The ependymal layer of the spinal cord is widely considered a neurogenic niche. Within this layer exists many cell types, but two are of particular interest here; ependymal cells (ECs) and cerebrospinal fluid contacting neurons (CSFcNs). ECs may be latent spinal cord stem cells, becoming active after spinal cord injury. ECs are responsive to GABA1, but the GABAergic source is unknown. CSFcNs have direct contact with the cerebrospinal fluid and are known to be GABAergic2 . We have studied the properties of these unusual cells to understand their function and how they might interact with ECs. We show here that there is 100% colocalisation between polycystic kidney disease 2-like 1 (PKD2L1, a specific marker for CSFcNs) and vesicular GABA transporter (VGAT) in both juvenile (P18-P20, n=3) and adult (P30-P55, n=6) mice, indicating that this is a reliable marker for these CSFcNs.
Glucagon like peptide-1 (GLP-1) is a hormone that is secreted by the gut after food intake and is also produced by a subset of specialised neurons located in the brainstem. GLP-1R agonists exendin-4 and liraglutide neuroprotective and neurotrophic effects in models of Alzheimer’s and Parkinson’s diseases3. Using immunofluorescence labelling we found that most, but not all, CSFcNs possess receptors for GLP-1 (90 ± 1.6%, n=4 adult mice). In calcium imaging experiments using acute spinal cord slices, obtained from VGATxGCaMP6f mice (n=7, P21-P33) terminally anaesthetised with sodium pentobarbital, CSFcNs respond to bath application of 1µM liraglutide through changes in their calcium activity levels, indicating that these receptors are functional.
We then asked if systemic administration of the GLP-1 agonist Exendin-4 could affect levels of cell proliferation in this neurogenic niche area, using the thymidine analogue 5-ethynyl-2’-deoxyuridine (EdU) as a marker of proliferating cells1. Injections of exendin-4 (0.5µM, i.p, every 24 hours for 4 days) in-vivo in healthy adult mice (n=5 control, n=5 test, 9 weeks old) resulted in a significantly lower number of EdU positive cells coexpressing an oligodendrocyte marker, (p=0.026, two sample t-test) and a significantly higher number of (p=0.05) neuronal marker expressing EdU positive cells compared to mice treated with vehicle. There was no overall difference (p=0.19) in the number of newly proliferated ependymal cells between groups, which is consistent with their limited proliferative nature in the physiologically normal state4. In simulated injury conditions using organotypic spinal cord slice cultures, application of 1µM liraglutide significantly increases the proliferation of cells within the ependymal layer after 10 days in culture (p=0.041).
To conclude, we have shown that the GLP-1 receptors on spinal cord CSFcNs are active, influencing the calcium activity of these cells. Furthermore, activation of GLP-1 receptors alters the number of proliferating spinal cells when given in-vivo, and results in an increased number of proliferating central canal cells in ex-vivo experiments. Further work will be done to explore the exact links between these cell types.