The SPTLC1 p.S331 mutation bridges sensory neuropathy and motor neuron disease and has implications for treatment

Abstract

Aims: SPTLC1-related disorder is a late onset sensory-autonomic neuropathy associated with perturbed sphingolipid homeostasis which can be improved by supplementation with the serine palmitoyl-CoA transferase (SPT) substrate, l-serine. Recently, a juvenile form of motor neuron disease has been linked to SPTLC1 variants. Variants affecting the p.S331 residue of SPTLC1 cause a distinct phenotype, whose pathogenic basis has not been established. This study aims to define the neuropathological and biochemical consequences of the SPTLC1 p.S331 variant, and test response to l-serine in this specific genotype.

Methods: We report clinical and neurophysiological characterisation of two unrelated children carrying distinct p.S331 SPTLC1 variants. The neuropathology was investigated by analysis of sural nerve and skin innervation. To clarify the biochemical consequences of the p.S331 variant, we performed sphingolipidomic profiling of serum and skin fibroblasts. We also tested the effect of l-serine supplementation in skin fibroblasts of patients with p.S331 mutations.

Results: In both patients, we recognised an early onset phenotype with prevalent progressive motor neuron disease. Neuropathology showed severe damage to the sensory and autonomic systems. Sphingolipidomic analysis showed the coexistence of neurotoxic deoxy-sphingolipids with an excess of canonical products of the SPT enzyme. l-serine supplementation in patient fibroblasts reduced production of toxic 1-deoxysphingolipids but further increased the overproduction of sphingolipids.

Conclusions: Our findings suggest that p.S331 SPTLC1 variants lead to an overlap phenotype combining features of sensory and motor neuropathies, thus proposing a continuum in the spectrum of SPTLC1-related disorders. l-serine supplementation in these patients may be detrimental.

Keywords: HSAN; S331; SPTLC1; l-serine; motor neuron; sphingolipids.

FIGURE 1

Pedigree and sequencing chromatogram of the SPTLC1 gene mutation. Family 1 (panel A) and family 2 (panel B). The proband is indicated by an arrow. Vertical arrows indicate the mutation site. Conservation analysis of the amino acid sequences between different species (panel C)

FIGURE 2

Summary of the clinical features from patients with p.S331 mutations in the SPTLC1 gene. (A–D) Patient 1 with SPTLC1 p.S331Y variant. (A) Tongue atrophy and fasciculation; (B, C) hands and foot deformities; (D) T1 axial MRI images of lower limbs muscles showing diffuse fatty replacement (T1 hyperintensity) of all leg muscles (lower panel) and most of the muscles in the thigh (upper panel), with partial preservation of quadriceps and adductors. (E, F) Patient two with SPTLC1 p.S331F variant. (E) Low stature and diffuse muscle hypotrophy with pes cavus; ambulation is still possible. (F) Muscle biopsy from the quadriceps (vastus lateralis) muscle. ATP staining at pH 4.3 shows predominance of type 1 dark fibres. A denervated fascicle is also spotted with atrophic angulated fibres. Bar = 20 μm

FIGURE 3

Study of skin innervation. (A–C) Digital confocal images from a distal site showing epidermal and dermal innervation in a young healthy control (A) and in the patients carrying p.S331 variants in the SPTLC1 gene (B, C). (A1–C1) Epidermal nerve fibres (ENF ‐ empty arrowheads in A1) are completely absent in patients' skin (B1, C1) compared with the control (A1). A severe reduction of nerves around sweat glands (B2, C2) and along erector pili muscles (B3, C3) is also evident in patients compared with the healthy subject (A2, A3). The image series 1 at lower magnification (10x) show the extent of epidermis and dermis denervation in both patients compared with the control; the image series 2 and 3 show at higher magnification (20×) the morphological abnormalities of sudomotor and pilomotor nerves in terms of cotton‐like aspect, more evident in patient two (full arrowhead in C2 and C3) compared with the control (A2, A3). In green: nerves marked with the pan‐neuronal protein gene product 9.5 (PGP9.5) antibody. In red: basement membrane and blood vessels marked with Collagen IV (Col IV) antibody. In blue: endothelia and epidermis marked with ULEX europaeus. Bar A1–C1 = 100 μm, A2–C2, A3–C3 = 50 μm

FIGURE 4

Neuropathological study of patient one's sural nerve biopsy. (A, C) Semithin sections of sural nerve biopsy (Toluidine blue; scale bars = 20 μm) from Patient 1 (A) display a complete loss of myelinated fibres compared with control subject (C). (B, D) Paraffin‐embedded longitudinal sections of sural nerve biopsy from Patient 1 (B) and a control subject (D) were stained with Mallory trichrome to show myelinated fibres (gold yellow), collagen fibrils (deep blue) and nuclei (red). Patient 1 displays a complete loss of myelinated fibres that are clearly identifiable in the control (D, black arrowheads) as well as nodes of Ranvier (D, black arrow). Instead, a massive deposition of collagen is present in Patient 1 compared with the control subject (B, white asterisks). Bar in B, D = 20 μm. (E–G) TEM on sural nerve biopsy of Patient 1 shows a remarkable increase of collagen fibrils (E, black asterisks); strikingly, we found only one myelinated fibre in all the nerve fascicles (E, black arrow). At higher magnification (F), this myelinated fibre displays clear signs of degeneration both in the Schwann cell cytoplasm (F, black arrow) and at axon level with detachment from the myelin membrane (F, black dotted arrow). (G) Unmyelinated fibres did not show signs of sufferance. Occasionally, we detected collagen pockets (G, black arrow). An increased number of mitochondria was also found (G, black asterisks). Bar in E–G = 1 μm

FIGURE 5

Sphingolipidomic analysis of serum samples of patients with SPTLC1 p.S331 variants. (A) HPLC‐MS analysis of serum samples with each row representing a different sphingolipid species. Row Z‐scores were generated and depicted in a heatmap form. LCB, long chain base; Ce r, ceramides; GlucCer, glucosylceramides; SM, sphingomyelins; deoxyCer, 1‐deoxyCeramides. (B) Serum levels of 1‐deoxy and 1‐deoxymethyl long chain bases after acid hydrolysis in control, classic HSAN1 (SPTLC1 p.C133Y/W) and SPTLC1 p.S331 samples. SA, sphinganine; SO, sphingosine. (C) Serum levels of C‐18 (d18:1 and d18:0) and C‐20 (d20:1 and d20:0) long chain bases after acid hydrolysis in control, classic HSAN1 (SPTLC1 p.C133Y/W) and SPTLC1 p.S331 samples

FIGURE 6

Sphingolipidomic analysis of patient derived fibroblasts with SPTLC1 p.S331 variants. (A) HPLC‐MS lipidomic analysis of fibroblasts harvested after 3,3‐D2 l‐serine treatment (3mM). Only the D2‐containing (+2) species were quantified to reflect the de novo rate of synthesis for each species. SA, sphinganine; SO, sphingosine; Cer, ceramide; GlucCer, glucosylceramides; SM, sphingomyelin. (B) HPLC‐MS lipidomic analysis of fibroblasts harvested after D4‐alanine (3 mM) supplementation. Only the D3‐containing (+3) deoxysphingolipid species were quantified to reflect the de novo rate of synthesis for each species. DHC, dihydroceramide. (C) The effect of increasing l‐serine supplementation on deoxy sphingolipids, total sphinganine and total ceramide levels in S331 fibroblasts. The height of columns represents the average of two technical replicates.

FIGURE 7

Genotype–phenotype correlation in SPTLC1 related disorders. S331 SPTLC1 phenotype shares HSAN and ALS features with clinical, electrophysiology, neuropathology and biochemical overlap, confirming SPLTC1 related disorders spectrum range from sensory autonomic neuropathy to pure motor neuron disease depending on site and effect of the different gene mutations.