**5. Impaired transport, more places to go**

178 Amyotrophic Lateral Sclerosis

 1(Quinlan et al., 2011),2(Kuo et al., 2005), 3(van Zundert et al., 2008), 4(Meehan et al., 2010), 5(Carunchio et al., 2010), 6(Pieri et al., 2009), 7(Bories et al., 2007), 8(Amendola and Durand, 2008), 9(Saxena et al.,

2009), 10(Johnston et al., 2000), 11(Turner et al., 2003b), 12(Mourelatos et al., 1996), 13(Li et al., 2010), 14(Jaiswal and Keller, 2009),15(Mattiazzi et al., 2002), 16(Nguyen et al., 2009), 17(Jaiswal et al., 2009), 18(Bilsland et al., 2008), 19(Damiano et al., 2006), 20(Bilsland et al., 2010), 21(De Vos et al., 2007),

22(Williamson and Cleveland, 1999), 23(Zhang et al., 1997), 24(Kieran et al., 2005), 25(Warita et al., 1999), 26(Alexianu et al., 2001), 27(Gowing et al., 2008), 28(Chiu et al., 2008), 29(Chiu et al., 2009), 30(Fischer et al., 2004), 31(Frey et al., 2000), 32(Pun et al., 2006), 33(Hegedus et al., 2007), 34(Hegedus et al., 2008), 35(Kennel et al., 1996), 36(Bruijn et al., 1997), 37(Munch et al., 2002), 38(Tortarolo et al., 2006), 39(Bendotti et al., 2001),

Fig. 1. Timeline of deficits in mutant SOD1 mice. Earliest reported deficits in the above properties are used. Different SOD1 mutants were normalized to dates of overt symptom onset. When differences in timing between mouse lines were large (as it was for protein ubiquitination, stress of the ER, and activation of astrocytes), the range is indicated in the timeline with (///). † Also found in embryonic cultured motoneurons. **\*** Different aspects of mitochondrial function were impaired at different time points. The first alteration in

membrane potential, is not altered until just before symptom onset17, while the function or

40(Dal Canto and Gurney, 1995), 41(Dal Canto and Gurney, 1994), 42(Kaiser et al., 2006).

function is decreased Ca2+ storage capacity19. Another property, mitochondrial

regulation of the electron transport chain is impaired slightly before membrane

potential16.

The lack of Ca2+-buffering proteins in vulnerable motoneurons make the mitochondria even more critical to their function. Mitochondria are normally highly mobile both in axons and dendrites (MacAskill et al., 2010). Mitochondrial movement can be halted by increased concentrations of ADP, so they tend to remain in compartments which are highly metabolically-active (Mironov, 2007). Mitochondrial movements are also regulated through Ca2+ signaling and synaptic activity (Rintoul et al., 2003, Yi et al., 2004, Macaskill et al., 2009). When glutamate binds NMDA- or certain AMPA-type receptor-channels, it allows the influx of Na+ and Ca2+ into the cell. The Ca2+-sensitive domain of Miro, the mitochondrial trafficking protein, then interacts with Ca2+ and the transport factors TRAK and KIF5, and pauses in its movement at active synapses (Rintoul et al., 2003, MacAskill et al., 2009). Postsynaptic NMDA receptors are also associated with PSD95 and with nitric oxide synthase (NOS) which, through nitric oxide (NO), also pauses mitochondrial movement (Rintoul et al., 2006). Once at a synapse, the mitochondria are probably tethered by neurofilaments, a process that depends both on the state of phosphorylation of the neurofilaments and a high mitochondrial membrane potential which indicates a high level of activity (Wagner et al., 2003).

In axons, but not in the soma of cultured SOD1 motoneurons, mitochondria are more sparsely distributed (De Vos et al., 2007), and *in vivo* mitochondria show more frequent pauses in their movements in pre-symptomatic SOD1 mice (Bilsland et al., 2010). Unfortunately, movement of mitochondria and other membrane-bound organelles has not yet been well studied in the dendrites of SOD1 motoneurons. If the mitochondria are similarly sparse in the dendrites, where most Ca2+ channels are located, this could have serious consequences for Ca2+ buffering. Spinal motoneurons of SOD1 mice show a significant proliferation in dendritic branches (Amendola and Durand, 2008) and an increased Ca2+ PIC (Quinlan et al., 2011), which could make mitochondrial motility in the dendrites more challenging. Without mitochondria to take up Ca2+ at the synapses, this would further exacerbate the low Ca2+ buffering in vulnerable motoneurons and any increased Ca2+ entry with synaptic inputs (Tortarolo et al., 2006). It is also worth noting that

Molecular and Electrical Abnormalities in the Mouse Model of Amyotrophic Lateral Sclerosis 181

In summary, not only are there fewer mitochondria present in the processes of SOD1 neurons (De Vos et al., 2007, Bilsland et al., 2010), but those that are present are impaired in functioning. This is likely to have dire consequences for both Ca2+ buffering and ATP

Misfolded proteins are degraded through autophagy (Yang and Klionsky, 2010). When the capacity of the cellular machinery in the ER to properly fold proteins is exceeded, cells react with the unfolded protein response (UPR) and signs of ER stress (reviewed by Ron and Walter, 2007). The UPR decreases most protein synthesis in the cell while upregulating synthesis of some ER proteins that assist in proper folding and processing of proteins. Another pathway, known as ER-associated protein degradation (ERAD), helps to clear the ER of misfolded proteins by exporting them to proteasomes where they are broken down (Bernasconi and Molinari, 2011). Proteins to be exported and degraded are marked by ubiquitination, a process in which ubiquitin molecules bind to the protein, tagging it for destruction (Bingol and Sheng, 2011). Normal ER function can be disrupted by blocking the ER-resident proteins from folding properly, inadequate functioning of the ubiquitinproteosome system, or failure to maintain a high level of Ca2+ inside the lumen of the ER

It is known that mice with the highest expression levels of mutant SOD1 protein have the earliest disease onset (Wong et al., 1995), and that markers for ER stress have been found in the spinal cords of sALS patients (Ilieva et al., 2007, Atkin et al., 2008, Ito et al., 2009). However, recent studies have shed more light on the role of protein degradation and ER stress in the pathology of ALS. In the first study, gene expression patterns from 3 different SOD1 mouse lines all showed an early increase in protein ubiquitination only in those motoneurons that are vulnerable to the disease. This is followed shortly by the UPR and signs of ER stress by P30 in SOD1G93A-high expressor mice (see Fig 1) (Saxena et al., 2009). In another study, cortical motoneurons from SOD1 mutant mice were compared to those from wild type mice that were fed a diet high in branched-chain amino acids (Carunchio et al., 2010). These branched chain amino acids are part of protein supplements that some athletes consume. Like mutant SOD1 neurons, cortical neurons from mice fed the highprotein diet were hyperexcitable compared with neurons from wild type mice on a normal diet. A return to normal levels of excitability after treatment with rapamycin was achieved for both the SOD1 and the amino- acid-supplement-treated cortical neurons (Carunchio et al., 2010). The protein kinase known as the mammalian target of rapamycin (mTOR) serves as an integration point for several cell signaling pathways. As its name suggests, mTOR is inhibited by rapamycin; it also inhibits protein degradation, and promotes increased cell size in some neurons (Lee et al., 2007). These results indicate that promoting autophagy with rapamycin can reduce abnormal excitability and could be beneficial for treatment of the disease (Carunchio et al., 2010). The third, most recent study described a mutation found in 5 different families, located in the gene encoding ubiquilin-2 as a novel genetic cause of fALS (Deng et al., 2011). The function of ubiquilin is to clear certain misfolded proteins during ERAD by shuttling ubiquitinated proteins from the ER to the proteasome, such that loss of ubiquilin leads to ER stress (Kim et al., 2008, Lim et al., 2009). The mutations in ubiquilin-2 found in ALS patients were also found to impair proteosome- mediated protein degradation *in vitro,* suggesting these mutations could be causing similar impairments in the

production in the large and metabolically-active SOD1 motoneurons.

**7. Protein degradation and endoplasmic reticulum stress** 

(Paschen, 2003).

the motoneurons that are vulnerable are the largest: the fast, fatiguable alpha motoneurons (Pun et al., 2006, Hegedus et al., 2007, Hegedus et al., 2008). Evidence for further increases size in SOD1 motoneurons is reviewed in the previous chapter by Elbasiouny et al. Perhaps the size of the motoneuron and deficits in transport go hand in hand to produce vulnerability.

Axon transport has been extensively studied and is likely to contribute to ALS and to several neurodegenerative diseases, reviewed by (De Vos et al., 2008). In ALS, both slow and fast axon transport appear to be altered (Zhang et al., 1997, Warita et al., 1999, Williamson and Cleveland, 1999, Kieran et al., 2005, De Vos et al., 2007, Bilsland et al., 2010). Excessive glutamate could cause these deficiencies: high levels of glutamate activate a family of mitogen-activated protein kinases that phosphorylate neurofilaments, thereby decreasing transport (Ackerley et al., 2000, Hiruma et al., 2003, Stevenson et al., 2009). This process can be induced by NMDA or AMPA, blocked by removal of extracellular Ca2+, or reduced by application of riluzole (Hiruma et al., 2003, Stevenson et al., 2009). The protein kinases JNKs, cdk/p35 and p38, which phosphorylate heavy and light chains of kinesin and medium and heavy neurofilament sidearms, may link glutamate neurotransmission and axon transport deficits (Kawasaki et al., 1997, Schwarzschild et al., 1997, Ackerley et al., 2000, Brownlees et al., 2000, Lee et al., 2000). Further suggesting this, p38 has been found to be activated in SOD1 mice and ALS patients (Raoul et al., 2002, Tortarolo et al., 2003, Ackerley et al., 2004). Axon transport deficiencies occur early, with reports of impaired axonal integrity and dieback from the neuromuscular junction occurring weeks in advance of onset of symptoms in SOD1 mice, and appearing in cultured embryonic neurons (Kennel et al., 1996, Zhang et al., 1997, Williamson and Cleveland, 1999, Frey et al., 2000, Fischer et al., 2004, Pun et al., 2006, De Vos et al., 2007, Hegedus et al., 2007, Hegedus et al., 2008, Bilsland et al., 2010). Strengthening these results, transgenic TDP-43 mice show significantly lower levels of expression of heavy and light neurofilaments, though axon transport itself has not yet been assessed (Swarup et al., 2011).
