Sendai Virus Induces Persistent Olfactory Dysfunction in a Murine Model of PVOD via Effects on Apoptosis, Cell Proliferation, and Response to Odorants

Abstract

Background: Viral infection is a common cause of olfactory dysfunction. The complexities of studying post-viral olfactory loss in humans have impaired further progress in understanding the underlying mechanism. Recently, evidence from clinical studies has implicated Parainfluenza virus 3 as a causal agent. An animal model of post viral olfactory disorders (PVOD) would allow better understanding of disease pathogenesis and represent a major advance in the field.

Objective: To develop a mouse model of PVOD by evaluating the effects of Sendai virus (SeV), the murine counterpart of Parainfluenza virus, on olfactory function and regenerative ability of the olfactory epithelium.

Methods: C57BL/6 mice (6-8 months old) were inoculated intranasally with SeV or ultraviolet (UV)-inactivated virus (UV-SeV). On days 3, 10, 15, 30 and 60 post-infection, olfactory epithelium was harvested and analyzed by histopathology and immunohistochemical detection of S-phase nuclei. We also measured apoptosis by TUNEL assay and viral load by real-time PCR. The buried food test (BFT) was used to measure olfactory function of mice at day 60. In parallel, cultured murine olfactory sensory neurons (OSNs) infected with SeV or UV-SeV were tested for odorant-mixture response by measuring changes in intracellular calcium concentrations indicated by fura-4 AM assay.

Results: Mice infected with SeV suffered from olfactory dysfunction, peaking on day 15, with no loss observed with UV-SeV. At 60 days, four out of 12 mice infected with SeV still had not recovered, with continued normal function in controls. Viral copies of SeV persisted in both the olfactory epithelium (OE) and the olfactory bulb (OB) for at least 60 days. At day 10 and after, both unit length labeling index (ULLI) of apoptosis and ULLI of proliferation in the SeV group was markedly less than the UV-SeV group. In primary cultured OSNs infected by SeV, the percentage of cells responding to mixed odors was markedly lower in the SeV group compared to UV-SeV (P = 0.007).

Conclusion: We demonstrate that SeV impairs olfaction, persists in OE and OB tissue, reduces their regenerative ability, and impairs the normal physiological function of OSNs without gross cytopathology. This mouse model shares key features of human post-viral olfactory loss, supporting its future use in studies of PVOD. Further testing and development of this model should allow us to clarify the pathophysiology of PVOD.

Fig 1. Location of olfactory epithelium in murine nasal cavity and section for histology.

The dotted line indicated the location of coronal section at Level III. Shaded areas are the ethmoid turbinates, which are covered by olfactory epithelium.

Fig 2. Time course of the incidence of olfactory dysfunction in mice infected with SeV and UV-Inactivated SeV.

Fig 3. Viral load of SeV by realtime RT-PCR over time.

(A) Olfactory Bulb. (B) Olfactory Epithelium.

Fig 4. Immunohistochemical detection of Sendai virus proteins in the olfactory epithelium 3 days post-infection.

The virus-specific staining was observed in the cytoplasm of some olfactory neurons (arrows) in SeV group (A) but none in UV-SeV group (B). (C) Isotype control (200X).

Fig 5. Histological changes by hematoxylin-eosin staining (HE) and immunohistochemical detection of olfactory marker protein (OMP) in the olfactory epithelium post Sendai virus infection.

(A) HE staining showed that there is no obvious loss of integrity of the olfactory epithelium in SeV group compared with UV-SeV group 10 days after infection. Macrophages, neutrophils and lymphocytic infiltration in the olfactory mucosa stained with hematoxylin-eosin is not evident. (400X) (B)Immunohistochemical detection of OMP demonstrated that there is no obvious loss of integrity of the olfactory epithelium in both two groups at each time point. (200X) (C)Isotype control (200X).

Fig 6. Apoptosis and proliferation of cells in the olfactory epithelium at Day 15 post-infection (200X).

Apoptotic cells are marked by green fluorescence mainly located in the middle layer of olfactory epithelium. (A) SeV group. (B) UV-SeV group. Cell proliferation (DNA synthesis marked by EdU) in the olfactory epithelium was located in the basal layer, labeled by red fluorescence. (C) SeV group (D) UV-SeV group.

Fig 7. Unit length labeling index (ULLI) of apoptosis and proliferation in olfactory epithelium in mice infected with SeV or UV- SeV.

(A) Apoptosis(B)Proliferation (Unit/mm).

Fig 8. Morphology and identification of primary cultured olfactory neurons (OSNs).

(A) At 24 hour after plating, some cultured cells had abnormal morphology, whereas others resembled classical OSNs. (40X). (B) Classical OSNs under high-powered magnification exhibit bipoloar morphology with a round cell body, a small thin process suggestive of an axon, and a short, thick process, suggestive of a dendrite (400X). (C) At 3 days, OSNs are identified by expression of type III β-tubulin under confocal laser microscopy, Synaptic connections between neurons were demonstrated. (800X). (D) At 7 days, some cultured neurons expressed neuronal markers (OMP). (800X).

Fig 9. Counts of β-tubulin III positive cells in primary cultured olfactory neurons by flow cytometry.

(A) Counts of β-tubulin III positive cells treated with phosphate-buffered saline only. (B) Counts of β-tubulin III positive cells treated with second antibody-FITC. (C)Counts of β-tubulin III positive cells labeled with β-tubulin III antibody (1:200) and second antibody-FITC. (D)Histogram demonstrate the β-tubulin III positive cells (P3), which represents the olfactory neurons.

Fig 10. Morphology changes of olfactory neurons in vitro after SeV infection.

(A) OSNs treated with UV-SeV at day 3 p.i. (200X). (B) OSNs treated with SeV at day 3 p.i. (200X). (C) OSNs at day 5 p.i. (200X). (D) Single OSN treated with SeV at day 5 p.i. demonstrates a normal bipolar morphology with a long axon under high-power. (400X). (E) Cell-to-cell fusion was observed in the LLC-MK2 cell lines treated with SeV at day 5 p.i. (400X). (F) and (G) SeV F protein was localized mainly in the OSN cell body at day 5 p.i. under confocal laser microscopy. (1:100 MAb dilution, 800X, 200X respectively).

Fig 11. Ca²⁺ responses of primary cultured OSNs after Sendai virus infection induced by olfactory stimulation.

(A) The OSNs in the control group whose fura-4 intensity is displayed in green and purple showed increases in intracellular calcium when stimulated with odor mixture that returned to basal levels after removal of the odors. The brown curve represented a possible untypical olfactory neuron without the bipolar morphology but having the similar cell body with OSNs, which nevertheless had the same response. The bottom curve was recorded from another type of cell with a large nucleus meaning fibroblast and shows no responses. (B) SeV infected OSN did not show calcium Ca²⁺ influx upon stimulation with odor, displayed in green and brown curve. The purple curve represented a cell without typical OSN morphology in the SeV group but having the similar cell body with OSNs, and shows a slow but continuous influx of Ca²⁺ induced by mixed odors and demonstrate a slow decreasing after withdraw of stimulation, which suggests a bit dysfunction in the regulation of cytosolic Ca²⁺.