Live-imaging of endogenous neurofascins reveals glial adhesion shapes developing nodes of Ranvier

The rapid transmission of electrical signals in the vertebrate nervous system depends on a highly ordered molecular architecture. Specifically, axons must organize themselves into specialized domains, such as the nodes of Ranvier, to allow for saltatory conduction (the "jumping" of electrical impulses between gaps in insulation). This process relies on a delicate cooperation between neurons and glia (support cells that wrap axons in myelin).

However, scientists have struggled to watch these microscopic domains assemble in real time. Most studies rely on fixed tissue, which provides only a static snapshot. Other studies use transgenic overexpression (forcing cells to produce extra proteins), but this often distorts the very structures being studied. Researchers report that overexpressing neurofascin significantly increases the apparent size of both nodes and paranodes .

This creates a misleading picture of the submicron domains that characterize healthy nerves.

A new study from the University of Edinburgh uses a novel genetic toolkit in zebrafish to solve this problem. By generating endogenous fluorescent reporters, the researchers reveal that the refinement of these nerve domains is actively driven by glial cells.

The limitations of static snapshots and overexpression

To understand how nerve insulation matures, researchers traditionally rely on two main methods. The first is immunohistochemistry (staining fixed, dead tissue with antibodies). This offers a high-resolution look at protein location, but it cannot capture real-time movements.

The second is transgenic overexpression. While this allows for live imaging, it often introduces significant artifacts. Because these methods flood the cell with more protein than it naturally produces, they can distort the submicron morphology of the domains. The authors report that overexpression leads to larger apparent node and paranode lengths . Consequently, the field has lacked a way to witness the natural "tightening" of these domains as they mature in vivo (within a living organism).

Engineering endogenous fluorescent reporters

The researchers bypassed these issues by using a PCR-tagging strategy to create knock-in zebrafish lines. Instead of adding extra copies of a protein, they integrated fluorescent sequences directly into the existing genomic loci of the zebrafish neurofascin paralogues, nfasca and nfascb. This ensures the proteins are expressed at their natural, physiological levels.

The architectural choice that makes this work is the location of the tag. The authors targeted the intracellular C-termini (the tail end inside the cell) of the proteins .

This approach allows for the simultaneous visualization of two distinct populations. Neuronal neurofascin (nfasca) is tagged with mRuby3 and glial neurofascin (nfascb) is tagged with mEGFP.

By focusing on the zebrafish, the team leveraged a unique evolutionary advantage. Following a genome duplication event, zebrafish possess two separate neurofascin genes [Figure 1B]. These genes have specialized to handle neuronal and glial roles respectively. This allowed the researchers to monitor the axon and the myelinating glia independently and in real time.

Observing the compaction of nodal domains

Using these tools to monitor the posterior lateral line (pLL) nerve, the authors measured how these domains change during development. They found that the nodes of Ranvier do not stay a fixed size. Instead, they undergo a period of "morphological refinement" or compaction.

The paper reports a significant reduction in the length of these domains between 3 and 5 days post-fertilization (dpf). Specifically, the authors measure the average length of nfasca-mRuby3 nodal domains decreasing from 0.95±0.08µm to 0.72±0.01µm [Figure 5B-C]. This represents a 25% reduction in size. Similarly, the distance between adjacent paranodal domains decreases by approximately 18% [Figure 5E-F]. Crucially, the researchers also observed that the clusters of voltage-gated sodium channels (the proteins responsible for the electrical impulse) shrink by roughly 28% [Figure 5H].

To determine if this compaction was driven by the axon or the glia, the researchers performed two key experiments. First, they inhibited ErbB signaling to prevent Schwann cells (the glial cells in the peripheral nervous system) from maturing [Figure 6A-D]. This effectively blocked the initial assembly of any nodes. Second, they used CRISPR-Cas9 to knock out the glial neurofascin (nfascb). In these "crispant" fish, the authors report a massive breakdown in refinement. The nodal gaps increased by 63%, and the nodal clusters were 65% longer than in healthy controls [Figure 7B-F].

Assessing the scope of the findings

While the study provides a high-resolution look at developmental dynamics, it is not without limitations. The researchers acknowledge that C-terminal tagging might still subtly influence the kinetics of protein trafficking. It might also affect how the proteins interact with the internal cytoskeleton (the structural framework of the cell).

Furthermore, because these are germ-line knock-ins, the fluorescent tags are expressed in all cells. In very dense, heavily myelinated tracts, this can make it difficult to achieve single-cell resolution. Finally, the study focuses on the peripheral nervous system of a zebrafish. While the fundamental biology of neurofascins is conserved, the authors note that it remains to be seen if these same glial-driven compaction mechanisms operate identically in the central nervous system (the brain and spinal cord).

A new model for nodal maturation

The verdict is clear: the maturation of the nerve insulation is a collaborative, multi-step process. The initial assembly of the node is triggered by glial cells. However, the subsequent "tightening" or compaction of the node is a specialized task driven by the glial neurofascin (nfascb) at the paranode.

This discovery shifts our understanding of the axon-myelin interface. Rather than viewing the node as a static structure once it is built, we must see it as a dynamic domain. It requires ongoing glial input to reach its optimal, compact state. For researchers, this implies that many pathologies involving the loss of nerve conduction speed might involve failures in this active, glial-mediated maintenance.

Figures from the paper