Olfactory ensheathing cell membrane properties are shaped by connectivity

Abstract

Olfactory ensheathing cells (OECs) have been repeatedly implicated in mediating plasticity, particularly in situ in the olfactory nerve in which they support the extension of olfactory sensory neuron (OSN) axons from the olfactory epithelium to the olfactory bulb (OB). OECs are specialized glia whose processes surround OSN axon fascicles within the olfactory nerve and across the OB surface. Despite their purported importance in promoting axon extension, and following transplants, little is known about either morphology or biophysical properties of OECs in situ. In particular, cell-cell interactions that may influence OEC function are largely unexplored. Here, we studied OEC connectivity and morphology in slice preparations, preserving tissue structure and cell-cell interactions. Our analyses showed that OECs form a matrix of cellular projections surrounding axons, unique among glia, and express high levels of connexin-43. Lucifer Yellow injections revealed selective dye coupling among small subgroups of OECs. Two types of OECs were biophysically distinguished with whole-cell voltage-clamp recordings: (1) with low-input resistance (R(i)), linear current profiles, and frequently dye coupled; and (2) with high R(i), nonlinear current profiles, and infrequent dye coupling. Pharmacological blockade of gap junctions changed OEC membrane properties such that linear OECs became nonlinear. Double recordings indicated that the appearance of the nonlinear current profile was associated with the loss of electrical coupling between OECs. We conclude that the diversity of OEC current profiles can be explained by differences in gap-junction connectivity and discuss implications of this diversity for OEC influences on axon growth and excitability.

Figure 1

OECs are closely associated with axons in the ONL. Immunohistochemistry in OB sections of S100-GFP mice (P26). A and B, OB layers indicating the locations shown at higher magnification to the right (square). ONL: Olfactory nerve layer; GL: glomerular layer; EPL: external plexiform layer. Nuclei were stained with DRAQ5. A′-A‴, Colocalization of GFP (green) with BLBP (red) in OECs. Insets: higher magnifications of the area inside the dashed square; Scale bar = 5 μm. Filled arrowheads: OEC fine processes; open arrowhead: OEC cell body; asterisk: next to endothelial cell. PGs: Periglomerular cells. B′-B‴, Immunoreactivity for NCAM, a marker for OSN axons (red), is in the lacunae (asterisks) defined by the matrix of fine OEC processes (arrowheads).

Figure 2

OEC projections align with neighboring axon bundles. A, Slicing procedure for horizontal slices; the anteroventral surface of the brain was glued to a vibratome stage and sliced. B, A horizontal OB slice, fixed and stained for nuclei with DRAQ5. Dashed line: limit between the olfactory nerve layer (ONL), where cells were recorded, and the glomerular layer (GL). C, Fluorescence images showing a LY-filled OEC at three different depths in the tissue slice (superficial to the left). Asterisk: cell body; white arrowheads: projections; ‘p’: recording pipette. D, DIC image of the areas in C. Black arrowhead: pipette tip. Axon bundles delineated with red and blue (dashed lines) run in nearly perpendicular directions. E, Cartoons of the outlined cell body and projections in C (green) and areas in D. Black arrows: orientation of axon bundles.

Figure 3

OECs are elongated cells with fine projections forming a matrix. A, Volume 3D-reconstruction of a LY-filled OEC (Grid: 5 μm). Nuclei were stained with DRAQ5. Inset: matrix of lamellar projections (dashed circle). B, Points (red) defined to estimate the length of OEC projections. Inset: the OEC main axis is in blue and a bounding box enclosing the cell is in yellow (see Materials and methods). C, Isosurface rendering using the green channel. The blue channel was used to generate a contour surface of the nucleus. D, Length of the longest segment from the cell body to the tip of lamellae, as indicated in B (n = 8). E, Length, width and depth of the bounding boxes enclosing each OEC, as shown in B (inset) (n = 8). * p < 0.001 (Friedman test). F, Surface and volume of isosurface renderings as the one shown in C (n = 8). G, Surface vs. volume plot (full line: linear fit; broken lines: 95% confidence interval; r² = 0.75; n = 8).

Figure 4

Ultrastructural correlate of the matrix formed by dye-filled OECs. A, Single optical section of a LY-filled OEC. Right panel: lacunae defined by fine interdigitations fitted to elliptical regions of interest (red) used for quantification (see Materials and methods). B, Electron micrograph of the olfactory nerve layer. Blue: contoured glial processes; red asterisks: axon bundles surrounded by glial projections. C, The same contour as in B showing the strategy used to quantify the average diameter of lacunae at the EM level. Maximum and minimum diameters are indicated with perpendicular red lines. D, Histogram showing the average diameter of lacunae measured in confocal single optical sections (black bars) or in electron micrographs (white bars).

Figure 5

Current profiles in OECs. A, Representative recordings of an OEC with non-linear (left) or linear (right) current profiles in response to a series of 20-mV, 400-ms voltage steps; holding potential: -80 mV. The series of 15 steps started at -160 mV (schematized below). B, I/V curves of the recordings in A, with currents measured at the times indicated with symbols. The range indicated by the dashed rectangle was fitted to a straight line (expanded below). C, Histograms of Ri (left) and Vr (right) from cells with non-linear (black bars) and linear (white bars) cells. Bell-shape curves: Gaussian fits to the non-linear cell (full line) or linear cell (broken line) data.

Figure 6

Dye coupling in OECs. A, Cluster of 11 dye-coupled OECs (green) with nuclei stained with DRAQ5 (blue). Grid: 10 μm. Arrow in A and B: recorded cell. GL: glomerular layer. Dashed lines delineate the glomeruli. Asterisks in B: nuclei of dye-coupled cells. C, Isosurface rendering of the green channel. The nuclear staining (blue channel) was used to generate contour surfaces of the nuclei of the injected cell (blue) and dye-coupled cells (purple). D, Segments: linear distance between the cell body of the injected cell and those of the coupled cells (red) and cluster main axis (blue, see Materials and methods). Inset: the same cluster oriented with north (N) pointing up and the equator indicated (dashed yellow line). E, Proportion of coupled cells vs. the proportion of nuclei in the northern hemisphere (N). Each point is a different cluster. The vertical and horizontal dotted lines correspond to symmetrical distribution of nuclei and coupled cells on both sides of the equator, respectively. The dashed line corresponds to slope=1. F, Dye intensity at the cell bodies of coupled cells vs. the linear distance to the injected cell. Each point is a different cell and each color is a different cluster, matching E. The lines in E and F are linear regressions.

Figure 7

Gap junction coupling in OECs. A, Percentage of the total recorded non-linear (NL) or linear (L) cells dye-coupled to other cells (black column). The complementary percentage of uncoupled cells is shown in white. B, Cluster size (number of stained cells) associated with non-linear (NL) or linear (L) cells. C, Representative recordings of membrane currents of two OECs recorded simultaneously (cell 1 at the top and cell 2 at the bottom). Left panel: responses of the two cells when cell 1 was stimulated with a series of 20-mV voltage steps; holding potential: -80 mV. The series of 13 steps started at -160 mV. Right panel: responses when the same stimulation protocol was applied only to cell 2. The command voltage (Com) is schematized above (cell 1) or below (cell 2) the traces of the stimulated cell. Scale bars: 2 nA (stimulated cell) or 100 pA (non-stimulated cell), and 100 ms. D, The same cells and experimental protocol as in C, during bath-application of the gap-junction blocker meclofenamic acid (MFA, 100μM). Scale bars: same as in D; inset scale bars = 0.5 nA, 100 ms. E, Average I/V curves (n = 8) of OECs subjected to the voltage steps in control conditions (open circles) and after application of MFA (filled circles). The current amplitude was measured at the time indicated with symbols in C and D. F, Average I/V curve of the MFA-sensitive current (n = 8). Error bars represent SEM.

Figure 8

Expression of Connexin 43 in OECs. A, Immunohistochemistry in an OB section of S100-GFP mice (p26). Low magnification of the OB layers indicating the locations where the images were acquired (square). Nuclei were stained with DRAQ5. A′-A‴, Clusters of Cx43 immunoreactivity (red) colocalize with GFP-labeled fine projections of OECs (arrowheads). Insets: higher magnifications of the area inside the dashed square; Scale bar = 5 μm. B, Average intensity of Cx43 immunoreactivity in each layer of the OB, normalized to intensity in the ONL. Error bars represent SEM. C, Single optical section of a z-stack, with lateral projections showing Cx43 immunoreactivity in a tissue slice with a LY-filled OEC. Inset: higher magnification of the dashed rectangle indicated at the right. Arrowhead: a Cx43 cluster colocalizes with an OEC fine projection. D, Isosurface rendering of the green and red channels of a z-stack similar to the one in C. The red surface was edited to eliminate the clusters not apposed to the cell. Arrowhead: contour surface of the nucleus generated with the blue channel. E, Electron micrograph of the ONL. Glial processes were contoured (blue); red lines: gap junction specializations. Inset: high magnification of the gap junction enclosed in the rectangle.