Hydrocephalus due to multiple ependymal malformations is caused by mutations in the MPDZ gene

Abstract

Congenital hydrocephalus is considered as either acquired due to haemorrhage, infection or neoplasia or as of developmental nature and is divided into two subgroups, communicating and obstructive. Congenital hydrocephalus is either syndromic or non-syndromic, and in the latter no cause is found in more than half of the patients. In patients with isolated hydrocephalus, L1CAM mutations represent the most common aetiology. More recently, a founder mutation has also been reported in the MPDZ gene in foetuses presenting massive hydrocephalus, but the neuropathology remains unknown. We describe here three novel homozygous null mutations in the MPDZ gene in foetuses whose post-mortem examination has revealed a homogeneous phenotype characterized by multiple ependymal malformations along the aqueduct of Sylvius, the third and fourth ventricles as well as the central canal of the medulla, consisting in multifocal rosettes with immature cell accumulation in the vicinity of ependymal lining early detached from the ventricular zone. MPDZ also named MUPP1 is an essential component of tight junctions which are expressed from early brain development in the choroid plexuses and ependyma. Alterations in the formation of tight junctions within the ependyma very likely account for the lesions observed and highlight for the first time that primary multifocal ependymal malformations of the ventricular system is genetically determined in humans. Therefore, MPDZ sequencing should be performed when neuropathological examination reveals multifocal ependymal rosette formation within the aqueduct of Sylvius, of the third and fourth ventricles and of the central canal of the medulla.

Keywords: Autosomal recessive inheritance; Foetal hydrocephalus; MPDZ pathogenic variants; Multifocal malformation of the ependyma; Neuropathology.

Fig. 1

Identification of three homozygous mutations in the MPDZ gene using targeted NGS. Pedigrees of the three families with individuals affected by congenital hydrocephalus represented in grey, blue arrows indicate the affected foetuses which underwent NGS screening. An asterisk depicts individuals who underwent Sanger sequencing for segregation analysis. Figures under the foetuses identify the foetuses included in the study (a) Sanger sequencing electropherograms obtained in probands, affected siblings and parents (b)

Fig. 2

US coronal section passing through the diencephalon displaying absent third ventricle (white arrow) with major dilatation of the lateral ventricles and rupture of the septum in foetus 1 (a) and with on MRI considerable thinning of the corpus callosum in foetus 2 (black arrow), small vermis with enlarged cisterna magna (white arrow) (b). All foetuses presented characteristic dysmorphic features associating prominent forehead, small nose with large nasal bridge and bulbous tip, small mouth and midface hypoplasia (Foetus 4) (c) short prominent philtrum and micro-retrognathism (d.) On macroscopic section (Foetus 1), the corpus callosum was extremely thinned with undiscernible third ventricle (arrow) (e) and bilateral frontal brain parenchyma loss observed in foetus 2 due to severe intraventricular hyperpressure (arrow) (f) In the mesencephalon, the aqueduct was also undiscernible (arrow) (g) with in foetus 1 severe deformation of the fourth ventricle (h)

Fig. 3

Main histological lesions consisted in hypoplasia of the SCO (thick arrow) and colliculi, small aqueduct with rosettes in its inferior part (thin arrow) (a) compared with an age-matched control whose SCO has a normal size (arrow) (b) In foetus 4, the patent cerebral aqueduct possessed multiple indentations with rosettes in its dorsal and ventral parts [OM x 100] (c) At higher magnification the rosettes were most often lined by ciliated ependyma, and surrounded by small round-shaped cells (arrows) (d) and in foetus 4 atresia of the ependymal canal was observed at the level of the decussation of the pyramids (arrow) [OM x 20] (e) The immature cells were either dispersed or clustered (arrow) (f) and similar lesions observed around third ventricle (arrows) at the level of the interpeduncular fossa, between the red nuclei (asterisks) (g) as well as close to the 4^th ventricle ependyma of the fourth ventricle (arrow) which was abnormal in shape (h)

Fig. 4

Immunohistochemical findings using GFAP antibody showed only scattered immunoreactive ependymal cells within the rosettes (thin arrow) (a) along with reactive gliosis (thin arrow) but clustered or isolated cells were negative (thick arrow) (b) rosettes and cells were strongly S100B immunoreactive (c) as well as vimentin positive in the ependyma of the central canal of the medulla (arrow) (d) and surrounding cells were also immunolabeled using nestin and SOX2 antibodies respectively (e, f)

Fig. 5

Confocal analyses performed on ependyma sections double stained with anti- MPDZ (green) and EMA (red) showed a lack of MPDZ positivity in foetus 1 with EMA immunoreactivity at the apex of ependymal lining (a, b and c) contrary to the control where multiple MPDZ dots (green) were strongly expressed at the apical and lateral sides of the ependyma cell membrane and in the underlying cytoplasm (d, e and f) Immature cells surrounding the aqueduct and rosettes were only immunoreactive for nestin (arrow) (g) compared to the control case in which the dentate gyrus (used as a positive control) contained multiple stem cells co-expressing nestin (green cytoplasmic) and PAX6 (red nuclear) (arrow) (h)