Screening for GFAP rearrangements in a cohort of Alexander disease and undetermined leukoencephalopathy patients.

Abstract Alexander disease (AxD), a fatal degenerative leukoencephalopathy, is caused by de novo heterozygous missense mutations in the Glial Fibrillary Acidic Protein (GFAP) gene. The pathological hallmark of the disease is the presence of Rosenthal fibers, cytoplasmic inclusions in astrocytes, composed mainly of GFAP, αB-crystallin and HSP27. To date, several patients with a typical presentation of the disease or displaying characteristic Rosenthal fibers in brain material have been reported with no GFAP mutation. Recently, several studies have demonstrated a correlation between Rosenthal fiber formation and wild-type GFAP overexpression, despite the absence of mutations. We tested the hypothesis that a GFAP gene rearrangement could modulate AxD severity or promote GFAP overexpression and aggregation, resulting in leukoencephalopathy. A QMPSF assay was validated for 11 exonic fragments: 3 in control genes ( CFTR , DSCR1 , F9 ) and 8 corresponding to GFAP exons. A total of 97 patients suspected of AxD were analyzed: 28 with a GFAP mutation; 69 with clinical and magnetic resonance imaging criteria compatible with the disease. Neither duplications nor deletions of GFAP were found, suggesting that GFAP coding-region rearrangements may not be involved in AxD or Alexander-related leukoencephalopathies. In addition, 80 patients with undetermined leukodystrophies, and negative for PLP1 , GJA12 , Sox10 and MCT8 mutations and PLP1 and Lamin B1 rearrangements, were tested. These patients were also negative for GFAP rearrangements. Other hypotheses should be investigated for a molecular diagnosis in patients with undetermined leukoencephalopathy: mutations in GFAP isoforms, splicing sites or regulatory regions, or defaults in genes encoding molecular partners of GFAP.