Abnormal fatty acid metabolism is a core component of spinal muscular atrophy

Abstract

Objective: Spinal muscular atrophy (SMA) is an inherited neuromuscular disorder leading to paralysis and subsequent death in young children. Initially considered a motor neuron disease, extra-neuronal involvement is increasingly recognized. The primary goal of this study was to investigate alterations in lipid metabolism in SMA patients and mouse models of the disease.

Methods: We analyzed clinical data collected from a large cohort of pediatric SMA type I-III patients as well as SMA type I liver necropsy data. In parallel, we performed histology, lipid analysis, and transcript profiling in mouse models of SMA.

Results: We identify an increased susceptibility to developing dyslipidemia in a cohort of 72 SMA patients and liver steatosis in pathological samples. Similarly, fatty acid metabolic abnormalities were present in all SMA mouse models studied. Specifically, Smn^2B/- mice displayed elevated hepatic triglycerides and dyslipidemia, resembling non-alcoholic fatty liver disease (NAFLD). Interestingly, this phenotype appeared prior to denervation.

Interpretation: This work highlights metabolic abnormalities as an important feature of SMA, suggesting implementation of nutritional and screening guidelines in patients, as such defects are likely to increase metabolic distress and cardiovascular risk. This study emphasizes the need for a systemic therapeutic approach to ensure maximal benefits for all SMA patients throughout their life.

Figure 1

Smn^2B/‐ mice have fat accumulation in the liver. Gross morphology (0.75X) and histology (H&E – 40X, Oil Red O – 400X) of Smn^2B/‐ livers showing fatty inclusions at P17–19 (A and B, E–G) but not P4 (C and D). Lipid profiling identified elevation of triglycerides at P19 in Smn^2B/‐ livers (H), with altered chain length (I–J). Other lipid classes, such as phospholipid, free fatty acids, diglycerides, cholesterol esters, unesterified cholesterol, and total cholesterol, were also misregulated in P19 Smn^2B/‐ livers (K–P). P4 lipid levels were unchanged from control (H, K–P). Scale bar: (A–D) 50 µm, (E and F) 10 µm. (N value for each experiment is as follows: N = 5–6 for A–D, 3–5 for E–G, 4 for H–P, unpaired two‐sided student's t‐test, *P ≤ 0.05, **P ≤ 0.01, and ***P ≤ 0.001).

Figure 2

Hepatic triglyceride misregulation is a common feature in different SMA models at symptomatic. Quantification of hepatic triglycerides showed a fivefold reduction in P5 Smn^‐/‐; SMN2 mice (A), a twofold reduction in P9 Taiwanese mice (B), and a threefold increase in P10 SmnΔ7 mice (C) in comparison to control littermates. Analysis of hepatic triglyceride levels for each SMA mouse model in A–C involved a comparison to their own control (N value for each experiment is as follows: N = 4–6 for A and B, 9–10 for C, 4–9 for D, unpaired two‐sided student's t‐test, *P ≤ 0.05 and **P ≤ 0.01).

Figure 3

Commonalities identified in expression of fatty acid metabolism genes between Taiwanese and Smn^2B/‐ mice. Volcano plot presentation of all changes (1.5X, P < 0.05) in a focused fatty acid metabolism PCR array in Smn^2B/‐ mice (A) and Taiwanese mice (C) identify general downregulation. Changes more than two‐fold are represented for Smn^2B/‐ (B) and Taiwanese (D). Analysis of commonalities between Smn^2B/‐ and Taiwanese are represented by Venn diagrams, which identify nine genes with similar changes (E), listed in (F). (N = 4, for Smn^2B/‐ mice, and N = 3 for Taiwanese mice).

Figure 4

Fat accumulation is first observed between P9 and P11 in Smn^2B/‐ mice and denervation is not sufficient to trigger hepatic steatosis. (A and B) Triglycerides and cholesterol esters quantification in livers from Smn^2B/‐ mice at different ages. (C–E) Oil Red O staining (400X) additionally showed increased fat at P9. (F–H) H&E staining (40X) of livers of 20‐week‐old SOD1^G93A mutant mice, a model of ALS, did not show hepatic fat accumulation in comparison to livers from Smn^2B/‐ mice, even though denervation is well‐established at this time point. (I) Triglycerides and cholesteryl esters quantification showed no difference between mutant SOD1^G93A and WT controls. Scale bar represents 50 µm in C and D (10 µm in the inset), and in F–H. (N value for each experiment is as follows: N = 4–6 for A and B, 3 for C–E, and 3–5 for I, unpaired two‐sided student's t‐test, *P ≤ 0.05, **P ≤ 0.01, ***P ≤ 0.001).

Figure 5

Smn^2B/‐ mice display dyslipidemia and abnormal fatty acid metabolism in skeletal muscle, but not in spinal cord. (A–D) Significant upregulation of total cholesterol, VLDL and LDL in the plasma of Smn^2B/‐ animals, while HDL levels were significantly lower. (E) Parameters for cardiovascular risks such as TC/HDL were significantly increased for Smn^2B/‐ mice. (F) Glucose trend lower early and plummet later in life in Smn^2B/‐ mice. (G,H) Every lipid class in the P19 Smn^2B/‐ skeletal muscle or spinal cord were at similar levels to WT. (I) Many genes involved in fatty acid metabolism were altered in P19 Smn^2B/‐ skeletal muscle in a focused fatty acid PCR array. (N value for each experiment is as follows: N = 10 for A–F, 5 for G, 4 for H–I, unpaired two‐sided student's t‐test, **P ≤ 0.01, ***P ≤ 0.001 and ****P ≤ 0.0001).