Inflammation-related pathology in the olfactory epithelium: its impact on the olfactory system in psychotic disorders

Abstract

Smell deficits and neurobiological changes in the olfactory bulb (OB) and olfactory epithelium (OE) have been observed in schizophrenia and related disorders. The OE is the most peripheral olfactory system located outside the cranium, and is connected with the brain via direct neuronal projections to the OB. Nevertheless, it is unknown whether and how a disturbance of the OE affects the OB in schizophrenia and related disorders. Addressing this gap would be the first step in studying the impact of OE pathology in the disease pathophysiology in the brain. In this cross-species study, we observed that chronic, local OE inflammation with a set of upregulated genes in an inducible olfactory inflammation (IOI) mouse model led to a volume reduction, layer structure changes, and alterations of neuron functionality in the OB. Furthermore, IOI model also displayed behavioral deficits relevant to negative symptoms (avolition) in parallel to smell deficits. In first episode psychosis (FEP) patients, we observed a significant alteration in immune/inflammation-related molecular signatures in olfactory neuronal cells (ONCs) enriched from biopsied OE and a significant reduction in the OB volume, compared with those of healthy controls (HC). The increased expression of immune/inflammation-related molecules in ONCs was significantly correlated to the OB volume reduction in FEP patients, but no correlation was found in HCs. Moreover, the increased expression of human orthologues of the IOI genes in ONCs was significantly correlated with the OB volume reduction in FEP, but not in HCs. Together, our study implies a potential mechanism of the OE-OB pathology in patients with psychotic disorders (schizophrenia and related disorders). We hope that this mechanism may have a cross-disease implication, including COVID-19-elicited mental conditions that include smell deficits.

Fig. 1. Structural alterations in the OB of the IOI mice.

GL glomerular layer, EPL external plexiform layer, MCL mitral cell layer, IPL internal plexiform layer, GCL granule cell layer. A OB size (n = 6 mice per condition): left panel: dorsal view of the OB, and schematic diagram of the OB measurement methods; right panel: quantification of the OB length and width. B DAPI staining of the OB (white) (n = 10 mice per condition): left panel: representative coronal section image of the OB. Scale bar, 500 μm; middle panel: schematic diagram of the OB coronal section layers; right panel: quantification of each area as % of control mice. C Immunohistochemistry with antibody against OMP (green), VGlut2 (red), CCK8 (red), and PGP9.5 (red) with DAPI staining in the GL (n = 6–12 mice per condition): left panel: representative images of the GL. Scale bar, 50 μm; right panel: quantification of the signal intensity of OMP, VGlut2, CCK8, and PGP9.5 as % of control. D Patch clamp recording of spontaneous EPSCs from OB tufted cells: left panel: raw traces; middle panel: averaged frequency; right panel: amplitude. For all cells, the holding potential was −70 mV under voltage clamp mode. ***p < 0.001, **p < 0.01, *p < 0.05.

Fig. 2. Behavioral deficits in the IOI mice.

A Olfactory habituation/dis-habituation test. The mice were subjected to three consecutive exposures to water and two distinct odorants, such as vanilla and banana. Control, n = 9 (4 males, 5 females); IOI, n = 7 (4 males, 3 females). B The schema of PR schedule of reinforcement task using an operant conditioning chamber. C Total number of lever presses, pellets received, and animal’s breakpoints in PR test session were shown. Control, n = 19 (9 males, 10 females); IOI, n = 17 (8 males, 9 females). D Free feeding test. The weight of the consumed standard lab chow (normal food pellet) or high-sugar pellet food in 60 min was measured under free-feeding conditions. Control, n = 14 (6 males, 8 females); IOI, n = 11 (8 males, 3 females). ***p < 0.001.

Fig. 3. Volumetric alterations in OB in FEP patients.

A Representative T1w (upper panel) and T2w (lower panel) images of the OB region of the same subject in the axial (left), coronal (middle), and sagittal (right) views. B Boxplot of total (left), right (middle), and left (right) OB volume between FEP patients and HCs. ***p < 0.001.

Fig. 4. Correlation between OB volume and expression profiles in ONCs.

The first principal component from PCA analysis was used to measure the total expression of a set of genes. A larger value means a higher expression level. A Dotplot of the expression of inflammation-related differentially expressed genes (DEGs) in ONCs versus right OB volume in FEP patients. B Dotplot of the expression of inflammation-related differentially expressed genes (DEGs) in ONCs versus right OB volume in human HCs. C Dotplot of the expression of human orthologs of the inflammation-related genes (IOI genes) in ONCs versus right OB volume in FEP patients. D Dotplot of the expression of human orthologs of the inflammation-related genes (IOI genes) in ONCs versus right OB volume in human HCs. *p < 0.05.