All the neuronal function that we perform in day to day life from cognition to behavior depends upon the intricate network of neurons joined together in form of a circuit in our nervous system. The formation of the neural circuit is a key process during early embryonic development that requires accurate and precise connectivity of neurons towards its synaptic target. During development of the nervous system each neuron fires its axonal projection towards its target which is guided by several guidance molecules that guides the axons towards its target .The tip of the axon and dendrites has a special structure known as ‘the growth cone ‘ that senses the guidance cues and helps in directional motility of the growth cone. (Lowery et al). The Growth cone during its journey to its correct synaptic partner faces a number of guidance cues that either substrate-bound or diffusible chemotropic molecules in the environment that guide towards its target. These guidance cues acts in additive or non-additive manner to either give an attractive or a repulsive signal to the growth cone ( Sahasrabudhe, et al,.(2012) There are receptors present on the growth cone that interacts with the extracellular guidance cues providing spatiotemporal information about the environment. These information are further integrated and interpreted by the intracellular signaling machinery and are transmitted to the effector system which in turn generates appropriate response. These response can range from an alteration of growth cone motility rate to its directional movement i.e. attraction or repulsion from the guidance cues.
Unlike any other motile cells, growth cone also has a high concentration of cytoskeleton dynamics that helps in proper navigation of the growth cone. Regulated cytoskeleton dynamics forms the basis for several key processes in growth cone driven motility which includes directional polarization, generation of protrusive extension and their stabilization and the development of mechanical and traction necessary for translocation. Actin and Microtubule (MT) primarily dominate the cytoskeleton of the neuronal growth cone. The actin component is present at the tip of the growth cone and helps in formation of filopodia and is involved in generation of protrusive and contractile forces in the cell with the help of myosin motor proteins (Lowery et al). Microtubules are less flexible and form a rigid, polarized network inside the growth cone and are critical for polarity and directional migration of the of the growth cone. The intermediate filaments are less abundant in the growth cone and their exact function is still elusive. The intermediate filaments lack polarity and have suggested a relative passive structural role. The actin and MT system are very much coordinated and regulated to bring efficient motility. Rho-GTPase signaling pathway is implicated in coordinated regulation of actin and microtubule cytoskeleton. Growth cone like other motile cells form adhesive contacts with the substrate, which is mediate by mechanical coupling between intracellular cytoskeleton elements and extracellular matrix (ECM).This further allows proper transduction of intracellular forces to the substrate resulting in traction pull. Thus the motility of growth cone is not smooth; it first pulls itself in forward movement which is opposed by the traction force by the substrate. The exact mechanisms mediating these processes remain poorly understood. The cellular and molecular interaction underlying the establishment of the neuronal circuitry has been very much challenging since decades. It has been just about two decades that with the integrated help of molecular cell biology, genetics, imaging technology and biophysical techniques, scientist have come to know integrated framework of the mechanism of growth cone motility, and the detailed molecular interaction in the establishment of neural circuit. However, the complete mechanism still remains elusive.
1.1.1 Structure of the growth cone:
The growth cone is the specialized, motile tip of growing neuronal processes that explores the external space for guidance cues and brings about directional motility. The growth cone has two types of protrusive structures with distinct morphological features- ‘at, sheet-like extensions called lamellipodia and long, ‘nger-like projections called as ‘lopodia (T.M. Gomez et al.) These structures not only differ in their morphologies but also in their underlying sub-cellular architecture. The lamellipodia are produced by the protrusive activities of a dense meshwork of cross linked actin while filopodia consists of long bundles of filamentous actin (F’actin) (Gomez et.al) Filopodial extensions dominate mobility of the growth cone. The filopodia is highly dynamic, and undergoes bouts of extension and retraction which make them ideally suited for exploring the environment for navigation of the growth cone. It has been demonstrated in the literature that filopodia plays the key role in sensory module and it bears necessary guidance receptors to interact and respond to the guidance cues. (E.W. Dent et al.)
Based on the underlying cytoskeleton dynamics, the growth cone is divided into three basic domains or regions : The Peripheral (P) domain contains long, bundled actin filaments (F-actin bundles), which form the filopodia and lamellipodia of the growth cone. The central (C) domain contains stable bundled Microtubules that enter the growth cone from the axon shaft. It also contains numerous mitochondria, vesicles and central actin bundles. (Gomez et.al and Dent et.al) and the transition (T) zone lies at the interface of P and C domain and contains acto-myosin contractile structures lying perpendicular to the F actin bundles forming a hemi-circumferential ring.
Figure 1. General structure of an extending growth cone. The P domain is dominated by dynamically extending and retracting filopodia, which senses the local environment for guidance cues. Growth is consolidated in the direction of filopodia that pick up positive guidance cues, while filopodia that sense negative cues are rapidly retracted. ((Lowery and Van Vactor, 2009)
1.1.2 Motility of the growth cone:
The motility of the growth cone takes place by a series of changes that could be characterized as
Protrusion, Engorgement and Consolidation (Helmbacher, F. et al. and Kania, A et al.) During navigation,as soon as the distal end of growth cone encounters a guidance cues, it activates a series of intracellular signaling pathways and starts forming a molecular clutch that links the substrate to the cytoskeleton elements. After formation of a clutch, it protrudes its Filopodia and lamellipodia and extends forward. This mechanism further strengthens the molecular clutch and anchors the actin to its substrate. Further, F-actin polymerization continues in front of the clutch site, resulting in the forward movement of peripheral domain. Therefore growth cone protrusion phase may be characterized by increased T domain and elongated filopodial extension.
In Protrusion phase the filopodia and lamellipodia extends forward and just after the protrusion phase, actin filaments move away from the region of adhesion and C domain characterized by the removal of actin clutch. As F actin clutch is removed, it re-orients itself from the C domain towards a new growth site. This is further followed by the microtubules invasion into this region, guided by T zone actin arcs and C domain actin bundles. Therefore, during engorgement phase, the C domain extends further while T zone is shortened.
Finally to form new segment of axon shaft, advanced C domain are consolidated at the proximal end of the growth cone and myosin II containing actin arches compresses the microtubule into a newly localized C domain, which is then followed by microtubule associated protein stabilization. (Suter et al.,) During the retraction of filopodia, F actin protrusive activity is suppressed in the area of retraction which is also promoted by Myosin II activity, thereby promoting consolidation of axon shaft.
These three stages of growth cone i.e protrusion, engorgement and consolidation are continuous and overlapping stages involved in formation of nascent axons, during new growth cone formation and axon branching.(Dent and Kalil et al.)
Figure 2: The stages of axonal outgrowth .Protrusion, engorgement and consolidation. (Lowery et al.)
1.1.3. Cytoskeleton makeup of the growth cone:
As we know that the change in cytoskeleton dynamics plays a very important role in growth cone migration. The rearrangement of cytoskeleton like F actin polymerization and microtubule assembly and disassembly plays the major role in growth cone motility.
The major proteins that form cytoskeleton of the growth cone and can be classified as:
‘ Actin Filaments or F-actin.
‘ Microtubule
Actin filaments:
Actin filaments are also called as microfilaments. They are double helical structure of actin protein. They are very flexible structure with a diameter of 5-9 nm and organize into a variety of linear bundles, two dimensional network and three dimensional gels. Though actin filaments are distributed throughout the cell but mostly they are concentrated at the cortex below the plasma membrane. A typical actin filament has two ends, one is plus end or barbed end and other is minus end or pointed end. To form an actin polymer, several actin monomers are added to either ends.
In growth cone, actin plus end or barbed end points toward the cell membrane and ATP actin is usually added to this end. Following hydrolysis of ATP actin forms ADP actin and ADP actin disassembles at minus end of actin filament that faces toward the T zone. The MBPs (monomer binding proteins) than comes into the picture and transports actin back to the leading end to facilitate further growth. [Pak, C. W., Flynn, K. C. & Bamburg, J. R. Actin-binding proteins take the reins in growth cones. Nature Rev.Neurosci. 9, 136’147 (2008).
Figure 9: Showing actin polymerization in growth cone.( Lowery et al.)
Microtubules:
Microtubules (MTs) are long hollow cylinders made up of tubulin. They are rigid structure than actin filaments and have a diameter of 25 nm. Microtubules (MT) are polarized structures and comprises of ?? and ?? tubulin dimers that are linearly assembled. A Proto-filament is formed by alternating linear array of ?? and ??-tubulin subunits.11-15 protofilaments combine to form the wall of Microtubule.
Like actin polymer, microtubules also have a plus and a minus end. Unlike in actin polymer where ATP and ADP is involved, in this case GTP and GDP is involved. Here GTP tubulin dimers are added at the plus end and following hydrolysis GDP tubulin dimers are dissociated from the minus end.
Figure 11: Microtubule and its organization.
In the case of growth cone, the microtubules exhibit dynamic instability and MT plus end faces towards the periphery. Its dynamic instability alternates between their relatively slow growths from their plus end to rapid plus-end disassembly that might be followed by recovery of plus-end assembly. During the growth cone motility, the MTs cycles through periods of growth, shrinkage and occasional pausing (Mitchison, et al.)
Numerous proteins are recognized in this process that binds to microtubules. Some of these proteins stabilize it like MT-associated protein 1B (MAP1B) while some are known to act as MT motors. eg Dynein and Kinesin (Hunter, A. W et al.) There are some other proteins also involved that are a part of protein family called as plus end tracking proteins (+TIPs) (Riederer et al.)
During navigation of growth cone towards its target, several signaling cascades are activated as a result of interaction of growth cone receptors with the guidance cues, thus bringing about dynamic cytoskeleton remodeling of the growth cone. The guidance cues are the several intracellular as well as extracellular molecules that are responded by the selective receptors present on the growth cone, bringing about directional navigation of growth cone.
The downstream effectors of Rho GTPases regulate all known aspects of the actin cytoskeletal machinery that affects growth cone protrusion, including F-actin polymerization and its retrograde flow. Formins, being among the major effectors of the Rho GTPases have therefore been implicated in numerous aspects of cytoskeletal regulation at the growth cone. For example, Diaphinous formins have been found to localize to the tips of growing filopodia and are required for their formation (Pellegrin et al.; Schirenbeck et al.)
1.2. Actin dynamics and its regulation:
1.2.1 Actin nucleator proteins:
Like any other cellular morphological processes, axon guidance is also dependent on the actin cytoskeleton dynamics. During developmental processes, cells receive a intracellular or extracellular signal and respond to it by the morphological changes including flattening, rounding up, dividing, establishing or breaking contacts with other cells or surfaces, and extending or retracting processes. All of these changes require quick and precise remodeling of the cytoskeleton.
The elongation process of actin filaments involves the addition of actin subunits while depolymerization is exactly the reverse. In most of the cytoskeleton structures, spontaneous assembly of actin subunits is not kinetically favored and the initial nucleation process is the rate limiting step in actin polymerization process. There are certain proteins like profilin that binds to free actin monomers that further hinders the spontaneous nucleation of actin filaments in vivo. (Goode and Eck, 2007).Therefore other cellular factors are required to aid and actively nucleate actin filaments. There are three major classes of actin nucleators identified till now that aids in polymerization of actin filaments.( Chesarone and Goode, 2009)
The first class of actin nucleators proteins include Actin related proteins (Arp) 2 and 3, sometimes collectively referred to as the Arp 2/3 complex. These proteins are 45% identical to actin and nucleate actin at the pointed end leading towards elongation of actin filaments at the barbed ends. The Arp2/3 complex binds to the existing filament of actin with Scar protein. Scar protein (a WASp releted protein) activates the Arp2/3 complex and in turn Arp2/3 complex results in addition of actin monomers hence forming a branch. Since Arp2/3 complex nucleate at 70?? angle, therefore it is attributed towards the formation of web like lamellipodia which helps the growth cone during navigation.
Figure : Activation of Arp2/3 complex by Scar (a WASp related protein) leading to branching of actin filaments. (Machesky, et al.,)
Formins comprise a second class of actin and function as hinged dimers that bind two actin monomers. Unlike the Arp2/3 complex, Formins follow the barbed ends of filaments during growth. Additionally, Formins also binds to Profilin and forms Actin-profilin complex which further accelerates actin monomer additions to the barbed end .Thus Formins are particularly implicated in the assembly of linearly bundled actin filaments.
Fig: Actin nucleation by Formin class of actin nucleators (Pollard et al,.)
The third class of actin nucleators are called as tandem-monomer-binding nucleators. This group
of actin nucleator comprises of proteins such as SPIRE and cordon bleu (COBL). They have their tandem G-actin binding motifs (typically WH2 domains), by which these proteins bring together actin monomers to form an actin nucleation seed (Quinlan et al.,; Ahuja et al.,).
Fig: A comparative analysis of the three classes of actin nucleators (Weston et al.,)
1.2.2. Formins , dynamics and regulation:
Formins or Formin Homology (FH) proteins are a class of widely expressed actin nucleator protein involved in nucleation and regulation of dynamics of actin polymerization and disassembly. Formins are dimeric proteins (Moseley et al, 2004), that acts as a key molecular regulator of cytoskeleton remodeling by novel mechanism. Each formin dimer captures two actin monomers during the process of actin nucleation (Goode and Eck, 2007).
Formins are the major effectors of Rho-GTPases and are usually activated by them. The first
formin to be discovered was Formin 1 (Fmn1), so named due to the notion that its disruption
caused limb deformities in mice, coupled with severe kidney defects (Zeller et al, 1989).
These defects were later attributed, however, to another nearby gene (Zuniga et al, 2004); but
the name is still applied to the entire family of proteins.
Eukaryotic species in general contain multiple homologous formin genes. For example, in mammals, fifteen different types of formin genes have been identified (Higgs and Peterson, 2005), while in yeast, it may comprise of two or three types of formins (Rivero et al, 2005). Formins across distantly related species have been characterized, and most formins have been found to share multiple regions of homology including three major regions of homology. FH 1 and FH2 domains are the chief Formin homology domains (Goode and Eck, 2007). FH2 domain has highly conserved 400 residues that is indispensable to formin function (Evangelista et al, 2002; Sagot et al, 2002).
Figure : Domain structure of the Diaphinous-related formins.
GBD: GTPase-binding protein; DID: diaphinous Inhibitory domain; FH: Formin Homology; GAP: G-protein activating protein; GEF: Guanine nucleotide exchange factor; DAD: diaphanous
autoregulatory domain; DD: dimerization domain. Unless activated by interactions of the GBD domain with GTP-Rho, the formin is maintained in an auto-inhibited state by binding of DAD to DID (Good and Eck, 2007)
In mammals, there are fifteen types of formins which have been divided into seven classes based upon their divergence of FH2 domains (Higgs and Peterson, 2005):
1. Original formins (Fmn)
2. Diaphanous formins (Dia)
3. ‘Inverted’ formins (INF)
4. Formin homology domain containing proteins (FHOD)
5. Disheveled-associated activators of morphogenesis (Daam)
6. Delphilin, and
7. Formin-related (or formin-like) proteins identified in leukocytes (FRL, FMNL)
The Actin filaments nucleated by formins, like those nucleated by the Arp 2/3 complex, grow at their barbed end. But in contrast to the Arp 2/3 complex, formins only nucleate linear unbranched actin filaments. Also unlike the Arp 2/3 complex, the FH2 domain does not share any structural similarity with actin. One mechanism that has been suggested for formin nucleation involves binding of the FH2 domain to actin polymerization intermediates, such as dimers (Pring et al, 2003; Xu et al, 2004) or trimers, and their subsequent stabilization to form the nucleus.
One of the more interesting activities of formins is that of processive capping. In the study by
Pruyne et al, Bni1pFH1FH2 was found to remain persistently associated with the growing barbed end even after actin nucleation, in contrast to the Arp 2/3 complex (which associates with the pointed end). Once activated, the FH2 domain forms an anti-parallel dimer. The processive capping activity of formins is facilitated by this dimerization of the FH2 domain (Harris et al, 2006), since disruption of dimerization prevents nucleation and processive capping (Moseley et al, 2004). Following successful nucleation, the FH2 dimer promotes the continual incorporation of more actin monomers at the barbed end, whilst remaining associated with that end by means of a mechanism that involves alternating contacts of the two FH2 monomers with barbed end actin subunits by means of identified actin-binding sites on the FH2 domain (Xu et al, 2004). Some of the most convincing evidence for processive capping has come from a study by Kovar and Pollard. When glass slide-immobilized formins are used to nucleate actin, the actin filaments grow away from the glass slide while remaining bound to the glass slide (Kovar and Pollard, 2004), indicating clearly that formins remain associated barbed ends of growing actin filaments even after nucleation and guide rapid insertion of new subunits. Furthermore, GFP-labelled Dia1 molecules are found to travel inside cell in straight paths at rate similar to those of actin polymerization (Higashida et al, 2004), offering more substantiation for processive capping. In addition to facilitating the addition of new subunits at the barbed end and thereby accelerating actin polymerization and filament elongation, processive capping also serves to prevent the binding of capping proteins.
Fig: Mechanism of Formin action (Baarlink et al, 2010)
Enabled/vasodilator-stimulated phosphoproteins (Ena/VASP) also aid elongation of actin filaments in a manner very similar to that employed by the formins. They too associate with growing barbed ends of actin filaments and prevent binding of capping proteins. They also seem to promote formation of unbranched linear actin filaments: in Ena/VASP-deficient lamellipodia, actin networks contain shorter filaments and are more highly branched, while excess Ena/VASP in lamellipodia essentially causes the reverse morphology, with longer, less branched filaments (Bear et al, 2002).
Monomeric actin subunits found in living cells is predominantly present in a form bound to profilin. ATP bound to polymerized actin is hydrolysed to ADP during depolymerisation. Binding of profilin to free actin subunits promotes exchange of bound ADP for ATP, recycling actin monomers to a pool of re-useable actin for addition to elongating filaments (Pollard et al, 2000). In addition to binding actin subunits, profilin also binds polyproline sequences. This allows the FH1 domains of formins, which is proline-rich, to interact with profilin (Chang et al, 1997). Profilin-FH1 interactions accelerate the rate of subunit addition to actin barbed ends and thereby elongation.
FH1 and FH2 are both located at the C-termini of the formins. No clear homology exists, however, at the N-terminal regions of members of the FHOD class of formins and the original namesake formins, to which class Formin-2, the major subject of this study, belongs. The N-terminal region, with the FH3 domain, serves a largely auto-regulatory function by intra-molecular binding .The formin stays thus until activated by an upstream Rho GTPase. Given that the carboxy termini of the formins, which include the FH1 and FH2 domains, are largely conserved, it has been postulated that a fair degree of redundancy also exists in their functions in most organisms (Leader et al, 2000).
1.2.3. Formin-2 as a cytoskeletal regulator
Formin-2 (Fmn2) is a formin homology protein that was first identified in the murine and human systems. It was found to share regions of high degrees of homology with the known formins Fmn1 and the Drosophila formin cappuccino, and was highly expressed in the developing and mature central nervous system (Leader et al, 2000). Since then, it has been implicated in a number of cellular processes involving actin and/or microtubule remodeling in diverse cell types across multiple species, such as asymmetric cell division in mammalian oocyte maturation (Kwon and Lim, 2011), human adipogenesis (Peng and Liou, 2012), and vesicle transport (Schuh, 2011). Fmn2 has also been shown to interact with Spir1, a member of the Spire family of WH2-containing actin nucleators (Vazcarra et al, 2011) to drive asymmetric oocyte division. Given such varied functions, coupled with the observation that cappuccino, its Drosophila ortholog, interacts with Slits (unpublished observations), Fmn2 is a prime candidate for being a novel regulator of cytoskeletal dynamics at the growth cone during axon guidance.
1.3 Studying axon guidance in zebra fish
Zebra fish is a teleost that has gained significant prominence in recent year as a model system to study several developmental processes due to ease of its easy maintenance, rapid growth, small size and their transparent body that allows to observe the developmental phenomenon live in the embryo. More importantly, it is relatively simple to generate gene knockdown embryos in zebra fish, such as by using morpholinos (Sumanas and Larson, 2002), than in other systems. The neural trajectories are easy to detect and study in detail at different stages of zebra fish due to its optically transparent brain. A few specific tracts of axon guidance have received especial attention in the last few years as models to study its functioning. One of such model that has proven particularly informative is the escape response network. The escape response network in zebra fish includes two mauthner cells on both the hemisphere near rhombomere 4 in hind brain that are the major regulator of the escape response network of the zebra fish. (lorent et al,.2001). Mauthner cells receive sensory input directly at their dendrites and send an axon to the spinal cord to deliver output to contract the trunk muscles along the contralateral side of body (Kohashi et al,.2008). There are several commissural neural that crosses that cross the midline and reaches the contralateral side plays an important role in the behavioral pattern associated with escape response. The study by lorent et al,.2001 implicated the role of one such commissural neuron called as ‘spiral fiber neurons’ present in rhombomere 3 in hind brain that controls the fast ‘C Turning movement ‘ of zebra fish. The ‘C Turning movement’ is when a fish is touched on the left side, it immediately turns its body toward right side. This is facilitated by the mauthner cells and its associated neural trajectories (lorent et al., 2001)
This neural network is of particular interest because of the fact that during establishment of this escape network, several axon trajectories crosses the midline and midline crossing is brought about by a very tight regulation of the growth cone motility by the guiding molecules that alter the cytoskeleton dynamics of the growth cone, thus facilitating the axonal motility. The guidance molecules and receptors like Slit /Robo, Semaphorins play a significant role in guidance the growth cone at the midline. Erskine et al, 2000; Ringstedt et al, 2000)
One other neural network that has proven informative is the visual system development pathway, which involves axons of Retinal Ganglion Cells navigating their way from the cell bodies in the retina through a series of critical decision points to reach their final targets in the visual processing center in the brain, called the optic tectum in fish (Erskine and Herrera, 2007). One such critical decision point is the optic chiasm, a major brain commissure on the midline where the optic nerves from the two eyes meet. Here, the growing axon must either chose to stay on the same body hemisphere and project ipsilaterally or cross over and project contralaterally. In species with eyes located laterally, such as zebra fish, which do not have binocular vision, all axons cross over at the chiasm to the other side .The retinotectal projection system has been used to identify many genes involved in axon guidance (Baier et al, 1996; )
Fig: A.) Mauthner(M) cells controlling the escape behavior in zebra fish (Kohashi et al,.2008)
B.) Retinotecctal projection in zebrafish, showing complete contralateral projection (Culverwell and Karlstrom, 2002)
1.4. Morpholino over other knock out strategies
1.5. Development of zebra fish:
1.5. Confocal Microscopy: It is an optical imaging technique ehich is used to increase optical
resolution and contrast of a micrograph by using point illumination and a spatial pinhole to
eliminate out-of-focus light in specimens that are thicker than the focal plane . It enables to
look at three dimensional structures of the sample. The images could be taken in spatiotemporal
manner in three dimensions i.e. X, Y and Z along with time as 4th dimension. This microscopic
technique has been applicable in the scientific and industrial fields and mainly in life sciences, semiconductor inspection and materials science. Confocal is able to image with an average resolution of 60 to 70nm but it can image the whole sample in three dimensions by taking stacks of each plane towards Z plane.
Chapter 2: Genesis of hypothesis and objectives of
the project
The stereotyped neural circuit development underlies the basic functioning of the nervous system. During early development all neurons extends their axons mediated by special structure at its tip called as ‘the growth cone’ toward its synaptic targets in order to form an efficient and accurate neural network. While numerous factors and diverse molecules have over the years been found to aid this process, the significance of the role of cytoskeletal regulators at the growth cone in this context can hardly be overstated. Traditionally, the major known nucleator of actin cytoskeletal elements at the edges of most cell protrusions has been the Arp 2/3 complex. Owing to the very nature of the mechanism employed by the Arp 2/3 complex in nucleating actin, it leads to the formation of branched mesh-like actin networks, which form the mainstay of lamellipodia occurring at the leading edge of most cell protrusions in non-neuronal cells.
However, on a morphological level, the leading edges of growth cones have profoundly different cytoskeletal architecture. For precise and prompt axon guidance to occur, the growth cone must possess immaculate sensory capabilities. In order to achieve this, it is more practical and efficient to send out long but less substantial sensory processes in different directions, and then consolidate growth in directions where positive cues are picked up, than to send out a veil of cytoplasm. To fulfill this purpose, filopodia dominate the leading edges of growth cones. The cytoskeletal architecture of filopodia is composed chiefly of linearized actin bundles found at the growing tip of filopodia. Expectedly Arp 2/3-mediated nucleation is irrelevant here as it leads to the branching of the actin filaments, but the more recently identified formins are crucial, since they nucleate linear actin filaments. Formin-2 (Fmn2), in particular is found to be enriched in developing nervous systems of vertebrates, and is this led us to hypothesize that it may have a role in axon guidance specifically and neural development in general.
The zebrafish (Danio rerio) model system was chosen for our study for primarily three reasons: their embryos are transparent, allowing for direct and easy observation of developing embryos; they are easily amenable to gene expression modifications; and they provide a relatively simple vertebrate system to investigate. Only embryonic stages were investigated as studies on adults would require the time-consuming process of generation of knockout mutants.