BetaIV spectrins are essential for membrane stability and the molecular organization of nodes of Ranvier

Abstract

High densities of sodium channels at nodes of Ranvier permit action potential conduction and depend on betaIV spectrins, a family of scaffolding proteins linked to the cortical actin cytoskeleton. To investigate the molecular organization of nodes, we analyzed qv(3J)"quivering" mice, whose betaIV spectrins have a truncated proline-rich "specific" domain (SD) and lack the pleckstrin homology (PH) domain. Central nodes of qv(3J) mice, which lack betaIV spectrins, are significantly broader and have prominent vesicle-filled nodal membrane protrusions, whereas axon shape and neurofilament density are dramatically altered. PNS qv(3J) nodes, some with detectable betaIV spectrins, are less affected. In contrast, a larger truncation of betaIV spectrins in qv(4J) mice, deleting the SD, PH, and ankyrinG binding domains, causes betaIV spectrins to be undetectable and causes dramatic changes, even in peripheral nodes. These results show that quivering mutations disrupt betaIV spectrin retention and stability at nodes and that distinct protein domains regulate nodal structural integrity and molecular organization.

Figure 1.

Identification and phenotype of qv^3J mice. A, Schematic of WT, qv^3J, and qv^4J mutant βIV spectrin. B, PCR-based screening for qv^3J homozygous mutant mice using the StyI restriction enzyme. C, Double immunostaining of rat sciatic nerve with antibodies against Nav channels (Pan Nav) or the SD domain of βIV spectrin (βIV SD) shows that the βIV SD epitope is absent in qv^3J mice. D, E, qv^3J mice clasp their hindlegs together and have significant gait abnormalities, including weakness, tremor, and a dragging of their hindlegs. F, Immunoblot analysis for a variety of proteins associated with myelinated axons, including Nav channels (Pan Nav, Nav1.2), Caspr, Kv1.2, CNP, MBP, β-actin, α-tubulin, and SD-containing βIV spectrin isoforms. Scale bars, 5 μm.

Figure 2.

Nodal Nav channel clusters are disrupted in the CNS of qv^3J mutants, but PNS nodes are normal. A, B, Double immunostaining for Nav1.6 (red) and Caspr (green). Nav1.6 clusters were often broader in qv^3J mutant mice (B, D, arrow and inset). C, D, Double immunostaining for Kv1.2 (green) and Nav1.6 (red). E, F, Double immunostaining for Caspr (red) and Nav1.6 (green). G, H, Kv1.2 (red) and Nav1.6 (green) immunostaining. Scale bars: A-H, 10 μm; insets, 2 μm.

Figure 3.

Node length and nodal membrane shape are distorted in qv^3J mutant mice. A, WT node of Ranvier from 4-month-old optic nerve. B-D, Optic nerve nodes of Ranvier from 4-month-old qv^3J mice are significantly longer (B). Paranodes in qv^3J mice have transverse bands (C, arrows). Many CNS nodes of Ranvier from qv^3J mice have prominent nodal protrusions. Arrows delineate the edges of the myelin sheathin A, Band D, E. F, Longitudinal sections of sciatic nerve nodes of Ranvier from 4-month-oldWT and qv^3J mice. Occasionally, nodal protrusions and vesicles were observed in qv^3J mice (F, arrow). G, H, Cross sections through sciatic nerve nodes of Ranvier from WT (G) and qv^3J (H) mice. Note that, although the membrane is deformed and there is an increase in vesicles in the axon in the qv^3J mutant, the Schwann cell microvilli are normal in appearance. Scale bars: A-F, 1 μm; G, H, 2 μm.

Figure 4.

Four-month-old qv^3J mice have dramatic changes in axon shape and cytoskeletal organization. A, B, Transverse optic nerve sections from WT (A) and qv^3J (B) mice show that qv^3J mouse axons are highly convoluted and not cylindrical. C, D, Longitudinal and transverse (inset) cross sections show that, compared with WT mice (C), qv^3J mice have a dramatic increase in the density of cytoskeletal elements (D). Nodes of Ranvier are delineated by arrows. E, F, High magnification of optic nerve cross sections shows qv^3J mice (F, arrow) have an increased density of neurofilaments compared with WT mice (E). G, Immunoblotting for neurofilament proteins shows that, in contrast to NF-L and NF-H, the amount of NF-M is increased in qv^3J mutant mice. The same blots were probed for α-tubulin as a control for protein loading. Scale bars: A, B, 3 μm; C, D, 1 μm; inset, 0.5 μm; E, F, 0.25 μm.

Figure 5.

Trafficking of βIV spectrin is not altered in qv^3J mice. Double labeling of WT or qv^3J optic (A-D) and sciatic (E-H) nerves using βIV NT (N-terminal; A, C, E, G) and Pan Nav (A′, C′, E′, G′) or AnkG (B, D, F, H) and Nav1.6 (B′, D′, F′, H′) antibodies. Double immunostaining of WT (I, I′) and qv^3J (J-L) sciatic nerve nodes of Ranvier withβIV NT (I-L) and anti-Caspr antibodies (I′, J′, K′, L′). In qv^3J opticnerve, βIVNT immunoreactivity is undetectable. However, in the qv^3J PNS, βIVNT was detected in reduced amounts at ∼40% of nodes (G, K, L). AnkyrinG immunoreactivity was detectable in qv^3J nodes but reduced in amount and in broader clusters in the CNS (D, arrows). Compared with WT mouse initial segments (M), Nav 1.6 immunoreactivity was reduced at axon initial segments in qv3J mutant mice (N). Scale bars: A-H, 3 μm; I-L, 5 μm; M, N, 10 μm.

Figure 6.

Compound action potentials from PNS and CNS nerves in qv^3J mice. A, Examples of CAPs recorded from WT and qv^3J sciatic and optic nerves at 25°C and 37°C. CAPs were recorded using suction electrodes. B, Peak conduction velocities. Error bars indicate ±SD.

Figure 7.

Nodes of Ranvier are disrupted in the PNS of qv^4J mutant mice. A, B, Immunostaining for Caspr (green) and Pan Nav channels (red) shows qv^4J mice have many disrupted nodes (B, arrow). C, D, Double labeling for Pan Nav channels (red) and Kv1.2 (green) shows Kv1 channels invade into nodal regions in the qv^4J mutant mouse (arrows). E-H, Double labeling with Pan Nav (red) and anti-ErbB2 (green) antibodies shows that, compared with WT (E, F) and qv^3J mice (G), Schwann cell microvilli in qv^4J mice (H) are disrupted. Note that, in F-H, axons run diagonally from bottom left corner to top right corner. Scale bars: A-D, F-H, 5 μm; E, 3 μm.