The most common cause of both amyotrophic lateral sclerosis (ALS) and frontotemporal lobar degeneration (FTLD) is a G4C2-repeat expansion in C9orf72. However, the lower limit for pathological repeats has not been established and expansions with different sizes could have different pathological consequences. One of the implicated disease mechanisms is haploinsufficiency. Previously, we identified expansion-specific hypermethylation at the 5' CpG-island near the G4C2-repeat, but only in a fraction of carriers (up to 36 %). Here, we tested the hypothesis that the G4C2-repeat itself could be the main site of methylation. To evaluate (G4C2)n -methylation, we developed a novel assay, which was validated by an independent methylation-sensitive restriction enzyme assay. Notably, both assays are qualitative but not quantitative. Blood DNA was available for 270 unrelated individuals, including 71 expansion carriers. In addition, we investigated blood DNA from family members of 16 probands, and 38 DNA samples from multiple tissues of 10 expansion carriers. Finally, we tested DNA from different tissues of an ALS patient carrying a somatically unstable 90-repeat. We demonstrated that the G4C2-expansion is generally methylated in unrelated carriers of alleles >50 repeats (97 %), while small (<22 repeats) or intermediate (22-90 repeats) alleles were completely unmethylated. The presence of (G4C2)n -methylation does not separate the C9orf72-phenotypes (ALS vs. ALS/FTLD vs. FTLD), but has the potential to predict large vs. intermediate repeat length. Our results suggest that (G4C2)n -methylation might sometimes spread to the 5'-upstream region, but not vice versa. It is stable over time, since (G4C2)n -methylation was detected in carriers with a wide range of ages (24-74 years). It was identified in both blood and brain tissues for the same individual, implying its potential use as a biomarker. Furthermore, our findings may open up new perspectives for studying disease mechanisms, such as determining whether methylated and unmethylated repeats have the same ability to form a G-quadruplex configuration.

The neurofibromatosis 2 (NF2) tumor suppressor protein, merlin, functions as a negative growth regulator; however, the molecular mechanisms that underlie merlin regulation remain elusive. Recent studies have implicated merlin phosphorylation in regulating merlin subcellular localization and growth suppression. P21-activated kinase (PAK), a downstream target of Rac1/Cdc42, directly phosphorylates merlin at Serine 518. In this report, we show that PAK2 directly phosphorylates wild-type merlin, whereas merlin truncation mutants with impaired GST-aminoterminal domain (N-term or NTD)/GST-carboxy-terminal domain (C-term or CTD) intramolecular association exhibit impaired S518 phosphorylation. We directly demonstrate that PAK2 phosphorylation impairs merlin N-term/C-term binding in vitro and in vivo. Lastly, we show that PAK2 phosphorylation impairs the ability of merlin to bind to two interacting proteins, CD44 and hepatocyte growth factor-regulated tyrosine kinase substrate (HRS), both critical for merlin growth suppression. These observations suggest that merlin S518 phosphorylation directly modulates merlin intramolecular and intermolecular associations important for the ability of merlin to function as a tumor suppressor.

The neurofibromatosis 2 (NF2) tumor suppressor gene product, merlin, belongs to the ezrin-radixin-moesin (ERM) subgroup of the Protein 4.1 family, which links cell surface glycoproteins to the actin cytoskeleton. Previous studies have suggested that phosphorylation of merlin, similar to other ERM proteins, may regulate its function. To determine whether merlin phosphorylation has functional consequences for merlin suppression of cell growth and motility, we generated doxycycline-regulatable RT4 schwannoma cell lines that inducibly express full-length merlin with mutations at two potential phosphorylation sites (amino-acid residues S518 and T576). Whereas a mutation at S518 that mimics constitutive phosphorylation (S518D) abrogates the ability of merlin to suppress cell growth and motility, the S518A merlin mutant, which mimics nonphosphorylated merlin, functions equivalently to wild-type merlin. Similar mutations involving T576, the analogous phosphorylation site in ERM proteins important for regulating their function, had no effect. In contrast to other functionally inactive missense merlin mutants, the regulated overexpression of S518D merlin resulted in dramatic changes in cell shape and the elaboration of filopodial extensions. These results provide the first direct demonstration that the S518D merlin mutation, which mimics merlin phosphorylation, impairs not only merlin growth and motility suppression but also leads to an acquisition of a novel phenotype previously ascribed to ERM proteins.

No abstract