Cobblestone lissencephaly represents a peculiar brain malformation with characteristic radiological anomalies, defined as cortical dysplasia combined with dysmyelination, dysplastic cerebellum with cysts and brainstem hypoplasia. Cortical dysplasia results from neuroglial overmigration into the arachnoid space, forming an extracortical layer, responsible for agyria and/or 'cobblestone' brain surface and ventricular enlargement. The underlying mechanism is a disruption of the glia limitans, the outermost layer of the brain. Cobblestone lissencephaly is pathognomonic of a continuum of autosomal recessive diseases with cerebral, ocular and muscular deficits, Walker-Warburg syndrome, muscle-eye-brain and Fukuyama muscular dystrophy. Mutations in POMT1, POMT2, POMGNT1, LARGE, FKTN and FKRP genes attributed these diseases to α-dystroglycanopathies. However, studies have not been able to identify causal mutations in the majority of patients and to establish a clear phenotype/genotype correlation. Therefore, we decided to perform a detailed neuropathological survey and molecular screenings in 65 foetal cases selected on the basis of histopathological criteria. After sequencing the six genes of α-dystroglycanopathies, a causal mutation was observed in 66% of cases. On the basis of a ratio of severity, three subtypes clearly emerged. The most severe, which we called cobblestone lissencephaly A, was linked to mutations in POMT1 (34%), POMT2 (8%) and FKRP (1.5%). The least severe, cobblestone lissencephaly C, was linked to POMGNT1 mutations (18%). An intermediary type, cobblestone lissencephaly B, was linked to LARGE mutations (4.5%) identified for the first time in foetuses. We conclude that cobblestone lissencephaly encompasses three distinct subtypes of cortical malformations with different degrees of neuroglial ectopia into the arachnoid space and cortical plate disorganization regardless of gestational age. In the cerebellum, histopathological changes support the novel hypothesis that abnormal lamination arises from a deficiency in granule cells. Our studies demonstrate the positive impact of histoneuropathology on the identification of α-dystroglycanopathies found in 66% of cases, while with neuroimaging criteria and biological values, mutations are found in 32-50% of patients. Interestingly, our morphological classification was central in the orientation of genetic screening of POMT1, POMT2, POMGNT1, LARGE and FKRP. Despite intensive research, one-third of our cases remained unexplained; suggesting that other genes and/or pathways may be involved. This material offers a rich resource for studies on the affected neurodevelopmental processes of cobblestone lissencephaly and on the identification of other responsible gene(s)/pathway(s).

Type II lissencephaly (type II LIS) is a group of autosomal recessive congenital muscular dystrophies (CMD) associated with defects in alpha-DG O-glycosylation, which comprises Walker-Warburg syndrome, Fukuyama cerebral and muscular dystrophy, or muscle-eye-brain disease. The most severe forms of these diseases often have a fetal presentation and lead to a pregnancy termination. We report here the first molecular study on fetal type II LIS in a series of 47 fetuses from 41 unrelated families. Sequencing of the different genes known to be involved in alpha-DG O-glycosylation allowed the molecular diagnosis in 22 families: involvement of POMT1 was demonstrated in 32% of cases, whereas POMGNT1 and POMT2 were incriminated in 15% and in 7% of cases, respectively. We found 30 different mutations in these three genes, 25 were described herein for the first time, 15 in POMT1, and five in POMT2 and POMGNT1. Despite sequencing of FKRP, FCMD, and LARGE, no definitive molecular diagnosis could be made for the other half of our cases. Preliminary results concerning genotype-phenotype correlations show that the choice of the first gene sequenced should depend on the clinical severity of the type II LIS; POMT1 and POMT2 for severest clinical picture and POMGNT1 for milder disease. The other genes, FKRP, FCMD, and LARGE, seem not to be implicated in the fetal form of CMD.