Introduction

STIM1 is necessary for correct growth cone motility in vitro. Reduced expression of STIM1 in rodent DRG sensory neurons causes a switch in growth cone turning response to BDNF from attraction to repulsion (Mitchell et al., 2012). Similarly, perturbing STIM1 function by overexpressing a dominant-negative or constitutively-active from of STIM1 in Xenopus spinal neurons results in a switch in the turning response to netrin-1 from attraction to repulsion (Shim et al., 2013). Importantly, attractive growth cone turning in response to both BDNF and netrin-1 requires spatially-restricted, sustained elevations in intracellular calcium, which is reliant on calcium release from the ER (Gomez and Spitzer, 1999; Zheng, 2000; Gomez et al., 2001; Wen et al., 2004b; Gasperini et al., 2009). Therefore, given that the ER is a finite source of calcium, it has been hypothesised that STIM1-mediated SOCE is required for the large amplitude increases in calcium required for growth cone attraction (Li et al., 2005; Gasperini et al., 2009; Mitchell et al., 2012). STIM1 is also necessary for sema3a-mediated growth cone repulsion (Mitchell et al., 2012), and sema3a is considered a largely calcium-independent guidance cue that does not elicit calcium release from the ER (Togashi et al., 2008), suggesting that SOCE- independent functions of STIM1 also contribute to the regulation of axon guidance. The zebrafish spinal motor system is an excellent model in which to study the mechanisms that regulate axon pathfinding in vivo. Similar to other vertebrate species, zebrafish spinal motor neurons are classified into subtypes that exhibit distinct axon trajectories that enable each subtype to innervate the correct target (Eisen et al., 1986; Myers et al., 1986; Westerfield et al., 1986). In the zebrafish embryo, each somite is initially innervated by three primary motor neurons: the rostral primary (RoP), middle primary (MiP), and caudal primary (CaP) motor neurons, which are identified by their combinatorial expression of LIM homoebox genes and their anatomical position within each spinal hemisegment (Eisen et al., 1986; Appel et al., 1995). Primary motor neuron axogenesis begins with the CaP axon exiting the spinal cord via the ventral root at approximately 17 hpf (Eisen et al., 1986; Zeller et al., 2002; Hilario et al., 2010; Plazas et al., 2013). Approximately 2 hr later, the MiP, then the RoP, begin axon pathfinding within the spinal cord, growing caudally to contact the CaP axon before also exiting the

spinal cord at the ventral root (Eisen et al., 1986; 1989; Pike et al., 1992; Bernhardt et al., 1998).

Axon pathfinding by primary spinal motor neurons is highly stereotypic (Fig. 3.1). All three primary motor neurons extend away from the spinal cord along the common pathway (Eisen et al., 1986; Myers et al., 1986; Westerfield et al., 1986), which is a ventral projection from the spinal cord to the nascent horizontal myoseptum between the medial surface of the myotome and the notochord (Bernhardt et al., 1998). Once each primary motor neuron axon reaches the horizontal myoseptum, they pause for up to 1 hr, before making vastly different pathfinding decisions (Eisen et al., 1986; Myers et al., 1986; Melancon et al., 1997; Beattie et al., 2000; Plazas et al., 2013). The CaP axon extends distal to the horizontal myoseptum to reach the ventral aspect of the notochord, before continuing towards the ventral muscle boundary. By 48 hpf, the CaP axon has branched at the ventral muscle boundary, with one prominent branch extending in a dorsal-rostral direction to innervate the rostral myotome, and the other prominent branch extending dorsal-caudal direction to innervate the dorsal myotome (Myers et al., 1986). In contrast, once the MiP axon reaches the horizontal myoseptum, it sprouts a dorsal collateral growth cone, retracts the initial ventral axon, and grows dorsally to innervate the dorsal myotome (Eisen et al., 1986; Myers et al., 1986; Westerfield et al., 1986). Whilst the RoP axon branches at the horizontal myoseptum, with branches extending laterally to innervate the regions of the myotome located proximal to the horizontal myoseptum (Eisen et al., 1986; Myers et al., 1986; Westerfield et al., 1986). Hence, CaP axons represent an isolated in vivo model of axon pathfinding in which to study the importance of STIM1 for axon guidance.

Many of the cellular mechanisms underlying axon pathfinding by CaP axons have been elucidated, including motility at intermediate choice points, with a myriad of overlapping guidance cues and receptors identified to date (for reviews see (Beattie, 2000; Lewis and Eisen, 2003; Schneider and Granato, 2003; Fetcho et al., 2008; Bonanomi and Pfaff, 2010)). Consistent with Xenopus and mammalian models of growth cone guidance (Davies et al., 1986; Li et al., 2005; Gasperini et al., 2009), where the role of STIM1 in growth cone motility has previously been investigated (Mitchell et al., 2012; Shim et al., 2013), cultured zebrafish motor neurons express the BDNF receptor TrkB and exhibit growth cone attraction towards a source of BDNF (Chen et al., 2013). Furthermore,

111 Figure 3.1: Axon pathfinding by spinal primary motor neurons in highly stereotypic. Schematic illustrating the stereotypic axon pathfinding decisions made by the rostral (RoP), middle (MiP) and caudal (CaP) primary motor neurons of the zebrafish spinal cord to reach their distinct muscle targets.

113 zebrafish netrin-1 and sema3a orthologs, which exhibit a high degree of homology with mammalian proteins (Chen et al., 2013), are required for axon pathfinding by CaP neurons in vivo (Lauderdale et al., 1997; Shoji et al., 1998; Zeller and Granato, 1999; Sato-Maeda et al., 2006; Plazas et al., 2013). As STIM1 is necessary for correct growth cone motility in response to BDNF, netrin-1 and sema3a in vitro (Mitchell et al., 2012; Shim et al., 2013), axon pathfinding by CaP axons represent an excellent model to study the importance of STIM1 for axon guidance in vivo.

This chapter addresses the hypothesis that STIM1 is required for correct axon pathfinding in vivo by examining the effect of reduced zSTIM1a expression on axon pathfinding by zebrafish spinal motor neurons in vivo. STIM1 was previously shown to be necessary for vertebrate and invertebrate development (Eid et al., 2008; Oh-Hora et al., 2008; Stiber et al., 2008). However, no study has described the importance of zSTIM1a expression or function for zebrafish nervous system development or motor axon pathfinding. Therefore, the effect of reduced zSTIM1a expression on embryogenesis and axon pathfinding by spinal motor neurons was investigated.

3.2. Methods