Loss of Slc35a2 alters development of the mouse cerebral cortex

Abstract

Brain somatic variants in SLC35A2 are associated with clinically drug-resistant epilepsy and developmental brain malformations, including mild malformation of cortical development with oligodendroglial hyperplasia in epilepsy (MOGHE). SLC35A2 encodes a uridine diphosphate galactose translocator that is essential for protein glycosylation; however, the neurodevelopmental mechanisms by which SLC35A2 disruption leads to clinical and histopathological features remain unspecified. We hypothesized that focal knockout (KO) or knockdown (KD) of Slc35a2 in the developing mouse cortex would disrupt cerebral cortical development through altered neuronal migration and cause changes in network excitability. We used in utero electroporation (IUE) to introduce CRISPR/Cas9 and targeted guide RNAs or short-hairpin RNAs to achieve Slc35a2 KO or KD, respectively, during early corticogenesis. Following Slc35a2 KO or KD, we observed disrupted radial migration of transfected neurons evidenced by heterotopic cells located in lower cortical layers and in the sub-cortical white matter. Slc35a2 KO in neurons did not induce changes in oligodendrocyte number, suggesting that the oligodendroglial hyperplasia observed in MOGHE originates from distinct cell autonomous effects. Spontaneous seizures were not observed, but intracranial EEG recordings after focal KO showed a reduced seizure threshold following pentylenetetrazol injection. These results demonstrate that Slc35a2 KO or KD in vivo disrupts corticogenesis through altered neuronal migration.

Keywords: CRISPR; cerebral cortex; cortical dysplasia; glycosylation; short-hairpin RNA.

Fig. 1.. SLC35A2 is highly expressed in the mouse cortex during development.

A) Western Blot of SLC35A2 expression relative to GAPDH depicting expression in mouse cortex across development. Both SLC35A2 isoforms were quantified together. E = embryonic day, P = postnatal day. Mean ± SEM. * P<0.05, ** P<0.01, *** P<0.001.

Fig. 2.. Slc35a2 knockout results in altered cortical lamination.

A) gRNAs targeting exons 2 & 3 spliced into pX458v2-R-spCas9 plasmid containing an mCherry fluorescent reporter. B) Neuro2A cells were transfected with double gRNA knockout (KO) plasmid and FACS sorted for mCherry fluorescence to create the Slc35a2 KO cell line. C) Slc35a2 KO cells have minimal Slc35a2 mRNA expression relative to GAPDH when compared to wild-type (WT) cells. D) KO cell lysates show no SLC35A2 expression compared to control lysates (scrambled gRNA sequence and WT). E) Slc35a2 KO mice at P0 show EGFP fluorescence indicating expression of KO plasmid in focal cortical area. F) Slc35a2 KO mice underwent transcardial perfusion and brain extraction at P4. G) EGFP+ control brain versus H) KO brain at P4 showing the distribution of positively transfected neurons (green). I) Control brain versus J) KO brain showing EGFP-positive cells in the context of their laminar destination with co-staining for cortical markers SATB2 (red, Layer II/III marker) and Ctip2 (magenta, Layer V marker). K) Proportion of malpositioned neurons (number of neurons not located in Layer II/III divided by total positively transfected neuron count) in Slc35a2 KO mice vs EGFP control transfected mice. L-M) GFP and Olig2 immunoreactivity in the cortex and white matter of a KO brain. (N) No significant difference in Olig2+ cell numbers were seen across groups. * P<0.05.

Fig. 3.. Slc35a2 knockdown is sufficient to disrupt cortical lamination.

A) shRNA sequences targeting Slc35a2 or a scrambled control sequence were expressed in a U6-shRNA plasmid with EGFP as a reporter. Mouse NIH/3T3 cells were transfected with plasmid DNA to assess KD efficiency. Slc35a2 mRNA expression was significantly reduced with shRNA treatment compared to scrambled control. B) Plasmid DNA was delivered to the lateral ventricle of E15.5 mouse embryos by IUE. C) Distribution of EGFP-positive cells in the mouse cortex at P1. D) KD resulted in significantly more neurons located in the deep cortical layers and subcortical white matter. Scale bars = 100 μm. Mean ± SEM. ** P<0.01, ** P<0.001.

Fig. 4.. Slc35a2 knockout results in a lower seizure threshold.

A) Mice (Slc35a2 KO, EGFP control, and WT littermates) were implanted with EEG electrodes. B,C) Video-EEG was recorded for 72 hours of baseline activity followed by PTZ threshold testing (55 mg/kg PTZ i.p). D) Slc35a2 KO mice had a reduced seizure threshold when compared to both control groups at age P120 and E) at P60. *P<0.05.