Peripheral axons of the adult zebrafish maxillary barbel extensively remyelinate during sensory appendage regeneration

Abstract

Myelination is a cellular adaptation allowing rapid conduction along axons. We have investigated peripheral axons of the zebrafish maxillary barbel (ZMB), an optically clear sensory appendage. Each barbel carries taste buds, solitary chemosensory cells, and epithelial nerve endings, all of which regenerate after amputation (LeClair and Topczewski [2010] PLoS One 5:e8737). The ZMB contains axons from the facial nerve; however, myelination within the barbel itself has not been established. Transcripts of myelin basic protein (mbp) are expressed in normal and regenerating adult barbels, indicating activity in both maintenance and repair. Myelin was confirmed in situ by using toluidine blue, an anti-MBP antibody, and transmission electron microscopy (TEM). The adult ZMB contains ∼180 small-diameter axons (<2 μm), approximately 60% of which are myelinated. Developmental myelination was observed via whole-mount immunohistochemistry 4-6 weeks postfertilization, showing myelin sheaths lagging behind growing axons. Early-regenerating axons (10 days postsurgery), having no or few myelin layers, were disorganized within a fibroblast-rich collagenous scar. Twenty-eight days postsurgery, barbel axons had grown out several millimeters and were organized with compact myelin sheaths. Fiber types and axon areas were similar between normal and regenerated tissue; within 4 weeks, regenerating axons restored ∼85% of normal myelin thickness. Regenerating barbels express multiple promyelinating transcription factors (sox10, oct6 = pou3f1; krox20a/b = egr2a/b) typical of Schwann cells. These observations extend our understanding of the zebrafish peripheral nervous system within a little-studied sensory appendage. The accessible ZMB provides a novel context for studying axon regeneration, Schwann cell migration, and remyelination in a model vertebrate.

Figure 1

Anatomical overview of adult zebrafish maxillary barbel (ZMB) regeneration. A: Morphological sequence of ZMB regeneration from 10 to 28 days postamputation. Proximal amputation (dotted line) induces regeneration of a new appendage (black). Barbel regeneration is highly variable; the length of the regenerated tissue is approximate. B: Gross morphology of a regenerated maxillary barbel after 28 days at 28°C. The amputation plane (dotted line) is marked by the blunt stump of the central rod of connective tissue, which does not regrow. nb, Nasal barbel; mb, maxillary barbel.

Figure 2

Normal and regenerating maxillary barbels express alternatively spliced transcripts for myelin basic protein (mbp). Top: Equal amounts of DNase-treated total RNA were analyzed by RT-PCR for the presence of mbp, a common marker of zebrafish myelination. The primers used amplify a doublet, indicating two alternatively spliced transcripts (see Buckley, 2010). Wild-type zebrafish embryos 5 days postfertilization (5 dpf) express both transcripts (+ = positive control). Similar bands were detected in normal adult barbels (norm) and the distal tips of regenerating barbels at 10 and 14 days postsurgery (10 dps and 14 dps). A parallel reaction with water as the template (H₂O, negative control) gave no product. Bottom: All RNA lanes amplified a single band for β-actin. The starting amounts of total RNA were identical to those in the top panel.

Figure 3

Myelination of the deep nerve tracts in the zebrafish maxillary barbel. Barbel tissues were collected from adult wild-type AB zebrafish and prepared as thin cross-sections for light microscopy. A: A thin (1 μm) plastic section of a normal barbel was stained with toluidine blue O. Myelinated axons with dark, ring-like sheaths appear in the dorsal nerve (dn) and ventral nerve (vn) areas surrounding the central rod of connective tissue (cr). B: A similar plastic section stained with DAPI (blue) and an antibody against myelin basic protein (MBP; green). Numerous bright punctae overlay the dorsal and ventral populations of neurons. C,D: Validation of the anti-MBP primary antibody on sections of zebrafish optic tectum. C: DAPI stain (blue), showing nuclei of the tectal layers. D: DAPI merged with anti-MBP (green). Immunostaining was observed in the expected layers of the tectum, specifically the stratum opticum (so) and stratum album centrale (sac). Positive staining in these layers has been used to validate other myelin-specific antibodies in zebrafish (compare Fig. 1 in Bai et al., 2011). sm, Stratum marginale; so, stratum opticum; sfgs, stratum fibrosum et griseum superficiale; sgc, stratum griseum centrale; sac, stratum album centrale; spv, stratum periventriculare.

Figure 4

Transmission electron microscopy (TEM) of maxillary barbel myelination. Barbel tissues were collected from adult wild-type AB zebrafish, embedded, and cut at 70 nm. To improve focus, A, B, and D were enhanced with a digital filter (Unsharp Mask) in Adobe Photoshop. A: Overview of the axon-rich ventral area in a normal adult maxillary barbel. Multiple small (1-2 μm) myelinated axons are surrounded by accompanying Schwann cells. B: In each barbel, there are several ‘giant’ axons. C: High-magnification TEM showing a single barbel axon (a) with numerous closely packed myelin lamellae (m). The axoplasm is full of smaller, grayish microtubules. D: Normal barbels also contain bundles of small, unmyelinated axons completely or partially wrapped by the cytoplasm of supporting cells. a, Axon, c, collagen bundle, en, endothelial cell, m, myelin, s, Schwann cell nucleus.

Figure 5

Development of myelination in the juvenile zebrafish maxillary barbel. Juvenile wild-type AB zebrafish (30-44 days postfertilization) were stained in whole mount to observe the outgrowth of maxillary barbel axons (antiacetylated tubulin, red) and/or the deposition of myelin (anti-mbp, green). DAPI was used as a nuclear counterstain (blue). The distal tip of each barbel projects rightward. A-C: Early maxillary barbel buds (mb, ~150 μm from base to tip) have one ventral nerve bundle that extends to the distal tip (arrows, B,C). At this stage, these axons appear unmyelinated (not shown). D-F: More developed barbel buds (~300 μm from base to tip) have more nerve fibers (arrows, E,F) and more projections to the ventral epithelium. Overt myelination is still absent (not shown). The anti-tubulin antibody also labels neuromast organs of the lateral line (arrowheads, E,F). G: Juvenile maxillary barbels (~1 mm long) have partial myelination of the ventral, but not dorsal, nerve tracts. Myelination proceeds distally and lags the growing axons. dn, Dorsal nerves; vn, ventral nerves. H: Ventral view of a juvenile maxillary barbel (~1 mm long). In this view, the ventral nerves (vn) are roughly divided into two lateral bundles (arrows, top panel). Only some of these tracts are myelinated (MBP, bottom panel), and myelination extends for only part of the way to the growing tip. [Color figure can be viewed in the online issue, which is available at wileyonlinelibrary.com.]

Figure 6

Myelination of mature axons in the adult zebrafish maxillary barbel. Barbel tissues were collected from adult wild-type AB zebrafish and prepared as whole mounts for immunohistochemical staining of acetylated tubulin (AC-TUB; red) and myelin basic protein (MBP; green). This figure shows the midsection of the barbel shaft; proximal is to the left and distal to the right. A: The adult maxillary barbel has two populations of nerves, dorsal nerves (dn) and ventral nerves (vn). A few small-caliber fibers run orthogonally (asterix). B: Most axons are myelinated, with thick MBP-positive sheaths; however, some fibers or regions of fibers appear unmyelinated (arrow) or with gaps (arrowhead). C: Merged view of the two channels. dn, Dorsal nerves; vn, ventral nerves. A magenta-green copy of this image is provided as Supporting Information Figure 1.

Figure 7

Ultrastructure of zebrafish maxillary barbel regenerates 10 days postsurgery (dps). Maxillary barbels were amputated from adult wild-type AB zebrafish and allowed to regenerate for 10 days. The regenerating portion of the barbel shaft was fixed and prepared as cross-sections for light and transmission electron microscopy (TEM). A: Light micrograph of a 10-dps regenerating barbel stained with toluidine blue O. The epithelial (e) and mesenchymal layers appear patchy, and the central rod (cr) is disorganized. The ventral nerve region (vn) lacks dark rings of myelin (compare Fig. 3A). B: Low-magnification TEM of the 10-dps ventral nerve region shows numerous electronlight voids (v) filled with cell debris, surrounded by fibroblasts (f) and collagen bundles (c). There are few intact axons (a). C: Similar magnification of a second ventral nerve region, showing structures as in B. D: Magnification of the boxed area in C. The right structure has six to eight loose myelin layers (arrowhead). E,F: Regenerates at 10 dps are full of large fibroblastic cells (f) filled with rough endoplasmic reticulum and Golgi, in close contact with cross-cut and longitudinally cut collagen bundles (c). a, Axon; bm, basement membrane; bv, blood vessel; c, collagen bundle; cr, central rod; e, epithelium; f, fibroblast; v, void; vn, ventral nerves. [Color figure can be viewed in the online issue, which is available at wileyonlinelibrary.com.]

Figure 8

Ultrastructure of zebrafish maxillary barbel regenerates 28 days postsurgery (dps). A: Maxillary barbels were amputated from adult wild-type AB zebrafish and allowed to regenerate for 28 dps. The barbels were fixed and prepared as cross-sections for light and transmission electron microscopy (TEM). Sections were cut to sample regenerating tissues ~0.5-1.0 mm distal to the amputation plane (far right, dotted lines). B: Light micrograph of a 28-dps regenerating barbel stained with toluidine blue O. The epithelium is compact, with regenerated goblet cells (gc) and taste buds (tb). Abundant myelinated axons fill a large ventral nerve trunk (vn) surrounded by a disorganized central rod (cr). Two small blood vessels (bv) carry erythrocytes. C: Electron micrograph of a second 28-dps barbel, showing structures similar to those showing in B. D: Magnification of the boxed area in C. The ventral nerve trunk is divided into several smaller bundles, each with 10-15 axons. Most axons appear myelinated, with accompanying Schwann cell nuclei (s). E: Magnification of the boxed area in D. Ventral nerve axons are heterogeneous, including large and small myelinated axons and smaller unmyelinated ones (asterisks). bv, Blood vessel; cr, central rod; e, epithelium; gc, goblet cell; tb, taste bud; s, Schwann cell nucleus; vn, ventral nerves; ZMB, zebrafish maxillary barbel; asterisk, unmyelinated axon. [Color figure can be viewed in the online issue, which is available at wileyonlinelibrary.com.]

Figure 9

Distribution of axoplasm area and myelin thickness in normal and regenerated (28 days postsurgery) zebrafish maxillary barbels. Adult wild-type AB zebrafish barbels were prepared for transmission electron microscopy, and 982 axons were measured for cross-sectional area (μm²) and thickness of myelin (μm), if present. For complete descriptive statistics see Table 2. A,B: Pooled, within-group histograms of axon cross-sectional area in normal (A) and regenerated (B) maxillary barbel axons. Myelinated and unmyelinated fiber distributions are interleaved. C,D: Pooled, within-group histograms of myelin thickness in normal (C) and regenerated (D) maxillary barbel axons. E,F: Cumulative frequency plots for myelinated axon cross-sectional area (E) and myelin thickness (F), comparing normal and regenerated tissue (solid vs. dotted line). Normal and regenerated barbels have equivalent distributions of axon caliber; regenerated axons have slightly thinner myelin.

Figure 10

g-ratio analysis of normal and regenerated (28 days postsurgery) maxillary barbel axons. Adult wild-type AB zebrafish barbels (three normal, three regenerated) were prepared for transmission electron microscopy and digitized for morphometric analysis, yielding a g-ratio = (axon diameter/fiber diameter) for each myelinated axon. Larger g-ratios indicate thinner myelin. For complete descriptive statistics see Table 2. A,B: Pooled, within-group histograms of g-ratio for normal (A) and regenerated (B) maxillary barbel axons. C,D: Scattergrams of g-ratio vs. axon diameter (μm) in normal (C) and regenerated (D) maxillary barbel axons.

Figure 11

Expression of promyelinating transcription factors in the distal regenerating zebrafish maxillary barbel. A: Maxillary barbels from adult wild-type AB zebrafish were amputated proximally and allowed to regrow for 28 days. Only the distal, regenerated tissue was processed for RNA isolation. B: Equal amounts of DNase-treated total RNA were split into separate one-step reaction tubes and simultaneously amplified by reverse-transcription polymerase chain reaction (RT-PCR). Primers for the promyelinating transcription factors sox10, pou3f1, and egr2a/b (=krox20a/b) all amplified the predicted bands (for product sizes see Table 1). To check RNA integrity, β-actin was amplified simultaneously as a positive control (+). A negative control (−) used the same master mix reagents + β-actin primers, but no RNA.