PMP2 regulates myelin thickening and ATP production during remyelination

Abstract

It is well established that axonal Neuregulin 1 type 3 (NRG1t3) regulates developmental myelin formation as well as EGR2-dependent gene activation and lipid synthesis. However, in peripheral neuropathy disease context, elevated axonal NRG1t3 improves remyelination and myelin sheath thickness without increasing Egr2 expression or activity, and without affecting the transcriptional activity of canonical myelination genes. Surprisingly, Pmp2, encoding for a myelin fatty acid binding protein, is the only gene whose expression increases in Schwann cells following overexpression of axonal NRG1t3. Here, we demonstrate PMP2 expression is directly regulated by NRG1t3 active form, following proteolytic cleavage. Then, using a transgenic mouse model overexpressing axonal NRG1t3 (NRG1t3OE) and knocked out for PMP2, we demonstrate that PMP2 is required for NRG1t3-mediated remyelination. We demonstrate that the sustained expression of Pmp2 in NRG1t3OE mice enhances the fatty acid uptake in sciatic nerve fibers and the mitochondrial ATP production in Schwann cells. In sum, our findings demonstrate that PMP2 is a direct downstream mediator of NRG1t3 and that the modulation of PMP2 downstream NRG1t3 activation has distinct effects on Schwann cell function during developmental myelination and remyelination.

Keywords: FABP8; NRG1t3; PMP2; Schwann cell; myelin.

FIGURE 1

PMP2 is regulated by NRG1t3. (A) RTqPCR analysis of MBP and PMP2 in control and NRG1t3OE sciatic nerves at 7 days of age. (B–D) Western blot analysis of PMP2 and MPZ in control and NRG1t3OE sciatic nerves at 7 days of age (B), in control, Bace1^−/−, NRG1t3OE, Bace1^−/−; NRG1t3OE, NRG1t3ΔC (Constitutively active nNRG1 type III) and Bace1^−/−; NRG1t3ΔC at 60 days of age (C,D). Tubulin (TUB) was used for the endogenous control for the Western blots. Error bars represent s.e.m., n = 3–11 mice, and each dot on the graph represents a different n. One-way ANOVA with Bonferroni correction (A,C,D) and two-tailed unpaired Student’s t test. (B). *p value <.05, ****p value <.0001. t3OE (NRG1t3OE).

FIGURE 2

PMP2 is not necessary for NRG1t3-mediated hyper-myelination. (A–E) Semithin analysis of control, Pmp2^−/−, NRG1t3OE, and Pmp2^−/−; NRG1t3OE sciatic nerves at 30 days of age. The thickness of myelin (B,C), the distribution of diameter of myelinated axons (D), and the density of myelinated fibers (E) were measured. (F,G) Measurements of compound muscle action potential (F), and nerve conduction velocity (G) of control, Pmp2^−/−, NRG1t3OE, and Pmp2^−/−; NRG1t3OE mice at 60 days of age. Error bars represent s.e.m, n = 3–11 mice, each point on the B, C–F, G graphs represents a different n. Each dot on the C graph represents a different myelinated fiber. One-way ANOVA with Bonferroni correction. ****p value <.0001.

FIGURE 3

PMP2 is required for NRG1t3-mediated hyper-remyelination. (A–E) Semithin analysis of control, Pmp2^−/−, NRG1t3OE, and Pmp2^−/−; NRG1t3OE crushed sciatic nerves at 20 dpi. The thickness of myelin (B,C), the distribution of diameter of myelinated axons (D), and the density of myelinated fibers (E) were measured. Error bars represent s.e.m, n = 3–6 mice, each point on the B, D and E graphs represents a different n, and each point on the C graph represents a different myelinated fiber. Scale bars, 10 μm. All images were acquired at the same magnification. One-way ANOVA with Bonferroni correction. (F) Western blot analysis for c-JUN and GAP43 on control, Pmp2^−/−, NRG1t3OE, and Pmp2^−/−; NRG1t3OE sciatic nerves in crushed nerves at 20 dpi. GAPDH was used as a protein loading control. (G) Western blot and densitometry analysis for PMP2, P0, and MBP on control, Pmp2^−/−, NRG1t3OE, and Pmp2^−/−; NRG1t3OE sciatic nerves in crushed nerves at 20 dpi. GAPDH was used as a protein loading control. (H) Western blot for P-AKT Ser473, total AKT, P-ERK 1/2 and total ERK 1/2 and densitometry analyses of P-ERK/tot ERK and P-AKT/tot AKT ratio in control, Pmp2^−/−, NRG1t3OE, and Pmp2^−/−; NRG1t3OE crushed sciatic nerves at 20 dpi. GAPDH was used as a protein loading control. *p value <.05; **p value <.01.

FIGURE 4

PMP2 is required to promote an increase in fatty acid uptake in sciatic nerve of mice overexpressing NRG1t3. Teased fibers from control, Pmp2^−/−, NRG1t3OE, and Pmp2^−/−; NRG1t3OE sciatic nerves at 20 days of age were incubated for 1 h with C16-labeled BODIPY (10 μM). Intensity of Bodipy was measured in the perinuclear region and in the internodal region. Error bars represent s.d., n = 12–25 teased fibers, and each point on the graph represents a different n. One-way ANOVA with Bonferroni correction. ****p value <.0001. Scale bar 20 μm.

FIGURE 5

PMP2 regulates NRG1t3-mediated ATP production and basal respiration during remyelination. (A) Schematic representation of the Mitochondrial stress test. (B) Oxygen consumption rate in control, Pmp2^−/−, NRG1t3OE, and Pmp2^−/−; NRG1t3OE in crushed sciatic nerves at 7 dpi. Recordings were taken every 393 s. Oligomycin was injected at 15 min, FCCP at 81 min, and Antimycin A/Rotenone at 107 min. The last recording was taken at 126 min. (C–E) basal respiration (C), mitochondrial ATP production (D), and total ATP production (E) were extrapolated from the OCR and ECAR measurements in control, Pmp2^−/−, NRG1t3OE, and Pmp2^−/−; NRG1t3OE in crushed sciatic nerves at 7 dpi. Error bars represent s.e.m, n = 13–19 nerves, and each point on the graph represents a different n. One-way ANOVA with Bonferroni correction. *p value <.05; **p value <.01; ***p value <.001; ****p value <.0001.