Expression of the ACE2 virus entry protein in the nervus terminalis reveals the potential for an alternative route to brain infection in COVID-19

Abstract

Previous studies suggested that the SARS-CoV-2 virus may gain access to the brain by using a route along the olfactory nerve. However, there is a general consensus that the obligatory virus entry receptor, angiotensin converting enzyme 2 (ACE2), is not expressed in olfactory receptor neurons, and the timing of arrival of the virus in brain targets is inconsistent with a neuronal transfer along olfactory projections. We determined whether nervus terminalis neurons and their peripheral and central projections should be considered as a potential alternative route from the nose to the brain. Nervus terminalis neurons in postnatal mice were double-labeled with antibodies against ACE2 and two nervus terminalis markers, gonadotropin-releasing hormone (GnRH) and choline acetyltransferase (CHAT). We show that a small fraction of CHAT-labeled nervus terminalis neurons, and the large majority of GnRH-labeled nervus terminalis neurons with cell bodies in the region between the olfactory epithelium and the olfactory bulb express ACE2 and cathepsins B and L. Nervus terminalis neurons therefore may provide a direct route for the virus from the nasal epithelium, possibly via innervation of Bowman's glands, to brain targets, including the telencephalon and diencephalon. This possibility needs to be examined in suitable animal models and in human tissues.

Keywords: ACE2; COVID-19; Nervus terminalis; SARS-CoV-2; brain infection; cathepsin; olfactory system.

Fig. 1.

Schematic sagittal section through a mouse head shows the orientation and planes of tissue sections from Fig. 2A, E, I and M. Sections within those planes were used for demonstration of double-immunolabeling and for cell counting. CNS, central nervous system; OB, olfactory bulb; OE, olfactory epithelium.

Fig. 2.

Examples of double immunofluorescent labeling for nervus terminalis neuronal markers GnRH (A, E) or CHAT (M) and ACE2 (B, F, J, N) in the medial region adjacent to the olfactory bulbs as indicated in Fig. 1. Panels A-D and E-H show slightly different focal planes to demonstrate the morphology of the two or three different neurons. Nuclei are stained with Hoechst 33258 (C, G, K, O). Merged images are shown in the last column (D, H, L, P). The neuronal somas labeled with GnRH (A, E) are co-labeled with ACE2 (B, F) as shown after merging (D, H). GnRH positive cells in the ACE2 knock-out mouse (I) are not labeled with ACE2 (J). The majority of cholinergic neurons are not labeled with ACE2 (M, N), as quantified in Fig. 3A. Control sections probed without primary antibodies or with control rabbit IgG had no detectable signal (not shown). Arrows and triangles indicate double-labeled neurons or lack thereof. Scale bars: 20 μm.

Fig. 3

A-F. Quantification of neurons labeled with nervus terminalis markers, virus entry proteins, and verification of the specificity of the ACE2 antibody. A. The large majority of GnRH-positive neurons is also ACE2-positive. In contrast, the majority of CHAT-positive (cholinergic) nervus terminalis neurons lack ACE2-expression. The total number of counted GnRH-positive or CHAT-positive neurons was set at 100%. Error bars represent ± SEM. A t-test shows that the colocalization difference between GnRH- and CHAT-positive nervus terminalis neurons is significant at p<0.0001. For further details, see Table S2. B. Western blot of ACE2 in wildtype (wt) mice and in ACE2 knock-out (KO) mice. The first two lanes (kidney) were loaded with 25 μg total protein, the lanes for olfactory bulb and cerebral cortex were loaded with 60 μg total protein, and probed with the R&D ACE2 antibody. No ACE2 protein was detectable in the ACE2 KO mice, proving that the antibody indeed recognizes ACE2. C-F. Example of GnRH-positive nervus terminalis neurons which are also cathepsin L-positive. C. One GnRH-labeled neuron is marked with a white arrow. D. The same neuron is labeled with the cathepsin L antibody (CatL, white arrow). E. The cell nuclei are stained with Hoechst nuclear dye. F. The three images are merged to show co-localization in the neuron indicated with the white arrow. All scale bars are 20 μm.

Fig. 4.

Peripheral projections of nervus terminalis (NT) neurons and their presumptive relationship with ACE2-expressing neurons in the olfactory epithelium and known SARS-CoV-2 infection. NT neurons innervate blood vessels (BV), Bowman gland (BG) cells, the olfactory epithelium (OE), and contact cerebrospinal fluid (CSF) spaces. Peripheral projections of NT neurons according to Larsell (1950), CSF contacts according to Jennes (1987). Cells expressing ACE2 are indicated in green, including sustentacular cells (SUS) and BG cells. Both of these cell types have been shown to express ACE2 (Bilinska et al., 2020; Brann et al., 2020; Chen et al., 2020; Ye et al., 2020; Zhang et al., 2020; Klingenstein et al., 2021). Cell types that have been documented to be infected by SARS-CoV-2 are indicated with pink asterisks. SARS-CoV-2 localization in SUS cells according to Bryche et al., 2020; Leist et al., 2020; Ye et al., 2020; Zhang et al., 2020; Zheng et al., 2020; de Melo et al., 2021, and in BG cells according to Bryche et al., 2020; Leist et al., 2020; Ye et al., 2020. BG cells furthermore express the protease furin (Ueha et al., 2021) which may facilitate virus entry into those nervus terminalis neurons which innervate BG cells.