Table 1.
Antibodies used in this study, their binding sites in olfactory epithelium and olfactory bulb, and possible functional implications.
Figure 1.
Immunohistochemistry of olfactory epithelium (OE) and olfactory bulb (OB) in adult NPC1+/+ and NPC1−/− mice (67–70 d).
Beta-tubulin (BT, A–D); Olfactory marker protein (OMP, E–H). (A) Regular columnar epithelium (OE) and olfactory nerve fibers in the lamina propria (LP). In contrast, the number of BT-positive cells is markedly decreased (B). There are large gap-like spaces in the suprabasal area of the epithelium (arrows). (C) Olfactory mucosa (right side) and nerve fiber layer of the OB show a normal morphology. Bundles of olfactory nerve fibers travel through gaps of the cribriform plate (CP) into the olfactory nerve layer (ONL) of the OB (left side). (D) ORN are more scattered and appear interrupted in the NPC1−/− animal. (E, G) OMP reactivity in NPC1+/+, detected in mature ORN, is comparable to the distribution of BT (see A and C). NPC1−/− mice (F, H) exhibit similar deficits as shown for BT in B and D. OMP antigen is also located in nuclei. Counterstained with hematoxylin.
Figure 2.
Immunohistochemistry of NeuN (A, B) and GFAP (C through F) in the OB.
Distribution of NeuN-positive neuronal cells do not reveal any obvious differences between NPC1+/+ (A) and NPC1−/− mice (B) (67–70 d). There is a clear increase of GFAP expression, especially in the glomerular layer (Gl) in NPC1−/− animals (D) compared to NPC1+/+ (C). The area marked by the rectangles is enlarged in E and F, respectively, demonstrating the astrogliosis of the granular layer (Gr) in NPC1−/− mice. Counterstained with hematoxylin.
Figure 3.
Immunohistochemistry of Gal-3 in the OE (A, B) and OB (D, E) (67–70 d).
(A) Gal-3 in NPC1+/+ is restricted to olfactory axon bundles in the lamina propria, whereas in NPC1−/− animals (B), large spot-like Gal-3 reaction product is detectable above and below the basal membrane. Note tender processes of these cells, presumably macrophages (arrows). These cells correspond most likely to those depicted in C. (C) Interface between olfactory mucosa and lamina propria. Electron microscopic resolution of two macrophages (red) filled with myelin-like material (arrows). These cells penetrate the basal lamina (BL). N- nuclei of horizontal basal cells; Ap- nucleus of an apoptotic cell. (D) NPC1+/+, cortical layers of the olfactory bulb express Gal-3 immunoreactivity only in nuclei of ensheathing cells within the nerve fiber layer (arrows). Deeper bulb areas, i.e., as glomerular (Gl), mitral cell (Mi), and granular (Gr) layers, are not affected. (E) Numerous Gal-3-positive cells occupy the glomerular and adjacent part of the external plexiform layer in an NPC1−/− animal. These cells have partly short and interrupted processes different from those of astroglia. Glomeruli are barely recognizable. (F, G) Double immunofluorescence, using antibodies against GFAP (red) and Gal-3 (green). (F) GFAP reactive astrocytes are mainly distributed within the glomerular layer. There are only a few Gal-3-positive microglia cells in NPC1+/+ animals. (G) Prominent microgliosis (Gal-3 expression) especially around glomeruli goes along with enhanced reactivity for GFAP, indicating which indicates astrogliosis. Macrophage activity is most likely associated with microglia rather than with activated astrocytes. Astrocytes do not co-express Gal-3.
Figure 4.
CatD (A–D) and ABCA-1 (E–H) immunoreactivity.
(A) Normal distribution of CatD reactivity in NPC1+/+; typical dot-like residues in cells of the lamina propria and basal cells of the olfactory epithelium. (B) Dramatically increased CatD reactivity is detected in all epithelial cells in NPC1−/−, especially suprabasal cells, most probably macrophages as demonstrated in Figs. 4C and 8. Also enhanced reactivity also of ensheathing cells of ORN bundles in the lamina propria (asterisk). (C) Overview of the interface between OE and OB in NPC1+/+ mice. (D) CatD reactivity is dramatically increased in all layers of the OB, including mitral cells (small arrows, upper right corner) and periglomerular cells in NPC1−/−. Gl, glomerular layer, CP, cribriform plate. (E, F) Reactivity of ABCA-1 is increased in nerve fibers of NPC1−/− animals (F), whereas supporting cells (arrows) of the OE displays similar intensities in both wild type (E) and mutant (F) animals.
Figure 5.
Western blot analysis of the olfactory system in NPC1−/− mice compared to normal controls.
A: NPC1 deficiency of the mice in the brain. B/C: OMP is reduced in OE and OB. D/E: The signal for gliosis marker GFAP is enhanced in the NPC1−/− mouse in both OE and OB. F/G: Marker of mature neurons MAP-2 shows a decreased signal in OE and OB. Pro-inflammatory and macrophage markers CatD (I) and Gal-3 (K) were elevated in OB. For CatD, a pronounced increase was observed also in OE (H). L/M: ABCA-1, a transporter known to respond to intracellular cholesterol dysregulation is up-regulated in the OE and OB of NPC1 deficient animals.
Figure 6.
Electron microscopic depiction of the olfactory epithelium of a young (32 d) NPC1−/− animal.
Some of the cells are exemplarily colored. The OE consists 1) of olfactory receptor neurons in the middle portion of the OE (O, green) with thin apical processes extending with knob-like structures into the mucous layer, 2) supporting cells, the nuclei of which lie in the upper third, and with basal footplates (S, yellow) above the basal membrane (magenta line), and 3) basal cells (B). There are many irregular, myelin-like inclusions within most of the cells (arrows). Most of these autophagosomes occur in subnuclear portions of supporting cells, but also, to a lesser degree, in perinuclear locations of ORN. Apical processes of ORN are not affected (see also Fig. 7). Singular, huge accumulations of myelin-like deposits are occasionally seen near the basal membrane, most likely corresponding to Gal-3 – and CatD-positive macrophages (M, red; see also Figs. 4,5). BG, Bowman glands of the lamina propria.
Figure 7.
Superficial part of the olfactory epithelium in an NPC1−/− animal.
Most prominent are perinuclear autophagosomes in supporting cells (yellow, arrows). ORN (green) occasionally contain such deposits, but appear normal at the surface. Tight junctions are intact (arrowheads). One microvillar cell (MV) is not affected.
Figure 8.
(A) Apical surface of the OE in an NPC1−/− animal. Myelin-like figures (arrows) seem to be shed from one of the OE epithelial cells into the mucous layer. (B) Characteristic pattern of multilamellar deposits within one of the ORN in the middle of the OE. (C) Lamina propria with an ensheathing cell (EC, blue), enwrapping axons of ORN (green). Membrane fragments and vesicular deposits are accumulated exclusively in dilated cisterns of EC, but not in ORN.
Figure 9.
(A) Low power resolution of the olfactory nerve fiber layer (ONL, right) and a glomerulus (GL, left). Similar to the situation in the lamina propria, numerous autophagosomes are seen in ensheathing cells (arrows). The glomerulus is almost free of these deposits. (B) High power resolution of a large degenerating astrocyte cell process (A) with remnants of membranes, vesicles and intermediate filaments (asterisk). Olfactory receptor neurons (ORN) are intact. (C) Dendritic compartment of a glomerulus. Synaptic connectivity seems not to be disrupted, and cells are largely free of deposits. D- mitral cell dendrite (green); P, periglomerular cell (red), A, astrocyte (blue); O, axon of olfactory receptor neuron. (D) Perikaryon of a mitral cell with autophagosomes containing membranous material. (E) Trigeminal ganglion cell (G) and an enwrapping satellite cell (S). The latter is packed with autophagosomes.
Figure 10.
On average, recordings from the olfactory epithelium exhibited decreased amplitudes after exposure to PEA (A), H2S (B), and C02 (C) in NPC1−/− mice.
Differences tended to be significant for adult (67 d) but not for young animals (31 d). Individual EOG recordings are shown on the right side. X-axis gives the time after volatile stimulation (at 0 sec) and y-axis shows the response in mV for a wild type (WT) and a NPC1−/− mouse (NP). The response amplitude is the minimum of each graph and the latency is the time between the stimulus and the minimum. Duration of recordings shown −400 msec.