**5. Gene function**

A hexanucleotide repeat expansion in the *C9ORF72* gene has recently been identified as the cause of chromosome 9p-linked ALS-FTD. The mechanism as to how this expanded repeat causes ALS is unknown. No causal mutations in *UNC13a* have been identified in ALS patients to date*.* Close examination of the reported function(s) of the proteins encoded by the *C9ORF72* and *UNC13a* gene may help to design strategies for determining the functional role of these loci in ALS and/or FTD.

In this section the current knowledge of the function of these genes will be discussed in light of a possible contribution to ALS pathogenesis.

### **5.1 Chromosome 9 open reading frame 72**

The *C9ORF72* gene encodes a protein of 481 amino acids. Alternative splicing of this gene is thought to produce five isoforms of which three are protein coding. Isoform 1 contains the entire sequence and consists of 481 amino acids, while isoform 2 and 3 have an asparagine to lysine change at amino acid 222 which results in the truncation of amino acids 223 to 481. Thus far, no C9ORF72 protein has been detected and nothing is known about the function of C9ORF72.

The *C9ORF72* gene has been sequenced in four linkage studies in 39 patients from different families, but no mutations have been identified. No changes in splicing, small deletions or duplications were detected in patients from an ALS-FTD family (Boxer et al., 2010). The gene has been sequenced in 16 sALS patients and 16 controls but no variants specific for sALS were identified (Laaksovirta et al., 2010). Hexanucleotide repeat expansions were recently found to be the most common cause of fALS and familial FTD and were also identified in sALS patients (Dejesus-Hernandez et al., 2011; Renton et al., 2011). The functional consequence of these repeat expansions are however unknown.

consisting of 599 ALS patients and 4144 control individuals two SNPs on chromosome 9p21.2 were found to be associated with ALS (Shatunov et al., 2010). A joint analysis including 4132 ALS patients and 8425 controls from this UK cohort and from previously published data from the UK, USA, Netherlands, Ireland, Italy, France, Sweden, and Belgium also showed significant association to the locus on chromosome 9p21.2 (Shatunov et al., 2010). In addition, replication of one of the associated SNPs on chromosome 9p21.2 was found in a GWAS performed in FTD patients when analyzing the ALS-FTD patients only. A different SNP in this locus was significantly associated with FTD (Rollinson et al., 2011). A trend towards significant genome-wide association between chromosome 9p21.2 and FTD was found when analyzing 426 FTD patients with TDP-43 pathology without mutations in the progranulin gene and 2509 control individuals (Van Deerlin et al., 2009). A replication study in Chinese and Japanese sALS patients failed to find association to one of the previously associated SNPs on chromosome 9p21.2 but this might be due to a lack of power

In summary, linkage studies in ALS-FTD families and GWAS in sALS, ALS-FTD and FTD patients provide compelling evidence for a role of chromosome 9p21.2 in ALS and/or FTD. As mentioned, recently two studies identified a GGGGCC hexanucleotide repeat expansion in intron 1 of the *C9ORF72* gene as the cause of chromosome 9p-linked ALS-FTD (Dejesus-

A hexanucleotide repeat expansion in the *C9ORF72* gene has recently been identified as the cause of chromosome 9p-linked ALS-FTD. The mechanism as to how this expanded repeat causes ALS is unknown. No causal mutations in *UNC13a* have been identified in ALS patients to date*.* Close examination of the reported function(s) of the proteins encoded by the *C9ORF72* and *UNC13a* gene may help to design strategies for determining the functional

In this section the current knowledge of the function of these genes will be discussed in light

The *C9ORF72* gene encodes a protein of 481 amino acids. Alternative splicing of this gene is thought to produce five isoforms of which three are protein coding. Isoform 1 contains the entire sequence and consists of 481 amino acids, while isoform 2 and 3 have an asparagine to lysine change at amino acid 222 which results in the truncation of amino acids 223 to 481. Thus far, no C9ORF72 protein has been detected and nothing is known about the function of

The *C9ORF72* gene has been sequenced in four linkage studies in 39 patients from different families, but no mutations have been identified. No changes in splicing, small deletions or duplications were detected in patients from an ALS-FTD family (Boxer et al., 2010). The gene has been sequenced in 16 sALS patients and 16 controls but no variants specific for sALS were identified (Laaksovirta et al., 2010). Hexanucleotide repeat expansions were recently found to be the most common cause of fALS and familial FTD and were also identified in sALS patients (Dejesus-Hernandez et al., 2011; Renton et al., 2011). The

functional consequence of these repeat expansions are however unknown.

(Iida et al., 2011).

**5. Gene function** 

C9ORF72.

Hernandez et al., 2011; Renton et al., 2011).

role of these loci in ALS and/or FTD.

of a possible contribution to ALS pathogenesis.

**5.1 Chromosome 9 open reading frame 72** 

Further studies will be needed to characterize the C9ORF72 protein and to establish the consequences of the intronic repeat on ALS pathogenesis.

### **5.2 UNCoordinated 13 homolog A (UNC13a)**

UNC13a is a member of UNC13 family of presynaptic proteins. The protein consists of 1791 amino acids but several isoforms exist. It contains a zinc-finger like C1 domain that is homologous to a diacylglycerol and phorbol ester binding region of protein kinase C (PKC), three C2 domains that are similar to the calcium binding regulatory regions of PKC and synaptotagmin, a calmodulin binding domain and two Munc homology domains (Basu et al., 2005).

In mammals, the Munc13 family comprises four homologous members, Munc13-1, Munc13-2, Munc13-3, and Munc13-4. Deletion mutants of Munc13-1 in mice, the murine homologue of UNC13a, shows that the protein is needed for presynaptic vesicle maturation and fusion competence in glutamergic hippocampal neurons (Augustin et al., 1999). GABA-ergic neurons in the hippocampus show no spontaneous or evoked synaptic transmission in absence of both Munc13-1 and Munc13-2 (Varoqueaux et al., 2002). Neuromuscular junction (NMJ) axon terminals contain Munc13-1 and a splice variant of Munc13-2 (Varoqueaux et al., 2005). Mice deficient in Munc13 due to a double knockout of Munc13-1 and Munc13-2 form specialized neuromuscular endplates. However, the distribution, size and shape of these synapses are altered. Also, muscle morphology is abnormal and a larger number of motor neurons is present in the spinal cord in Munc13- 1/2 knockout mice, probably as a result of defective apoptosis. Furthermore, evoked synaptic transmission is impaired in these mutants but spontaneous transmission is unchanged (Varoqueaux et al., 2005). This indicates that vesicle priming in NMJs is partially independent of Munc13-1 or Munc13-2. However, despite the unchanged spontaneous transmission, muscle innervation is aberrant in Munc13-1/2 knockout mice (Varoqueaux et al., 2005).

As exemplified by the defects observed in Munc13-1 and Munc13-1/2 knockout mice, it is plausible that a disruption in UNC13a expression affects motor neurons and muscle innervation. The effect of UNC13a on glutamate exocytosis is also interesting since Riluzole, the only drug with a proven effect on ALS, is a glutamate release inhibitor. Therefore, *UNC13a* is an interesting candidate gene to be investigated further for a role in ALS pathogenesis.
