A PIK3R2 Mutation in Familial Temporal Lobe Epilepsy as a Possible Pathogenic Variant

Abstract

Temporal lobe epilepsy (TLE), the most common form of medically refractory focal epilepsy in adults, often requires surgery to alleviate seizures. By using next-generation sequencing, we identified a PIK3R2 mutation (NM_005027.4: c.265C > T; NP_005018.2: p.Arg89Cys) in a family with mesial temporal lobe epilepsy. PIK3R2 encodes p85β, the regulatory subunit of Class IA phosphoinositide 3-kinase (PI3K) and the mutation we identified in PIK3R2 seems to function unexpectedly as a possible pathogenic variant. The mutation is predicted to be potentially pathogenic by multiple bioinformatics tools. Through a functional assay, we verified that the mutation enhances the function of PI3K in induced pluripotent stem cells (iPSCs) derived from peripheral blood mononuclear cells (PBMCs) of the proband. Finally, pathological testing of the resected temporal lobe cortex showed that the expression of PIK3R2 was significantly higher in patients with refractory temporal lobe epilepsy than in those of non-epileptic diseases as a control group. It can be inferred that PIK3R2 might play an important role in the development of TLE.

Keywords: PIK3R2; familial temporal lobe epilepsy; genetic epilepsy; iPSCs; temporal lobe epilepsy.

FIGURE 1

Brain MRI images of the proband. (A–F) Brain MRI images of different levels of the proband show hippocampal atrophy. (A) FLAIR in the coronal plane; (B,E) T2WI in the sagittal plane; (C,F) T2WI in the axial plane; (D) T2WI in the coronal plane.

FIGURE 2

The family tree of the proband and the mutant site of PIK3R2 identified in the patients. (A) The family tree of the proband shows that the mother is a suspected epilepsy patient, the proband and her second sister are TLE patients, while her father and first sister are not epilepsy patients. (B,C) The PIK3R2 mutation (c.265 C > T; p.Arg89Cys) was detected in the proband, her mother, and her second sister, showing a heterozygous missense mutation. A variant c. 264 T > C was also detected, however, it was a synonymous variant and found to be an SNP site, which does not affect the function of the protein. (D) The influence of the mutation c.265 C > T (p.Arg89Cys) on the function of protein was predicted to be “probably damaging,” “damaging,” and “disease causing” by PolyPhen-2, SIFT, and Mutation Taster, separately; and the CADD score of this variant is also shown.

FIGURE 3

Structure modeling of the complete protein of PIK3R2 and the local structure of the mutation site in wild-type and mutant protein. (A) The complete structure of the p85β subunit: SH3 domain (pink), Rho-Gap domain (orange), SH2 1 domain (white), SH2 2 domain (blue), and other regions (green). (B) The position information of four major domains on the UniProt Database. It shows that the mutation site c.265 C > T (p.Arg89Cys) locates near the SH3 domain. (C,D) The local structure of the mutation site in wild-type (C) and mutant (D) protein. It shows that the formation of two hydrogen bonds inside Arg-89, Pro-88, and Asn-684 makes the local structure stable enough, while this stability is broken by Cys-89 replacing Arg-89.

FIGURE 4

The characters of mutations in the PIK3R2 gene recorded in gnomAD database. (A) The allele frequency (AF) distribution of PIK3R2 mutations, x-axis means the AF values, and we have converted the original value to –log10, and y-axis means the number of the mutations. It shows PIK3R2 is highly conserved as most mutations detected in the gene are rare mutations (MAF < 0.001). (B) The annotation information of PIK3R2 mutations, there are 13 types of annotated mutations in total, and the mutations with the highest frequency are the intron variants (40.67%).

FIGURE 5

Protein–protein interaction (PPI) network of PIK3R2 and TLE-related genes. It shows that 32 TLE-related genes interact with PIK3R2, based on the information on STRING Database. Two genes in blue color are AKT1 and MTOR, which are components of the PI3K/Akt/mTOR pathway, showing the important role of this pathway in TLE. The area of the circles is proportional to the combined score between PIK3R2 and TLE-related genes.

FIGURE 6

PIP3 expression levels in iPSCs derived from the proband and a healthy control. (A,B) Immunofluorescence staining of PIP3 in iPSCs derived from the proband (A) and the healthy control (B). Images were taken with the same exposure time. Scale bar corresponds to 20 μm. (C) Elevated PIP3 expression is evident in iPSCs derived from the proband. The average PIP3 level among the cells is based on the mean fluorescence intensity, which is calculated with ImageJ software. A two-tailed t-test was used to compare the mean fluorescence intensity. ** indicates a significant statistical difference (p < 0.01). Error bar indicates SEM.

FIGURE 7

p85β expression levels in the temporal lobe cortex of patients with TLE and controls. (A) The expression of p85β is negative in temporal lobe cortex tissues of controls. (B) The expression of p85β is significantly positive in the temporal lobe cortex tissues of TLE patients. Scale bar corresponds to 40 μm. (C) The statistical results of the percentage of the positive area of p85β in temporal lobe cortex of patients with TLE and controls. Elevated expression of p85β is detected in TLE patients. Mann–Whitney test was used to compare the percentage of the positive area for the non-normal distributions of the data. *** indicates a significant statistical difference (p < 0.001). Error bar indicates interquartile range.