FGF-2 is involved in maturation and target innervation

Nigrostriatal wiring occurs between E18 und P4, followed by maturation until P28.

During this developmental process in the substantia nigra those mDA neurons which failed to establish adequate projections to the striatum undergo natural cell death with main peaks at P2 and P14 (Burke, 2003, Burke, 2004, Prakash and Wurst, 2006). Most probably multiple trophic inputs are required for establishment of functionally adequate nigrostriatal projections. In other neuronal systems FGF-2 participates in neuronal network establishment in very diverse manner (compare (Umemori, 2009)). FGF-2 was shown to regulate axon guidance of spinal motor neurons (Shirasaki et al., 2006), target recognition of the Xenopus retinal ganglion cells (McFarlane et al., 1995, Webber et al., 2003, Webber et al., 2005), as well as synaptic differentiation, inducing synaptic vesicle aggregation in Xenopus spinal cord neurons (Dai and Peng, 1995) and promoting neurite elongation and branching of hippocampal neurons in vitro (Li et al., 2002) and regulate their spine morphology in vivo (Zechel et al., 2009). Consistent with this are the findings of the present study, suggesting that FGF-2 loss compromises adequate nigrostriatal pathway formation, including target recognition and maybe also path finding. First evidence therefore is provided by decrease of apoptotic bodies missing caspase-3 immunoreactivity at stage PO in FGF-2 deficient mice, indicating a probability of reduced wiring control allowing inadequately wired fibers and the respective cells to retain. Whether the significantly more caspase-3 negative apoptotic profiles within the wild type compared to FGF-2 deficient SNpc represent the late apoptotic stages (Brecht

et al., 2001) or a different caspase-3 independent apoptotic cell death (Jeon et al., 1999) remains unclear. Postnatal MFB lesioning experiments showed that besides the conventional apoptotic signaling routes a caspase-3 independent cell death in the mDA system is feasible (Jeon et al., 1999). The authors investigated developmental cell death in the SNpc, natural developmental neuron death, and induced developmental death following either striatal target injury with quinolinic acid or dopamine terminal lesion with intrastriatal injection of 6-OHDA. Caspase-3 dependent apoptosis was shown to occur during all three investigated cell death models. However, after 6-OHDA lesion only 16%

of apoptotic profiles were caspase-3-positive in contrast to 59% after striatal target injury. Since the immunohistochemical techniques were unchanged, the authors suggested, that this difference might be due to mDA specific toxicity of 6-OHDA.

Therefore they discussed a possibility of a different caspase-3 independent apoptotic process in this context.

The reduced ontogenic cell death is correlated with increased fiber outgrowth in FGF-2 deficient VM and FB explant co-cultures. The heterogenous co-cultures missing FGF-2 in FB or VM, respectively, show a similar phenotype. Compared to co-cultures lacking FGF-2 in both VM and FB they have significantly shorter fibers. Additionally, if compared to pure wild type co-cultures they show significantly wider tracts. This indicates a complex interplay between mesencephalic and telencephalic and/or diencephalic FGF signaling during pathfinding. Previous VM explant culture studies revealed a biphasic TH-ir fiber outgrowth, first occurs the glia-independent straight long distance outgrowth followed by astrocyte-dependent network forming short distance fiber outgrowth (Johansson and Stromberg, 2003). Another study indicates that GDNF specifically regulates the astrocyte dependent TH-ir fiber outgrowth, by inducing the migration of astrocytes (Bjerken et al., 2007). In fact, GDNF has been hypothesized to be a leading striatum-derived neurotrophic factor for mDA neurons (Burke, 2006). One study also reports a cooperative requirement of GDNF for FGF-2 mediated neuroprotection in

hippocampal neuron cultures, showing a regulatory function of FGF-2 on GDNF expression (Lenhard et al., 2002).

However, it remains to be resolved how FGF-2 loss affects the astrocytes during the nigrostriatal pathway formation, which isoforms participate hereby as well as what are the sources of FGF-2 and the respective receptors. On the other hand a retrograde (Ferguson and Johnson, 1991) as well as anterograde transport of FGF-2 within nigrostriatal pathway (McGeer et al., 1992) imply a paracrine or autocrine mechanism involving the secreted 18 kDa FGF-2 isoform in maturation and maintenance of the nigrostriatal system. Hereby, the reports of induction of apoptosis via FGFR3 in the peripheral nervous system (Jungnickel et al., 2004) as well as via FGFR2 in cortical systems (Maric et al., 2007) support a hypothetical involvement of the secreted FGF-2.

However, additionally to extrinsically mediated, programmed cell death pathways, also a direct effect of intrinsic mDA derived FGF-2 should be considered. Apart from the fact that the LMW-FGF-2 is the predominant form during development, the 22kD FGF-2 is also present in developing rat brain (Giordano et al., 199FGF-2), suggesting that HMW-FGF-2 might also be involved in the regulation of adequate nigrostriatal pathway formation. In fact, the HMW-FGF-2 has already been shown to induce apoptosis in vitro (Ma et al., 2007). Accordingly, there are several putative sources for FGF-2 which may ensure intact pathfinding and an adequate response to natural cell death initiation (Fig.

18): 1.) extrinsic VM derived FGF-2, which activates FGF-signaling via transmembrane FGFR1; 2.) extrinsic FGF-2 produced by striatal astrocytes can also activate the FGFR located on the membrane surface or become internalized by dopaminergic cells and be transported to the soma; 3.) internalized extrinsic or intrinsic mDA neuron derived FGF-2 may modulate intracellular responses of other signaling pathways, in the soma, nucleus or axonal growth cones.

Figure 18. Putative sources of FGF-2 during nigrostriatal pathway formation. 1.) extrinsic VM derived FGF-2, which activates FGF-signaling via transmembrane FGFR1; 2.) extrinsic FGF-2 produced by striatal astrocytes can also activate the FGFR located on the membrane surface or become internalized by dopaminergic cells and be transported to the soma; 3.) internalized extrinsic or intrinsic VM derived FGF-2 may modulate intracellular responses of other signaling pathways, in the soma, nucleus or axonal growth cones.