties of the longitudinal muscle in
female locusts implies acquisition of the ability by the
muscle to tolerate large extensions of more than 8 mm.
Experimental inactivation of endocrine glands corpora
allata (CA) inhibits maturation of oviposition properties
of the longitudinal muscle. This suggests that JH
(juvenile hormone), whose synthesis in CA is triggered by
brain signals, is necessary for maturation of oviposition
properties of that muscle. Indeed, administration of JH in
female locusts with inactivated CA reverses the situation
by enabling maturation of oviposition properties in the
longitudinal muscle (Rose et al., 2001).
The indirect flight
muscles (IFMs) represent two muscle groups, the dorsal
longitudinal muscles (DLMs) and dorso-ventral muscles (DVMs).
The development of indirect flight muscles (IFMs) takes
place in two stages: the first stage of generation of the
pool of myoblasts, which is nerve-independent, and the
second nerve-dependent stage when motoneurons determine
the size of the myoblast pool by establishing a critical
threshold of the myoblast pool. However, differences are
observed in the effect of innervation between DVMs and
DLMs (figure 6.6).
Denervation of the
DVMs leads to a myoblast pool that falls below the
threshold so that neither fusion takes place nor mature
muscle fibers to develop. The fact that DLMs develop
despite denervation and the reduced myoblast pool may be
related to the fact that myoblasts in the case of DLMs
form on the persisting larval muscles that
serve as scaffold for patterning the mature muscle as
opposed to the larval DVMs that degenerate during
metamorphosis and have to be patterned and form de novo (Fernandes
and Keshishian, 2005).
Unilateral
denervation of these muscles leads to two major effects.
First, a significant decline in the proliferation rate of
myoblasts, which may be the cause of the observed reduced
myoblastpopulation and smaller muscle size in denervated
hemisegments. Second, it prevents myoblast patterning,
leading to failure to form muscle Anlagen and muscles.
The motoneuron influences both the number of cells
available for fusion, as well as potentially regulates the
fusion events themselves. This in our view is an elegant
mechanism for controlling muscle fiber differentiation
during myogenesis, and may have evolved as a way to ensure
that muscle primordia develop into muscles that meet the
diverse demands placed on them by the nervous system. (Fernandes
and Keshishian, 2005)
Local innervation is
directly and crucially involved in the proliferation and
distribution of myoblasts and myogenesis in Drosophila
(Currie and Bate, 1991).
Figure 6.6 . Neuromuscular development of the indirect flight muscles (IFMs)
during the first 24 hours of pupal development. The IFMs
consist of the dorsal longitudinal muscles (DLMs) and the
dorsoventral muscles (DVMs, I, II, III). (A) 8 hours APF
(after pupa formation). Three persistent larval muscles (9,
10, 19) give rise to the DLMs. The larval nerves (intersegmental
nerve, ISN, innervating dorsal targets and the segmental
nerve, SN innervating ventral and lateral targets) have
retracted their larval neuromuscular junctions. 1, 2 and 3
indicate regions of nerve cuts that resulted in complete,
partial and transient denervation respectively. (B) 12 hours
APF. The larval muscles flatten and elongate, and adult
specific nerve outgrowth is seen. In the region of the DVMs
smaller outgrowths are noticeable. (C) 16 hours APF. The
larval muscles split as myoblasts fuse with them to begin
formation of the six DLM fibers. Simultaneously, the nerve
also undergoes reorganization. Higher-order nerve branches
arise at this time. This is the earliest time that DVM
fibers can be seen. (D) 24 hours APF. The adult
neuromuscular pattern is formed. The DLMs (a-f) and DVM III
are innervated by the posterior dorsal mesothoracic nerve (PDMN),
which arises from the restructuring of the ISN. DVM I and II
are innervated by the ADMN and the mesothoracic accessory
nerve respectively (From Fernandes and Keshshian, 1998).
In
Drosophila, denervation causes reduction of myoblast proliferation and
alteration of segregation of myoblasts in dorso-ventral
muscles (figure
6.7) (Fernandes and Keshishian, 2005).
Neurectomy of the larval leg nerve before metamorphosis in
the moth Manduca sexta prevents proliferation and normal
migration and accumulation of myoblasts in respective
regions, so that in 26% of cases no muscles develop in adult
legs (Consoulas and Levine, 1997).
These experimental
facts demonstrate that motor neurons are necessary for
muscle formation during insect metamorphosis.
During metamorphosis
motoneurons not only withdraw larval synapses and rispecify
adult nerve branches and synaptic morphology, but they also
rispecify adult dendritic arbors, neuronal connections, and
circuitries in the brain (Kent and Levine, 1988).
Denervation of indirect flight muscles (dorso-lateral
muscle, which forms on larval muscle scaffold and
dorsoventral muscle, which forms de novo) has shown that
innervation is necessary for maintaining the size of
myoblast population (Fernandes and Keshishian, 1998). The
breakdown of the abdominal intersegmental muscles in the
saturniid silkmoths takes place in the presence of hormone
ecdysone and in the absence of JH (juvenile hormone) but
investigators have concluded that
The actual breakdown is triggered by a neural mechanism. The
latter consists of a sudden curtailing or cessation of the
outflow of impulses in the motor nerves which innervate the
abdominal muscles… By chronic electrical stimulation of the
nerves, the breakdown of the muscles can be opposed or
prevented. (Lockshin and Williams, 1965)
An essential
relationship between the nervous system and myogenesis has
also been observed in vertebrates. The differentiation of
myoblasts, which starts the development of skeletal muscles
within somites, depends on the expression of MyoD genes (Alves
et al., 2003) coding for transcription factors that induce
expression of genes for muscle specific proteins (MSPs) (Alves
et al., 2003; Te and Reggiani, 2002) and Myf5. These
muscle-specific proteins are also expressed according to an
identical timetable in each somite (Xu et al., 2000). Again,
signals for expression of MyoD genes (among them, Hedgehog
and Wnt) originate in the neural tube/notochord adjacent to
somites (Te and Reggiani, 2004) as it is also concluded from
the fact that, when separated from that axial structure,
somites do not express MyoD genes (Alves et al., 2003).
Experimental misexpression of Myf5 and MyoD in the chick
neural tube results in ectopic skeletal muscle development (Delfini
and Duprez, 2004). In Xenopus laevis maternal MyoD [acting
downstream the Pax3 (Arnold and Winter, 1998)] is present in
the egg, although the gene for MyoD is the earliest
muscle-specific gene to be expressed during gastrulation
(Hopwood et al., 1989).
Figure 6.7. (A) 16 hour APF (after pupa
formation) primordia for DVM I, II and the TDT jump muscle
in innervated control side. (B) 16 h APF (denervated
side): myoblasts are not organized into discrete DVM and TDT
primordia, and formed a single undifferentiated mass (From
Fernandes and Keshishian, 2005).
Studies on the development of
muscles in the paraxial mesoderm (on both sides of the neural
tube) and somites in chicks have shown that signals from the
adjacent dorsal neural tube, and to a limited extent from the
ventral neural tube, are basic inducers of myogenesis in these
embryonic structures (figure
6.8). Ablation of the neural tube
prevents formation of muscles in these structures (Stern et
al., 1995). Innervation may also have suppressive effect on
gene expression in muscles. This is the case, e.g., in rats
where denervation of particular muscles, within one week,
increases 150-200 fold expression of muscle transcription
factors, such as MyoD, Myf-5, and myogenin (Voytik et al.,
2005).
Figure 6.8 .
Proposed models of axial structure-dependent paraxial
mesoderm myogenesis. For somites (A) the major signal
(thick, solid arrow) comes from the dorsal neural tube and
may be mediated in part by Wnt-1 or other Wnts in this
region of the neural tube such as Wnt-3a. The expression
pattern of Wnt-1 and Wnt-3a as determined by Hollyday et al.
(1995) is indicated by white hatching. Weaker myogenic
signals (thin, solid arrows) emanate from the ventral neural
tube and notochord and act directly on somites. These
ventral signals may interact in some way (bracket) with the
dorsal signal causing a synergistic increase in the number
of MHC(myosin heavy chain)-positive cells induced.
Alternatively, the synergy could be due to intra-axial
signaling. For example, the ventral neural tube/notochord
might signal the dorsal neural tube to boost the dorsal
signal (white broken arrows). For segmental plate tissue
(B), the dorsal signal remains the major signal (thick
arrow) and could be mediated in part by Wnt-1 or other Wnts.
The ventral neural tube may emit a weak positive myogenic
signal (thin arrow), but our study shows no evidence for
synergy between dorsal and ventral signals for segmental
plate myogenesis.
Abbreviation: NC, notochord (From Stern
et al., 1995).
Innervation is necessary for the fusion
of muscle fibers and their transdifferentiation into
electrocytes (denervation prevents these processes) in the
weakly electric fish, Sternopygus macrurus, and interruption
of the neural input leads to the dedifferentiation of
electrocytes back into muscle fibers (Unguez and Zakon,
2002).
The central nervous system controls and
regulates the development of target muscles not only
directly, via peripheral nerves, but also indirectly, by
long-range action, via the neurohormonal signal cascades.
Often, both modes of control are operational in the
development of the same muscles, giving rise to a binary
neural control of myogenesis. This is the case, e.g., for
the development of the laryngeal muscle in Xenopus laevis.
In juvenile animals, the laryngeal muscle is similar,
female-like, in both sexes. After metamorphosis, as a result
of androgen secretion, male individuals develop the
male-specific muscle. Denervation of the muscle, however,
causes its atrophy in male amphibians, whereas androgen
administration causes its hypertrophy (Tobias et al., 1993).
It is worthwhile to note that the
laryngeal nerve, under normal conditions, may be involved in
the effects produced by the androgen since both laryngeal
muscle and motoneurons, after metamorphosis, express
androgen receptor. Experimental evidence has shown that
The nerve is required for maintenance of existing fibers and
denervation results in cell death. (Tobias et al., 1993)
In chicks, upon entering the chick
limbs, motor neurons release the RA-synthesizing RALDH-2
enzyme, inducing there muscle cell differentiation and
muscle formation (Berggren et al., 2001). The development of
muscle fiber types in chick embryos coincides with the
penetration of nerves in these muscles. Rightly, this has
been interpreted as indicative of the role of the local
nerves in the differentiation of muscle fiber types.
However, the fact that the embryonic denervation performed
before the nerves enter the muscles impairs the muscle
growth indicates that the local innervation is necessary for
the proper growth and survival of muscles, but not for the
initial muscle differentiation (Butler et al., 1982).
Skeletal muscle fibers are of two
subtypes, slow and fast muscles, depending on
ultrastructural morphology, contractile physiology, and
susceptibility to fatigue. These differences are related to
expression of different types of proteins and enzymes in
each myocyte subtype. The specific programs of gene
expression for each subtype are determined by the activity
of motor nerves innervating each subtype of skeletal muscle
fibers. Firing patterns of motoneurons innervating slow
muscles sustain Ca2+ at levels that activate the
calcineurin-NFAT (nuclear factor of activated T cells)
pathway leading to dephosphorylation and nuclear
localization of NFAT proteins (figure
6.9).
Figure 6.9. Model for a calcineurin-dependent
pathway linking specific patterns of motor nerve activity to
distinct programs of gene expression that establish
phenotypic differences between slow and fast myofibers. MEF2
is shown to represent the requirement for collaboration
between activated NFAT proteins (transcription factors) and
muscle-restricted transcription factors in
slow-fiber-specific gene transcription, but other proteins
(not shown) also are likely to participate.
Abbreviations: NFAT, four calcium/calcineurin
regulated proteins; MEF2, myocyte enhancer factor (From Chin
et al., 1998).
Firing
of motor nerves innervating fast fibers is infrequent and
insufficient to maintain calcineurin in active state, NFAT
proteins remain phosphorylated and are not localized in the
nucleus to bind DNA and the fast fiber proteins are
expressed.
Mechanical stress also may change the patterns of gene
expression and it has been observed that it can activate
synthesis of IGF-I, known for its growth promoting activity
in muscle fibers (figure
6.10). However, it is not known
whether the observed changes in the size and properties of
muscle fibers under conditions of mechanical stress are
direct result of that stress on muscles or are related to
changes in other tissues or organs and signals of neural
origin. Ideas have been expressed that both pathways are
operational and overlap in metazoans (Tidball, 2005).
Figure 6.10.
Potential mechanisms through which mechanical signals and
insulin-like growth factor (IGF)-1-derived signals may be
mediated through overlapping pathways. Solid arrows indicate
cause and effects or interactions that are likely to be
direct. Dotted arrows indicate interactions for which there
may be unknown intermediates.
Abbreviations: IRS, insulin response substrate; NFAT, nuclear
factor of activated T cells; mTOR, mammalian target of
rapamycin; PI3K, phosphatidyl-inositol-3 kinase; ?, unknown
downstream events that result from Akt association with the
actin cytoskeleton (From Tidball, 2005).
In Drosophila, the
indirect flight muscles, DLMs (dorsal longitudinal
muscles) and DVMs (dorso-ventral muscles), develop in
different ways: DLMs develop from myoblasts of existing
larval muscle fibers, while DVMs develop de novo.
Motoneurons regulate muscle formation not by providing any
essential survival factor or by stimulating myoblast
migration but by regulating imaginal pioneer cells, which
serve as myoblast fusion targets and is believed to
prefigure the DVM muscle fibers and myoblast
proliferation, which is correlated with the expansion of
motoneuronal terminal arbors on the muscle fiber surface.
Denervation inhibits myoblast proliferation (Fernandes
and Keshishian, 2005; figure 6.11).
Figure 6.11.
The roles of innervation during indirect flight muscle
myogenesis. The events associated with the DLM and DVM
formation are compared. For both muscle types myoblast
proliferation occurs in a nerve-independent fashion prior
to the onset of fusion events (0–12 h APF) to generate the
initial pool of myoblasts. DLM myogenesis: the persistent
larval fibers are responsible for segregating the
myoblasts into fibers which eventually split lengthwise.
Nerve-dependent proliferation of myoblasts replenishes the
myoblast pool, so that the appropriate fiber size is
attained. When denervated, segregation of myoblasts still
occurs normally. However, there is a delay in fusion and a
reduction in subsequent proliferation of myoblasts,
accounting for reduced muscle sizes. DVM myogenesis: the
innervating motoneuron, along with the Duf-positive
founder cells is responsible for segregation of myoblasts
into the distinct DVM primordia. Nerve-dependent
proliferation replenishes the myoblast pool. When
denervated, the myoblasts are incapable of segregating
into discrete primordia, and proliferation is also
reduced. In addition, Duf-positive cells are no longer
reliably evident by 24 h APF.
Abbreviations:
APF, after puparium formation; DLMs, dorso-lateral
muscles; DVMs, dorso-ventral muscles; Duf, a type of
“imaginal pioneer” cells (From
Fernandes and Keshishian, 2005).
Based on experimental
evidence, it is concluded that the
motoneuron influences both the number of cells available
for fusion, as well as potentially regulates the fusion
events themselves. This in our view is an elegant
mechanism for controlling muscle fiber differentiation
during myogenesis, and may have evolved as a way to ensure
that muscle primordia develop into muscles that meet the
diverse demands placed on them by the nervous system. (Fernandes
and Keshishian, 2005)
Development of secondary
myotubes
(elongated multinucleate structures, arising from the
fusion of several
myoblasts)
during rat embryogenesis always occurs in contact
with,
never at a distance from,
the primary
myotube
innervation
zone. Additionally, at least some, if not all, secondary
myotubes
make direct contacts with a nerve terminal (Duxson
et al. 1989). Based on the fact that formation of primary
and secondary
myotubes
in vivo occurs exclusively at the zones of
innervation,
it has been concluded that nerve terminals regulate the
fusion of
myoblasts
and formation of
myotubes
(Duxson
and
Sheard,
1995). In later stages of development,
innervation
suppresses expression of
myogenin,
a muscle specific transcription factor, in muscles of the
fetal hind limb. But if
denervation
of the muscle is performed,
myogenin
expression resumes
(Buonanno
et al., 1993). Development of muscle stretch receptors
also
requires, as a necessary condition, the presence of
sensory
innervation
(Soukup
and
Zelena,
1985).
Muscle
development often
requires both hormonal and neural regulation, the
binary neural control system.
In vitro experiments show that both ecdysone and
neurons separately stimulate proliferation of
myoblasts (Luedeman
and Levine, 1996).
An extremely curious and compelling
case of the CNS control and regulation of the muscle
development is the development of the dorsal external
oblique 1 (DEO1) muscle during metamorphosis in the
tobacco
hawkmoth,
Manduca sexta
(Hegstrom
et al., 1998). During metamorphosis, larval muscles are
radically remodeled; out of the
five muscle fibers which
the larval DEO1 muscle consists of,
only one serves as
Anlage
for the development of the adult muscle while the rest of
them degenerate and are eliminated. The remaining fiber
starts myoblast proliferation to grow into the adult
muscle.
Investigators
now have an answer to the question: What
let
that specific muscle fiber to survive, although it was
under the same hormonal regulation as the eliminated
fibers? That muscle fiber is the only that, during
metamorphosis, increases expression of a specific
isoform
of
ecdysone
receptor (EcR-B1), which by binding to ecdysone
stimulates
myoblast
proliferation. But this statement is just a description of
an experimental fact, not an explanation, hence it leads
us to the next
question: What selects this particular muscle fiber to
express that specific
ecdysone
receptor, while
four remaining, equally vibrant fibers, do not
express it?
The qualitative similarity of these fibers suggests that
selection is made externally ,
from outside the fibers. Indeed, experimental evidence has
shown that the selection is made by the local
innervation.
The terminal arbor of the
motoneuron
innervating the DEO1 muscle during metamorphosis recedes
from four of the muscle fibers and remains in contact only
with the surviving fiber (Truman and Reiss, 1995), on
which the axon terminal arbor releases a diffusible agent
of still unknown nature (Hegstrom
et al., 1998). The conclusion that
ecdysone
receptor (EcR-B1) expression is determined by the
motorneuron
is also
corroborated
by
the fact that experimental denervation prevents both
EcR-B1 expression and myoblast proliferation. In the
absence of
innervation,
the four muscle fibers, just before
ecdysis,
express only EcR-A, which is known to be related to
apoptosis, determining thus their fate, i.e. elimination
of those fibers.
That the innervation is
essential for the fiber to respond to ecdysteroids is
shown by the results of the denervation experiments.
Denervation entirely eliminated the upregulation of EcR-B1
if performed before the upregulation occurred. If
performed after the upregulation had commenced,
denervation dramatically reduced the expression by 24 hr
and eliminated it by 48 hr later
(Hegstrom
et al., 1998).
The development of the dorsal external oblique 1 (DEO1)
muscle in
Manduca sexta,
thus, is
function of a binary neural
control mechanism; on the one hand, it is mediated via the
cerebral
PTTH
(prothoracicotropic
hormone)-ecdysone
axis and, on the other, it is carried out via local
innervation.
Left-Right Asymmetry
There is a general
consensus
that in vertebrates the left-right
(LR)
asymmetry is determined by asymmetric expression of
the
nodal
gene. Downstream, this gene induces expression of pitx2,
but it is also known that nodal expression is
induced by the transcription factor, Vg1, a member of the
TGF-beta
family, which are induced by hormones (estrogen and
gonadotropins)
of signal cascades that ultimately originate in the CNS
(see
chapter
1,
section Endocrine Control of Secreted Proteins
and Growth Factors).
In chicken, along with the asymmetric expression of
Sonic hedgehog (Shh)
in
Hensen’s
node, the Lefty-1 gene
is involved in determining the left-right asymmetry
and
its
expression is induced by a hormone,
the RA (retinoic acid) (Tsukui
et al., 1999). It is known that during early development
the neural tube is the main source of RA; it produces and
releases more RA than any other structure (the RA level in
the neural tube is 29 times higher than in the heart) (Maden
et al., 1998).
Recent evidence from studies
on
frogs and chicks shows that the left-right asymmetry is
determined earlier by a maternal factor. The
maternal
factor is the neurotransmitter serotonin, which determines
the left-right asymmetry as early as the 4-cell stage
(Fukumoto
et al., 2005), i.e., before the initiation
of the expression of zygotic genes.
Left-right asymmetry
comprises not only asymmetric position of organs and parts
in the body but also the asymmetric looping of many organs
and parts, which is the most general marker of LR
asymmetry in metazoans (Adam et al., 2003). This last form
of asymmetry is probably the best known as far as the
determining molecular mechanisms are concerned.
Organ looping in insects is
determined by the level of juvenile hormone secreted by
corpus allatum (or the ring gland) in response to brain
signals (allatostatins/allatotropins). By investigating
the looping asymmetry of genitalia in Drosophila,
Adam et al. (2003) found that a mutation in the gene
Fas2 (fasciciclin 2) in Drosophila males
leads to abnormal levels of JH during the pupal stage and
rotation defects of their genitalia and spermiduct (figure
6.12). They concluded that a link exists between organ
looping and the retinoic-like juvenile hormone (Adam et
al., 2003). They also present evidence on the similarities
between the invertebrate and vertebrate asymmetries in
organ looping and the fact that a terpenoid, JH,
determines organ looping in invertebrates may suggest that
the vertebrate terpenoid, RA (retinoic acid), may also be
involved in organ looping asymmetry in vertebrates.
Figure 6.12.
Model for Fas2 and
JH action in Drosophila organ looping. (A)
Fas2 is specifically expressed and required in
neurosecretory cells innervating the corpora allata, to
control JH titers. JH is released into the circulating
hemolymph and reaches the genital disc to control its
rotation non-autonomously. (B) Parallels between
the vertebrate and the Drosophila pathways
controlling asymmetric organ looping. In vertebrates, two
conserved pathways have been implicated in LR asymmetry,
the retinoic acid (RA) and the nodal pathways. RA
plays a dual role as it is also involved in organ looping.
In Drosophila, mutants exist that can lead to a
reversion of genitalia rotation, suggesting that they are
involved in LR asymmetry (collectively represented by a
question mark). The analysis of Fas2spin
shows a role for JH, a RA analog, in the control of organ
looping and suggests an evolutionary conserved function of
terpenoids in the control of asymmetric organ looping
(From Adam et al., 2003).
Asymmetry of the first
pair of chelipeds is characteristic of many crustaceans.
Snapping shrimps
Alpheus
heterochelis
have a pair of asymmetric chelae (claws), the big one -
the snapper and the smaller one - the pincer, with
different morphology and functions (figure 6.13).
The first serves for defensive responses and the second is
used for burrowing and feeding. Both chelae start as
identical structures during the early juvenile stages and
only by the sixth juvenile stage start differentiating
into snapper and pincer. What determines which of the
chelae will become the snapper and which will become the
pincer?
Genes and their products,
humoral factors, such as hormones, growth factors or other
secreted proteins in the body fluids are excluded from
playing any essential role. It was observed that removal
of one of the claws before their differentiation (until
the fourth juvenile stage) induces the remaining claw to
develop into a snapper and, in adult shrimps, removal of
the snapper induces transformation of the pincer into a
snapper, while a new pincer develops on the removed
snapper’s site. In other experiments it is observed that
cutting the relevant nerve in the snapper claw induces
transformation of the pincer into a snapper (Read and
Govind, 1997).
Based on the experimental
evidence, it has been proposed that
Neural influences from the transforming pincer-to-snapper claw restrict
regeneration of the contralateral claw to a pincer type
thereby ensuring bilateral asymmetry in adult shrimps.
(Young et al.,
1994)
It is concluded that the
inhibition by the intact snapper claw is centrally
determined (Read and Govind, 1997) and
The
snapper-based inhibition of the pincer claw is neural in
origin. (Read and Govind, 1997)
The observation that the
intact snapper claw inhibits the development of the
contralateral claw to a snapper has been explained with
the more intense innervation of the snapper (the snapper
has over 13,000 axons whereas the pincer has only 10,000)
(Read et al., 1991).
Neural Control of the
Development
of
the
Neuroendocrine
System
The hypothalamus is a part of the brain but, due to
its regulatory functions in homeostasis, it is not
buffered by the blood-brain barrier. In humans it begins
to develop by the 6 gestational week and in
mice
by the
embryonic day 13.
The hypothalamic
Anlage
is encircled by neurons and fibers containing GABA (gamma-aminobutyric
acid)
and its
synthetizing
enzyme GAD67 (glutamic
aciddecarboxylase67).
.....................
.....................
Figure 6.14.
Model of Rathke’s pouch patterning at E (embryonic stage)
11.0. The boundary between the oral ectoderm expressing
Shh and the nonexpressing region forming Rathke’s
pouch functions as an organizing center for ventral gene
induction, including Bmp2, directing
patterning and proliferation of the gland. FGF8/FGF10
expressed in the infundibulum functions antagonistically
to SHH/BMP2. These ventrodorsal SHH/BMP and dorsoventral
FGF activity gradients within Rathke’s pouch lead to the
induction of several temporally and spatially restricted
transcription factors, several of which are listed on the
right, that are postulated to combinatorially divide
Rathke’s pouch into zones with different identities. These
zones are proposed to impose the determination of cell
lineages at this developmental stage (From
Treier et al., 2001).
Neural Control of the
Development of Sensory
Organs
Sensory organs in vertebrates mainly develop from cranial
placodes, thickenings of the embryonic head ectoderm,
which in turn develop from
an
earlier structure, the PPE (preplacodal ectoderm). PPE
develops next to the neural tube and is characterized by
neurogenic potential and by expression of specific genes
before the formation of the neural plate. It forms within
the cranial neural plate border region and may be induced
by the same maternal factors that determine the formation
of the neural plate (Baker and Bronner-Fraser, 2001), but
other interactions between the neural plate and epidermis
may also contribute to the differentiation of placodes and
to
their
later subdivisions. Placodes are essential for the
development of sensory structures in vertebrates and
invertebrates.
The olfactory placodes
develop
within the neural plate. As parts of the common PPE
domain, nasal precursors segregate from the lens placode
and in the process of their differentiation and migration
give rise to the olfactory neurons
and the olfactory epithelium (Whitlock, 2004).
Let’s illustrate the neural tube/CNS control of the
development of
the
sensory organs with the development of
the otic vesicle and
inner ear in mammals.
Inner Ear
It was believed that
the
otic
placode
is induced by
Fgfs
and
Wnts
of the
chordamesoderm
and hindbrain,
but
recent studies show that the head
mesoendoderm,
including the
chordamesoderm,
are not necessary for the early
placode
induction.
It is
Fgf8 and Fgf3 signals,
released from the hindbrain
rhombomere
4 that determine the early induction and maintenance of
the
otic
placode
and inner ear patterning in
the
zebrafish
(Leger and Brand, 2002).
Hindbrain
signals are necessary and sufficient for establishing the
dorso-ventral
axis of the inner ear.
In this context, it is noteworthy that
a specific change in the
dorsoventral
axis of the inner ear
can be
induced
by reversing the
dorsoventral
axis of the hindbrain.
Shh
(Sonic hedgehog) signals from
the
neural floor plate and notochord are required for ventral
inner ear structures and differentiation of auditory cells
(Bok
et al., 2005;
Riccomagno
et al., 2002;
figure 6.15).
The establishment of axial
polarity and compartmentalization of the otic vesicle as
well is believed to be a function of the hindbrain. The
otic cup develops adjacent to two hindbrain compartments,
r5 and r6 (rhombomeres 5 and 6) whose boundary is aligned
with the middle of the otic cup (figure 6.16). r5
and r6 may send different signals to anterior and
posterior otocyst, respectively. They may send signals
exclusively to the dorsal half of the vesicle, determining
thus formation of its separate AM (anteromedial) and PM (posteromedial)
compartments. Lateral identity is determined later by the
lateral-most part of the invaginating cup and could be
influenced by the distance from the hindbrain (Brigande
et al., 2000).
Figure
6.15. Model depicting how inner ear polarity is generated in response to
extracellular signals. Wnt signals secreted by the dorsal
neural tube (upper dark area) are necessary and sufficient
for the regulation of Dlx5/6 in the dorsal otocyst
(light gray). This contrasts with the function of Shh
emanating from the notochord, which specifies ventral otic
cell fates by regulating the transcription of genes,
including Pax2 and Ngn1 (lighty shaded
areas). The combination of Wnt and Shh signaling
activities are required for the maintenance of Gbx2
along the dorsomedial side of the otocyst. Shh also
restricts Wnt pathway activation, and consequently
Dlx5/6 expression, to dorsal regions of the otic
vesicle. An additional signal secreted from the dorsal
neural tube, possibly a Bmp family member, restricts Shh
target genes to the ventromedial domain of the otocyst. A
balance between these dorsal and ventral signals is key to
promoting regional identity within the otic epithelium
with the ultimate goal of coordinating the morphogenesis
of vestibular and auditory organs (From Riccomagno et al.,
2005).
Neural Control of Heart
Development
After
formation of the neural tube and the CNS, in many
vertebrate species, the heart is the next
organ to develop. The neural tube sends signals (Wnt3 and
Wnt8) that inhibit the induction of
cardiogenesis
and promote blood cell differentiation of mesoderm along
the whole of its length, except for the region where the
heart normally develops. It has been possible to induce
ectopic
heart formation in other parts of the body by simply
activating the GSK3, or other Wnt antagonists (Schneider
and
Mercola,
2001;
Tzahor
and Lassar, 2001, Marvin et al., 2001).
Wnt antagonists from the endoderm permit the
development of heart in a restricted region of mesoderm
where they overwhelm the neural tube Wnt signaling,
leading to expression of Nkx2-5 and synergistic
action of BMP (Zaffran and Frasch, 2002; figure 6.17).
In invertebrates such as Drosophila, by contrast,
Wnt-s
are necessary for the development of the heart.
The neurotransmitter serotonin, via its receptor 5-HT2B,
regulates differentiation and proliferation of the
developing and adult heart
(Nebigil
et al., 2000). Suppression of 5-HT2B synthesis
causes developmental anomalies and midgestational death of
the embryo.
The neurotrophic factor,
BDNF
(brain-derived
neurotrophic
factor), besides its
differentiative
actions on neurons expressing the
Tk
receptor tyrosine
kinase,
has an
angiogenic
effect on endothelial cells and is necessary for
maintaining the stability of the
intraventricular
walls (Donovan et al., 2000).
The hormone
RA (retinoic acid)
signaling determines the proportion of cells that will
differentiate into myocardial progenitor cells from a pool
of
multipotential
cells in a way that the population size of
cardiomyocytes
is inversely related to the RA level
(Keegan et al., 2005).
The
very low level of RA in the heart suggests that
the source of the hormone
may be the RA-rich
neural tube/spinal cord.
Apoptosis, the programmed cell death, also has a special
role in heart morphogenesis. It is observed to occur in
mouse hearts (myocardium and myocardial epithelium) on
days 11 through 16 of their embryonic development (Abdelwahid
et al., 1999).
Figure 6.16. A
schematic showing possible inductive influences of
hindbrain on the development of putative compartments in
the otic vesicle. (A) The r5-r6 boundary in the
hindbrain is aligned with the A-P lineage-restriction
boundary in the otic cup. Thus, r5 and r6 are
appropriately positioned to send different inductive
signals (arrows) to anterior and posterior otic placode,
respectively. Because only the dorsal half of the placode
makes intimate contact with the hindbrain cells because of
the absence of a basal lamina between the two structures,
inductive signals that require direct cell contact may not
extend to the ventral rim of the otic field, as shown. (B)
It is proposed that the ventral tissue does not acquire
lateral identity until the cup stage, and is separated
from the dorsomedial tissue by a region that acquires
sensory competence. The signaling sources for either
lateral or sensory identity are unknown, but they could be
influenced by distance from the hindbrain and/or proximity
to lateral ectoderm or mesoderm. (C) A schematic of
otic vesicle formation, combining information from fate
mapping and gene expression domains of
SOHo and Bmp4. The
gray region in the center of the field is proposed to
correspond to a sensory-competent region that will
intersect with the broader gene expression domains to form
the sensory patches and perhaps the ganglion cells. Only
the formation of the anterior crista (ac) and posterior
crista (pc) is shown.
Abbreviations:
cd, Cochlear duct; ed, endolymphatic duct; r, rhombomere
(From Brigande et al., 2000).
Figure 6.17.
Pathways of cardiac
induction in vertebrates (From
Zaffran and Frasch, 2002).
O ptosis is essential for the developmental remodeling and
shortening of the complex embryonic outflow tract, i.e.
for the transformation of the initial tubular structure
between the single primitive ventricle and the aortic sac
into a permanent outflow tract connecting the ventricles
with the arterial trunks in 4-8 days chick embryos
(Watanabe et al., 1998) (on the neural control of
apoptosis see
Control of Apoptosis in Vertebrates
later in this chapter).
Besides the neural tube inductions, neurotransmitters, and
apoptosis, an essential role in the development of the
cardiovascular system play neural crest cells, which
migrate there to participate in formation of the heart.
The region of the neural crest that is involved in the
process is known as cardiac neural crest and is
located between the mid-otic
placodes
and the caudal limit of the somite 3 (Kirby et al., 1993).
Cardiac neural crest cells that migrate from the
neural tube/CNS
to the heart region differentiate into various mesenchymal
cell types that initiate formation of the outflow tract
and the smooth muscle of the aortic arches (Phillips et
al., 1987; Creazzo et al., 1998)
(figure 6.18).
Due to the important contributions of neural crest cells
in the development of the vertebrate heart, the latter is
regarded as a new heart, part of a new cardiovascular
system composed of modular units, only some of which
existed in ancestral invertebrates
(Fishman and Chien, 1997).
Neural crest cells also surround the pharyngeal arch
arteries and “populate aortico-pulmonary septum and
conotruncal
cushions prior to and during overt septation of the
outflow tract and surround the thymus and thyroid as these
organs form”
(Jiang
et al., 2000).
Figure
6.18. Schematic
representation of NCC contribution to the developing
arterial system in the chick (stage 40), as deduced from
quailchick chimeras and retrovirally infected embryos.
Density of dots corresponds with relative contribution of
NCCs to periendothelial tissue or the media.
Abbreviations:
AoAr, aortic arch; AoS, aortic sac; AsAo, ascending aorta;
BA, brachiocephalic artery; CA, carotid artery; CO,
coronary artery; DA, ductus arteriosus; DAo, dorsal aorta;
DsAo, descending aorta; PA, pulmonary artery; PT,
pulmonary trunk; SA, subclavian artery; and III, IV, and
VI, pharyngeal arch arteries (From Bergwerff et al., 1998).
Neural crest cells mutant s
for the type I receptor ALK2 in mouse embryos show
impaired
migration leading to morphological defects in the heart
outflow tract and aortic arch (Kaartinen
et al., 2004).
Ablation of the cardiac neural crest leads to anomalies of
the outflow and inflow tracts of the heart and the aortic
arch arteries (Miyagawa-Tomita,
1991) (figure 6.19).
Figure 6.19.
Flow chart to illustrate the functional and structural
consequences of cardiac neural crest ablation.
Abbreviations:
DORV indicates double-outlet right ventricle; PTA,
persistent truncus arteriosis; and Ao, aorta (From Kirby
and Waldo, 1995).
Vasculogenesis
and Angiogenesis
The processes of vasculogenesis (formation of a primary
network of capillaries or the vascular plexus) and the
differentiation of endothelial cells from their precursors
begin approximately at the time that the embryonic heart
starts to develop. A group of 5 VEGFs (vascular
endothelial growth factors)
types,
all of them downstream elements of signal cascades under
ultimate control of the CNS, have essential roles in
vasculogenesis and hematopoiesis (Carmeliet et al., 1996;
Liang et al., 2001), in stimulating proliferation of
vascular endothelial cells (Ash and
Overbeek,
2000), and migration of endothelial cells
(Esser
et al., 1998). Their action is mediated by receptors
tyrosine
kinases
Flk-1 (VEGF-R2) and Flt-1 (VEGF-R1) (Millauer
et al., 1993), with the latter being essential for the
formation of embryonic capillaries
(Fong et al., 1995). Another secreted protein, FGF-2
(fibroblast growth factor-2),
also
known as basic
FGF,
has been demonstrated to be a potent stimulator of the
embryonic angiogenesis when applied to the quail
chorioallantoic membrane at embryonic day 7
(Parsons-Wingerter
et al., 2000). FGF-2
is also secreted by chick
chorioallantoic
membrane and has a rate-limiting role in the
vascularization
of the
chorioallantoic
membrane (Ribatti
et al., 1995). A spatially differential supply of
VEGF-A
determines vascular branching pattern
(Ruhrberg
et al., 2002).
Evidence about
the
immediate regulation of angiogenesis by the nervous system
has been presented
recently. The neurotransmitter dopamine inhibits the
vasculogenic
and
angiogenic
actions of the cytokine
VPF/VEGF
(vascular permeability factor/vascular endothelial growth
factor) by inducing the
endocytosis
of the
VEGF
receptor 2 (Basu
et al.,
2001). But, probably the most compelling
fact so far on the direct involvement of the CNS in
the
embryonic angiogenesis was reported by
Mukouyama et al.
(2002). They observed that in mutant mice lacking sensory
nerves in limbs, large diameter vessels do not branch
normally into intermediate vessels, but directly into
small-diameter vessels. Moreover,
they observed that during embryogenesis, the sensory
neurons
themselves
secrete VEGF, which determines the cell differentiation
and patterning of arteries in their vicinity (Mukoyama
et al., 2002), thus explaining the old anatomic
observation on the general association of arteries
with
peripheral nerves. In the embryonic
limb skin,
the nerve is the principal source of
VEGF.
The nerve-derived VEGF164 induces in vessels’
endothelium the synthesis of NRP1 (neuropilin1), an
artery-specific coreceptor of VEGF164. This
fact may also explain why the arteriogenic effects of
sensory nerves are restricted to vessels in close
proximity of the nerves (Mukoyama et al., 2005; figure
6.20).
Based on the experimental
evidence investigators concluded that
Peripheral nerves provide a
template that determines the organotypic pattern of blood
vessel branching and arterial differentiation in the skin,
via local secretion of VEGF. (Mukoyama et al., 2002)
It is demonstrated that in
coculture with presomitic mesoderm, the neural tube is
able to induce formation of a perineural vascular plexus.
Based on the demonstrated roles of the neural tube in
vascular patterning, the neural tube is considered to be
the midline signaling center for vascular
patterning in higher vertebrates. (Hogan et al.,
2004). During embryogenesis, the brain and spinal cord
release signals (with VEGF as a central player) for
differentiation and migration of somitic angioblasts in
their direction and form around them the so-called
perineural vascular plexus (PNVP) (figure 6.21), as
a necessary source of oxygen and nutrients for the
development of the CNS in vertebrates.
Another example of the neural control of vasculogenesis is
formation of the retinal vasculature.
The lower
portion of the signal cascade regulating the development
of
the
retinal
vascular pattern includes retinal neurons, which by
secreting
PDGFA
(platelet-derived growth factor A)
stimulate proliferation of
astrocytes.
The latter in turn, by secreting
VEGF,
determine blood vessel formation in retina (West et al.,
2005; figure 6.22).
Note that the signal cascade for the development of the
retinal vasculature starts with retinal ganglion neurons.
Development of the Gastrointestinal Tract
The specification of different parts of this tract is
related to the spatio-temporal pattern of expression of
various Hox genes along the embryonic antero-posterior
(A-P) axis. But the ordered expression of Hox genes
during the embryonic development in vertebrates is
regulated by the hormone RA (retinoic acid). In turn, RA
is synthesized by enzymes RALDH (retinaldehyde
dehydrogenase) –1, –2, and –3 (Niederreither et al. 2002).
RA signaling plays a crucial role for the establishment of
the A-P axis, as well as for the development of numerous
organs, including the nervous system, lung, digestive
tract, kidneys, eyes, etc. RA synthesis is chiefly
function of retinaldehyde dehydrogenase (RALDH) enzymes.
Starting with the early embryonic development, the
synthesis of RALDH-2 is closely related to the neural
tissue and motor
neurons, which extend their axons to the
periphery
indicating a potential role of retinoic acid in nerve
influences on pheripheral differentiation.
(Berggren et al., 1999)
The neural tube, which is
adjacent to the gastrointestinal endoderm (figure 6.23),
may be an important source of the RA for the development
of the gastrointestinal tract.
In chick embryos, the neural tube/spinal cord in its
length up to the hindbrain has the highest levels of RA
which gradually decrease with the increased
distance from the neural tube (somites
à
lateral plate) (Maden et al., 1998) and in retinoic
acid-deficient mice lack of RA in endoderm causes agenesis
of the dorsal pancreas (Martin, 2005).
The neural tube/CNS is also crucially involved in the development of the gastrointestinal tract in another direct
way. Migratory cells from the neural tube reach the region of the prospective gastrointestinal tract and
participate in the development of the organs of the tract.
Figure 6.20.
Schematic models for nerve-mediated arterial
differentiation and vascular branching in the limb skin. (A)
Proposed sequence of events in vascularization of limb. A
low concentration of VEGFA, or a distinct nerve-derived
signal (‘factor X’) promotes nerve-vessel alignment,
followed by VEGFA/NP-1-dependent arteriogenesis in
nerve-aligned vessels. (B) VEGF promotes
arteriogenesis via an NRP1-mediated positive-feedback
loop. All vessels are initially equivalent. Nerve-derived
VEGFA promotes arterial differentiation and NRP1 amplifies
the VEGFA effect due to increased sensitivity to VEGF164
in vessels in close proximity to nerves (N).
Abbreviations: A, artery; NPR-1 – neuropilin-1, an artery-specific coreceptor for VEGF that is induced
by VEGF (From Mukoyama et al., 2005).
Figure
6.21. Neural
tube patterning of blood vessels – a model. (A)
Diagram showing cross-sectional view of a mouse embryo at
8.5 dpc. The boxed area is enlarged in B-D. (B)
Initially, the presomitic mesoderm does not contain
committed vascular precursor cells. (C) A VEGF-independent
signal from the neural tube induces FLK1 expression in a
subset of presomitic mesoderm cells. (D) FLK1
expressing angioblasts are now competent to respond to the
neural tube-derived VEGFA containing signal. The
angioblasts now express additional vascular markers, such
as FLT1, and they migrate and assemble the PNVP. Other
unidentified signals emanating from the neural tube may
also contribute to PNVP patterning in vivo.
Abbreviations:
dpc, day post-coitum; IM, intermediate mesoderm; LPM,
lateral plate mesoderm; NC, notochord; NT, neural tube;
PNVP - perineural vascular plexus; PSM,
presomitic mesoderm (From Hogan et al., 2004).
Very early during
hepatogenesis, VENT (ventral neural tube) cells originating
in the ventral regions of the neural tube appear in liver
trabeculae, where they express hepatocyte markers and
display hepatocyte morphology. At stage E5 VENT cells reach
the stomach and duodenum where they contribute to formation
of their mucosa and almost at the same time (E6), VENT cells
from the ventral neural tube reach other regions of the
gastrointestinal tract (Dickinson et al., 2004). The
above evidence shows that development of the
gastrointestinal tract depends on the patterns of RA
synthesis as well as on the cellular and inductive
contributions of the migratory neural tube cells. Given the
immediate proximity of the endoderm of the prospect
gastrointestinal tract to the neural tube, the latter may be
the source of the RA (in view of the very high secretion of
RA from the neural tube) necessary for the development of
the tract.
................................................................................
..............................................................................
>>NEXT>>
|
