Responsiveness of sphingosine phosphate lyase insufficiency syndrome to vitamin B6 cofactor supplementation

Abstract

Sphingosine-1-phosphate (S1P) lyase is a vitamin B6-dependent enzyme that degrades sphingosine-1-phosphate in the final step of sphingolipid metabolism. In 2017, a new inherited disorder was described caused by mutations in SGPL1, which encodes sphingosine phosphate lyase (SPL). This condition is referred to as SPL insufficiency syndrome (SPLIS) or alternatively as nephrotic syndrome type 14 (NPHS14). Patients with SPLIS exhibit lymphopenia, nephrosis, adrenal insufficiency, and/or neurological defects. No targeted therapy for SPLIS has been reported. Vitamin B6 supplementation has therapeutic activity in some genetic diseases involving B6-dependent enzymes, a finding ascribed largely to the vitamin's chaperone function. We investigated whether B6 supplementation might have activity in SPLIS patients. We retrospectively monitored responses of disease biomarkers in patients supplemented with B6 and measured SPL activity and sphingolipids in B6-treated patient-derived fibroblasts. In two patients, disease biomarkers responded to B6 supplementation. S1P abundance and activity levels increased and sphingolipids decreased in response to B6. One responsive patient is homozygous for an SPL R222Q variant present in almost 30% of SPLIS patients. Molecular modeling suggests the variant distorts the dimer interface which could be overcome by cofactor supplementation. We demonstrate the first potential targeted therapy for SPLIS and suggest that 30% of SPLIS patients might respond to cofactor supplementation.

Keywords: SGPL1; SPL insufficiency syndrome; pyridoxal 5′-phosphate; sphingolipidosis; sphingosine phosphate lyase; sphingosine-1-phosphate; vitamin B6.

Conflict of interest statement

CONFLICT OF INTEREST

Piming Zhao, Isaac Liu, Jeffrey Hodgin, Peter Benke, Jeremy Selva, Federico Torta, Markus Wenk, James Endrizzi, Olivia West, Weixing Ou, Emily Tang, Denise Goh, Stacey Tay, Hui-Kim Yap, Alwin Loh, Nicole Weaver, Bonnie Sullivan, Austin Larson, Megan Cooper, Khalid Alhasan, Abdullah Alangari, Suha Salim, Evren Gumus, Karin Chen, Martin Zenker, Friedhelm Hildebrandt and Julie Saba declare that they have no conflict of interest.

© 2020 SSIEM.

Figures

FIGURE 1

Immunological and biochemical responses to vitamin B6 treatment in patients 1 and 4. Repeated blood counts in patient 1 show (A) the absolute lymphocyte count (ALC) and (B) percent lymphocytes increased in response to pyridoxine treatment. Blood counts before treatment, N = 14. Blood counts after initiation of treatment, N = 17. *P-value for absolute lymphocyte count before and after treatment = .0025; P-value for percent lymphocytes before and after treatment = .0013. Repeated blood counts in patient 4 show (C) the absolute lymphocyte counts and (D) percent lymphocytes increased in response to PLP and pyridoxine treatments. Blood counts before treatment, N = 4. Blood counts after initiation of treatment, N = 13. *P-value for absolute lymphocyte counts = 7.2 × 10−5; P-value for percent lymphocytes = .0002. In patient 4, T-cell CD3, CD4, and CD8 subsets were evaluated over time. Absolute subsets (E) and percentage of each subset (F) increased after B6 was initiated. The P-value for percent CD3, CD4, and CD8 before and after treatment = .011, .017, and .007, respectively; P-value for absolute CD3, CD4, and CD8 levels before and after treatment = 1.5 × 10−4, 2 × 10−4, and 2 × 10−4, respectively. (G) Concentrations of S1P and dihydro-S1P measured in plasma of five controls matched by age, gender, and ethnicity to patient 4. Data for patient 4 represent four different time points collected before any treatment was started. For each group, mean and interquartile values are represented. (H) Concentration of patient 4 plasma S1P (d18:1), measured by targeted mass spectrometry before (blue) and after (black) starting treatment with pyridoxal 5′-phosphate. Time is indicated in months after the first plasma sample was taken

FIGURE 2

Sphingosine phosphate lyase (SPL) protein expression and enzyme activity in sphingosine phosphate lyase insufficiency syndrome (SPLIS) fibroblasts. Fibroblasts from SPLIS patients 1 to 4 and a healthy control were grown to confluence in B6 replete medium plus 10% FBS, harvested by cell scraping, pelleted, and whole-cell extracts were used for both immunoblotting and SPL activity assays. (A) SPL protein expression in primary fibroblasts from patients 1 to 4; N = 3 per group. For patients 1 to 3 fibroblasts vs control, P < .05; For patient 4 fibroblasts vs control, P < 0.01. Significance was calculated based on image quantification results shown in Figure S6. (B) SPL activity in primary fibroblasts from patient 1 to 4. N = 4 per group. For each of the four patient-derived fibroblasts vs control, P < .0003. These results are representative of three separate experiments, with similar results obtained using either primary or transformed fibroblasts

FIGURE 3

Sphingolipid profiles in sphingosine phosphate lyase insufficiency syndrome (SPLIS) fibroblasts. Primary and transformed skin fibroblasts from SPLIS patients 1 to 4 and a healthy control were grown to confluence in B6 replete medium plus 10% fetal bovine serum (FBS), harvested by cell scraping, pelleted, and whole-cell extracts were used for analysis of sphingoid bases and sphingoid base phosphates. Results are shown for primary fibroblasts: (A) Sphingosine-1-phosphate (S1P); (B) Dihydrosphingosine-1-phosphate (DihydroS1P); (C) Sphingosine; (D) Dihydrosphingosine (DHS). Each sample was run in triplicate (N = 3); *For S1P levels: patient 1 vs control, P = .03; patient 2 vs control, P < .0006; patient 3 vs control, P = .001; patient 4 vs control, P < 5.7 × 10−6. For DihydroS1P levels: patient 1 vs control, P = .07; patient 2 vs control, P = .04; patient 3 vs control, P = .02; patient 4 vs control, P < .0003. For all other results, * P < .05. These results are representative of three separate experiments, with similar results obtained using either primary or transformed fibroblasts

FIGURE 4

A sphingosine phosphate lyase insufficiency syndrome (SPLIS) patient-derived fibroblast line responds to exogenous PLP and B6 vitamers. Patient 4 fibroblasts were maintained for 1 week in B6-deficient medium and 10% fetal bovine serum (FBS), then switched to B6-deficient medium plus B6 vitamer. Cells were harvested by cell scraping, pelleted, and whole-cell extracts were used for immunoblotting, sphingolipid quantitation or sphingosine phosphate lyase (SPL) assays. (A) SPL protein levels determined by immunoblotting. N = 3 per condition. For pyridoxal (PL) vs vehicle, no significant difference was observed. For PLP vs vehicle control (Ctrl) and pyridoxamine (PM) vs control, P < .05. Significance was calculated based on image quantification results shown in Figure S7. (B) Sphingosine-1-phosphate (S1P) and (C) Sphingosine in cells treated with the indicated B6 vitamer; N = 4 per group. *For S1P: P < .0003 for all vitamers vs control. + For sphingosine: P ≤ .01 for all vitamers vs control. These results are representative of three separate experiments. (D) Patient 4 transformed fibroblasts were maintained for 1 week in B6-deficient medium and 10% FBS, then switched to B6-deficient medium-plus vehicle control (Ctrl) or a range of PN doses from 0.3 to 50 μM. Cells were harvested by cell scraping, pelleted, and whole-cell extracts were used for S1P quantitation after 1 week of incubation in the stated conditions. N = 3 per condition. *For 0.3, 1.0, and 10 μM vs vehicle, P < .005. + For 50 μM vs vehicle, P < .002. (E) Patient 4 transformed fibroblasts were maintained for 1 week in B6-deficient medium and 10% FBS, then switched to B6-deficient medium-plus vehicle control (Ctrl) or plus 50 μM PN. After 1 week of incubation in the stated conditions, whole-cell extracts were harvested and SPL enzyme activity measured. N = 3 per condition; *P = .035 for PN vs control

FIGURE 5

Response of sphingosine phosphate lyase insufficiency syndrome (SPLIS) patient fibroblasts to PN treatment correlates with in vivo response. Fibroblasts from patients 1 to 4 were maintained for 1 week in B6-deficient medium and 10% FBS, then switched to B6-deficient medium plus 50 μM PN. After 1 week of incubation in the stated conditions, whole-cell extracts were harvested and sphingolipid quantitation performed. (A) S1P; (B) Sphingosine; (C) dihydroS1P (DHS1P); (D) dihydrosphingosine (DHS). N = 3 per group; *P < .05 or greater significance. For patient 1, PN vs vehicle: S1P, P < .0004 and Sphingosine, P = .03. For patient 2, PN vs vehicle and patient 3, PN vs vehicle, no significant differences in S1P or sphingosine were observed. For patient 4, PN vs vehicle: S1P, P < .002 and sphingosine, P < .0007. These results are representative of three separate experiments

FIGURE 6

Potential mechanism of B6 interaction with SPL R222Q mutant protein. (A) PyMol-generated illustration of pyridoxal 5′-phosphate (PLP; black arrows) bound to wild type sphingosine phosphate lyase (SPL) monomers at the dimer interface. Monomers within the dimer are designated by green or yellow carbon atoms, while the PLP moiety contains cyan carbon atoms. Helix 243–252 (which would be distorted by substitution of glutamine for arginine) is shown in orange. R222 side chains are shown in deep blue. (B) Electrostatic PyMol surface representations for wild type and R222Q SPL homodimers illustrating electrostatic repulsion (red) at the dimer interface of the R222Q mutant enzyme. The white box represents the dimer interface and the arrows point to the locations of R222 substituted by Q222 in each mutant monomer