Homozygous mutation of STXBP5L explains an autosomal recessive infantile-onset neurodegenerative disorder

Abstract

We report siblings of consanguineous parents with an infantile-onset neurodegenerative disorder manifesting a predominant sensorimotor axonal neuropathy, optic atrophy and cognitive deficit. We used homozygosity mapping to identify an ∼12-Mbp interval identical by descent (IBD) between the affected individuals on chromosome 3q13.13-21.1 with an LOD score of 2.31. We combined family-based whole-exome and whole-genome sequencing of parents and affected siblings and, after filtering of likely non-pathogenic variants, identified a unique missense variant in syntaxin-binding protein 5-like (STXBP5L c.3127G>A, p.Val1043Ile [CCDS43137.1]) in the IBD interval. Considering other modes of inheritance, we also found compound heterozygous variants in FMNL3 (c.114G>C, p.Phe38Leu and c.1372T>G, p.Ile458Leu [CCDS44874.1]) located on chromosome 12. STXBP5L (or Tomosyn-2) is expressed in the central and peripheral nervous system and is known to inhibit neurotransmitter release through inhibition of the formation of the SNARE complexes between synaptic vesicles and the plasma membrane. FMNL3 is expressed more widely and is a formin family protein that is involved in the regulation of cell morphology and cytoskeletal organization. The STXBP5L p.Val1043Ile variant enhanced inhibition of exocytosis in comparison with wild-type (WT) STXBP5L. Furthermore, WT STXBP5L, but not variant STXBP5L, promoted axonal outgrowth in manipulated mouse primary hippocampal neurons. However, the FMNL3 p.Phe38Leu and p.Ile458Leu variants showed minimal effects in these cells. Collectively, our clinical, genetic and molecular data suggest that the IBD variant in STXBP5L is the likely cause of the disorder.

Figure 1.

(A) Pedigree with affected male and female siblings of consanguineous parents indicating likely autosomal recessive inheritance. *DNA collected for mapping and subsequent segregation of mutations. (B) Homozygosity mapping for chromosome 3 performed using data obtained from Illumina 660 Quad SNP chips. A single peak was identified using MERLIN between rs2673391 and rs1355563 using both parametric (black, LOD = 2.31) and non-parametric (green, LOD = 1.81) analyses. (C) Segregation of the mutation in STXBP5L and FMNL3 with disease was confirmed by Sanger sequencing in all available family members. Sequence traces for the parents and the affected sibs III:2 and III:6 are shown. (D) STXBP5L has 14 WD repeats (as indicated by the key); the Val1043Ile mutation occurs in the 14th repeat (indicated by the arrowhead and line), which is adjacent to the C-terminal R-SNARE domain (as indicated by the key). (E) FMNL3 has multiple domains (GBD, GTPase-binding domain; DID, diaphanous inhibitory domain; DD, dimerization domain; formin homology 1 (FH1), 2 (FH2) and 3 (FH3) domains; DAD, diaphanous autoregulatory domain). The Phe38Leu and Ile458Leu are located in the N-terminus and between DD and FH1 domains, respectively. Sequences related to (F) STXBP5L and (G) FMNL3 were identified by tblastn search, as well as HomoloGene and OrthoDB (6). Full-length protein sequences were then aligned using the CLUSTALW algorithm and sorted by distance; a section of the full-length alignment of each protein is shown with the mutant residue marked with the arrowhead.

Figure 2.

(A) Expression of STXBP5L, the closely related STXBP5 and FMNL3 compared with PPIA as measured by semi-quantitative RT-PCR, size-fractionated by agarose gel (3%) electrophoresis, stained with ethidium bromide and visualized under UV light. STXBP5L and STXBP5 expression was also assayed in brain, spinal cord and heart by real-time RT-qPCR and normalized to HPRT1 expression (right panel). Note that STXBP5L expression is significantly (Student's two-tailed t-test) higher than STXBP5 in the brain and spinal cord. (B) Effect of STXBP5L p.Val1043Ile mutation (STXBP5L-MT) compared with STXBP5L-WT and STXBP5 on the secretion of hGH in rat PC12 cells ectopically expressing comparable levels of these proteins (inset western blot; anti-Myc immunoreactive signals from three independent western blots quantified using ImageJ software and normalized to the housekeeping anti-β-tubulin signals were as follows: Lane 2, 1.00 ± 0.09; Lane 3, 1.00 ± 0.15 and Lane 4, 0.94 ± 0.05). STXBP5L-MT exhibits enhanced suppression of hGH secretion compared with STXBP5L-WT. For each sample, data were normalized to unstimulated (light bars) levels of hGH. Columns represent the mean of three independent experiments performed in triplicate; error bars show SDs. Statistical significance was measured between all groups using a Student's two-tailed t-test. All data were corrected for multiple comparisons using the false discovery rate step-up method. (C) Membrane capacitance (calcium-dependent exocytosis) is reduced in the STXBP5L-MT compared with STXBP5L-WT. The figure shows the change in membrane capacitance at 30, 60, 90 and 120 s after calcium infusion in PC12 cells ectopically expressing with WT, MT or GFP alone proteins. The WT versus MT were significantly different at P < 0.05 at the 60 and 90 s time points.

Figure 3.

(A) STXBP5L, but not variant STXBP5L, promotes axonal outgrowth. Primary hippocampal neurons expressing EGFP only or with either STXBP5L (STXBP5L-WT) or the mutant STXBP5L p.Val1043Ile (STXBP5L-MT) protein were stained for marker proteins of dendritic (MAP2, cyan) and axonal (TAU1, red) structures. Cells were counterstained with DAPI (blue). Representative images of transfected neurons are shown. (B) Mean primary axonal length of transfected neurons. (C) Mean number of neurite (i.e. axonal, dendritic and total) termini. *P < 0.01 Student's two-tailed t-test. Error bars represent ±SD. Experiment conducted in triplicate with at least 30 neurons scored per replicate (in total, Control n = 107, STXBP5L-WT n = 100 and STXBP5L-MT n = 102). (D) Mean primary axonal length and (E) Mean number of neurite (i.e. axonal, dendritic and total) termini of neurons overexpressing FMNL3-WT, p.Phe38Leu, p.Ile458Leu or p.Phe38Leu and p.Ile458Leu. No difference in axonal outgrowth or arborization among different treatments was observed. Error bars represent ±SD. Experiment conducted in triplicate with at least 35 neurons scored per replicate (in total, FMNL3-WT n = 118, p.Phe38Leu n = 116, p.Ile458Leu n = 112, p.Phe38Leu and p.Ile458Leu n = 111).