Cerebrospinal Fluid-Contacting Neurons (CSFcNs) in the spinal cord are localised around the central canal, with an end bulb projecting into the cerebrospinal fluid (CSF). Based on their morphology and responses to various stimuli in non-mammalian vertebrates, they are hypothesised to play a mechanosensory and chemosensory role by detecting changes in CSF movement and composition. Within the CNS, PKD2L1 (Polycystin 2 Like 1, Transient Receptor Potential Cation Channel) is an ion channel subunit that is a cell-specific marker for CSFcNs and is a known to respond to both mechanical stimulation and changes in pH. The current study uses calcium imaging to examine effects of mechanical stimulation or changing pH on the level of activity of CSFcNs in spinal cord slices from mice in which the calcium sensor GCaMP6f had been expressed in CSFcNs.
Using two photon microscopy at room temperature (20-22 oC), the spontaneous Ca2+ spike rate in CSFcNs was 0.148 ±0.097 Hz (median ± IQR, n = 127 cells, N = 15 mice). The mean calcium spike rate was 3-fold higher when CSFcNs were mechanically stimulated through cell attached recordings of extracellular action potentials (EAPs; 0.43 ± 0.18 Hz, mean ± SD; n = 15, N = 10). Strikingly the different rates we measured with EAPs and Ca2+ imaging are almost identical to previous electrical recordings comparing spontaneous activity in wildtype (0.42 Hz) vs animals lacking the mechano-sensitive PKD2L1 channel (0.16 Hz) (Orts-Del’Immagine et al., 2016). Placement of an electrode against CSFcNs therefore elevates their spike rate and Ca2+ activity, likely due to the mechano-sensitive, large conductance and Ca2+ permeable PKD2L1 channels expressed in these cells.
The effects of pH on calcium signalling in CSFcNs in spinal cord slices were examined at room temperature (20-22 oC) using epifluorescence imaging (n = 24 cells). In initial analyses average spike amplitude decreased significantly (p = 5.3 x10-7; Student’s t-test) when the pH of bathing artificial cerebrospinal fluid was decreased from pH 7.34 (1.38 ± 0.05 DF/F; mean ± SEM) to pH 6.47 (1.07 ± 0.03 DF/F). However, at lower pH the sensitivity of GCaMP6f is reduced and when this was taken into account and calcium peaks normalised (Cho et al., 2017), the average spike amplitude significantly increased to 1.68 ± 0.05 DF/F; p = 2.09 x10-6; Student’s t-test).  This is consistent with increases in firing rate reported in whole cell patch clamped CSFcNs in response to similar pH changes in lamprey spinal cord (Jalalvand et al, 2016).
These data suggest that calcium imaging is an appropriate method to study mechanical stimuli in CSFcNs, but that caution is required when interpreting changes in calcium sensors in response to pH manipulations.  CSFcNs appear to sense and respond to mechanical stimulation and changes to CSF pH in the mammalian spinal cord. How this relates to their function in mammals remains to be determined.